linguistic analysis - Professor Binkert's Webpage - Oakland University

LINGUISTIC ANALYSIS
LECTURE NOTES AND WORKBOOK
Peter J. Binkert
Oakland University

Copyright © The Langtech Corporation 2012
The Langtech Corporation
643 Hazy View Lane
Milford, Michigan 48381-2159
All rights reserved, including those of translation into foreign languages. Except for the
quotation of short passages for the purposes of criticism and review, no part of this publication
may be reproduced or transmitted in any form or by any means, electronic or mechanical,
including photocopy, recording, or any information storage and retrieval system, without prior
permission in writing from the publisher.

PREFACE
This book is not a traditional introductory textbook for linguistics. Although many of the traditional
topics are discussed in the following chapters, the book does not aim at a comprehensive summary
of the discipline. Rather, the intent is to present to the student a unified theory of human language
which demonstrates specifically that the structure of human language is not arbitrary and that it
derives directly from genetically determined faculties of human beings. The essential nature of
human language is not a matter of choice, convention, or whim. Cultural diversity among different
societies and linguistic diversity among different languages and dialects reflect superficial variations
on this basic, biologically determined structure.
The central theme of this text is quite simple: human language reflects the capacities and limitations
of human beings. There are three facts supporting this theme. First, all normal human beings,
regardless of the languages they speak and the cultures they represent, have the same basic biological
makeup. Second, all children learn whatever language they are exposed to in whatever cultural
setting; children are not predestined to learn specific languages. Third, language acquisition in all
children proceeds in a uniform and predictable fashion despite widely varying environmental
conditions.
These facts clearly indicate that the structure of all languages must be based on the nature of human
beings. This is not an original idea; indeed, much of the current work in linguistics in all theoretical
frameworks proceeds from this position. For example, it is basic to Noam Chomsky’s concept of
linguistic universals.
Where this text differs from most introductory texts is in the manner in which the various
subdivisions within linguistics are described. At every point possible, the development of our theme
will be bolstered by specific arguments that have reference to human biology and psychology. All
components of language will be related to each other, rather than presented as discrete units, so that
elements of linguistic structure, change, and variation are integrated. Our discussion of different
languages and different cultures will show how linguistic divergence is constrained by linguistic
universals.
Generally, linguistics departments on university campuses are grouped with the social sciences
(psychology, sociology, anthropology, etc.) or humanities (philosophy, music, art, literature, etc.).
In either of these contexts, linguistics is in a distinctive position to make contributions to the history
of ideas. Linguistic argumentation, that is, the justification of grammars, is a highly developed
methodology that can be used to make predictions about the nature of man and mind. Traditional
justification of theoretical models in the natural and physical sciences derives from experimentation.
The techniques of justification in biology and physics are familiar to every student. In linguistics
also, it is possible to formulate hypotheses of considerable rigor and subject them to scientific
scrutiny that leads directly to their verification or refutation. This book aims to show the
introductory student how this is possible.

More than anything else, this text is designed to disabuse readers of the many misconceptions that
surround the study of language. These fallacies include, among other things, the idea that different
dialects, different cultural patterns, and even different languages reflect different levels of human
competence, as well as the prevalent idea that grammatical structure is haphazard and unjustifiable.
Although it may be difficult to believe at this point, the study of grammar can lead to provocative
and interesting ideas about the nature of human beings and the origin of cultural diversity.
Needless to say, one cannot attain such a high level of generality about language, or anything else,
without attention to detail. Therefore, in the following chapters, students will be introduced to some
of the technical vocabulary of linguistics. The aim is not to memorize facts per se, although mastery
of some details is essential before application can begin; rather, it is to show how rigorous
investigation can lead to meaningful generalizations. Lists of memorized facts rarely stay in people’s
minds for very long, but genuine comprehension of issues does. Moreover, such comprehension can
have a significant impact on one’s life. In this regard, this text aims to satisfy some of the major
objectives of general education, which include, among other things, helping students understand and
master basic techniques for the analysis and synthesis of ideas. Such techniques involve the ability
to gather, organize, and interpret data, to separate what is significant and interesting from what is
irrelevant and trivial, and to formulate hypotheses of real explanatory power. Effective thinking
occurs when one is able to uncover the essential nature of any given problem and to propose
reasonable solutions consistent with available resources.
ii

TABLE OF CONTENTS
Preface. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . i
Table of Contents.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . iii
Introduction: Linguistics and the Job Market. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1
Chapter One: Fundamentals of Linguistics.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23
1.1 Grammatical Characterization and Grammatical Realization. . . . . . . . . . . . . . . . 23
1.2 The Branches of Grammar and Linguistics. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 28
1.3.1 Native Language Acquisition and Maturation. . . . . . . . . . . . . . . . . . . . . . 36
1.3.2 Cerebral Lateralization.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 36
1.4 The Study of Grammar. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 38
1.5 Speech and Writing. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 39
1.6 Linguistic Argumentation. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 40
1.6.1 Diagnostics.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 41
1.6.2 Justifying Structure: Units. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 41
1.6.3 Justifying Structure: Labels. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 44
1.7 Universal Grammar (UG). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 46
1.8 Levels of Adequacy. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 51
1.9 Relating Linguistic Competence and Linguistic Performance.. . . . . . . . . . . . . . . 56
1.10 The Generality of Linguistic Universals. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 60
1.11 Linguistic Universals. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 67
1.12 Language Variation. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 70
1.13 Language and Culture. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 71
1.14 Types of Languages. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 74
1.15 The Evolution of Human Language.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 78
Exercises for Chapter One. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 88
Chapter Two: Phonetics and Phonology.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 93
2.1 Phonetics. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 93
2.1.1 Vowels. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 93
Figure One: the Vocal Apparatus.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 94
Figure Two: English Vowels. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 95
Figure Three: Distinctive Features for English Vowels. . . . . . . . . . . . . . . . . . . . . . . . . . 96
2.1.2 Consonants.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 97
Figure Four: English Consonants, Liquids, and Glides. . . . . . . . . . . . . . . . . . . . . . . . . . 98
2.1.3 Liquids and Glides. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 100

Figure Five: The Major Phonological Categories. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 100
Figure Six: Distinctive Features for English Consonants, Liquids & Glides. . . . . . . . . 102
2.1.4 Review of Issues. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 103
2.1.5 Suprasegmentals. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 105
2.1.6 Syllables.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 105
2.2 Phonology. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 107
Figure Seven: Phonological and Phonetic Representations. . . . . . . . . . . . . . . . . . . . . . 113
2.3 Phonotactics. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 114
2.4 Morphology. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 115
2.5 The Biological Basis of Phonology. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 117
2.5.1 First Set of Observations. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 119
2.5.2 Second Set of Observations. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 121
2.5.3 Third Set of Observations. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 123
2.6 The Phonological and Lexical Components of a Grammar.. . . . . . . . . . . . . . . . 123
2.7 English Spelling.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 125
2.8 Summary of Rule Formalism.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 127
2.9 Notes on Syllables. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 127
Summary of English Sounds and Spelling.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 128
Exercises for Chapter Two.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 130
Appendix A: Answers to Transcription Exercises.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . 141
Appendix B: Phonological Rules.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 146
Appendix C: Slash–dash Notation. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 149
Appendix D: Phonology Problem.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 150
Appendix E: Morphology Problem.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 151
Appendix F: English Vowel Shift. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 152
Appendix G: English Morphology.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 153
Chapter Three: Syntax. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 155
3.1 Linguistic Argumentation Revisited. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 155
3.1.1 Linguistic Argumentation: Sentence Types. . . . . . . . . . . . . . . . . . . . . . . 155
3.1.2 Linguistic Argumentation: Parts of Speech. . . . . . . . . . . . . . . . . . . . . . . 158
3.1.3 Linguistic Argumentation: Phrases.. . . . . . . . . . . . . . . . . . . . . . . . . . . . 159
3.1.4 Linguistic Argumentation: Strategies. . . . . . . . . . . . . . . . . . . . . . . . . . . 161
3.2 Ordering Constraints. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 162
3.2.1 Testing Hypotheses: Reference, Omission, and Placement. . . . . . . . . . . 163
3.2.2 Three Other Diagnostics to Determine Structure.. . . . . . . . . . . . . . . . . . 166
3.2.3 Language and Memory.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 167
iv

3.3 Traditional and Structural Grammar. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 168
3.4 Tense and the Internal Structure of Sentences. . . . . . . . . . . . . . . . . . . . . . . . . . . 174
3.5 Determiners and the Internal Structure of Noun Phrases.. . . . . . . . . . . . . . . . . . 178
3.5.1 Abstract Elements: Ø. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 179
3.5.2 Phrase Structure Rules: First Proposal.. . . . . . . . . . . . . . . . . . . . . . . . . . 183
3.6 The Endocentricity Constraint. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 186
3.7 The Lexicon. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 188
3.8 A Note on Scientific Inquiry. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 191
3.9 X–bar Syntax. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 193
3.10 Generalizing Phrase Structure: S and NP. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 198
3.11 Adjuncts (Modifiers) and Complements. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 200
3.12 Residential Grammar (RG). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 211
Figure Eight: Features for Some English Morphosyntactic Categories.. . . . . . . . . . . . . 227
3.13 Transformational Generative Grammar. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 235
3.14 Nontransformational Generative Grammar. . . . . . . . . . . . . . . . . . . . . . . . . . . . . 238
3.15 Command Relations in Phrase Structure. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 239
3.16 Empty Categories.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 242
3.17 Word Order Variations. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 247
3.17.1 Adverbs in French and Italian. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 247
3.17.2 Nonconfigurational Languages. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 251
3.17.3 The Ordering of V1 Constituents in Italian and Hebrew . . . . . . . . . . . . 254
3.18 The English Auxiliary. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 259
3.18.1 The Facts and Generalizations. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 259
3.18.2 Problems with the Classic TG Analysis.. . . . . . . . . . . . . . . . . . . . . . . . . 262
3.18.3 Resolution. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 270
3.18.4 Summary of Nominals and Verbals in RG. . . . . . . . . . . . . . . . . . . . . . . 273
3.18.5 Parses Illustrating the English Auxiliary. . . . . . . . . . . . . . . . . . . . . . . . . 274
Exercises for Chapter Three.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 282
Appendix A: Answers to Exercises. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 286
Appendix B: Latin Syntax .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 304
Appendix C: Summary of Tree Structures. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 306
Supplement One: Grammar Review. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 309
Syntactic Categories (The Parts of Speech). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 309
Syntactic Constructions. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 311
Morphology. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 312
Some Inflectional Categories. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 313
v

Supplement Two (Part I): Typological Classification. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 314
Supplement Two (Part II): Typological Classification. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 315
Supplement Three (Part I): Historical Classification. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 316
Supplement Three (Part II): Historical Classification.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 317
Supplement Four (Part I): Languages of Africa and the mid East. . . . . . . . . . . . . . . . . . . . . . . . 318
Supplement Four (Part II): American Indian Languages. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 319
Supplement Four (Part III): Indo–european Languages.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 320
Supplement Four (Part IV): Languages of East Asia & the Pacific.. . . . . . . . . . . . . . . . . . . . . . 321
Supplement Four (Part V): Other Languages of Europe and Asia.. . . . . . . . . . . . . . . . . . . . . . . 322
Supplement Five: Languages and Language Groups. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 323
Supplement Six: Some Alphabets. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 325
Supplement Seven: Sample Parses from the Langtech Parser.. . . . . . . . . . . . . . . . . . . . . . . . . . 327
Sample Parses Illustrating Basic Sentence Patterns.. . . . . . . . . . . . . . . . . . . . . . . . . . . . 327
Supplement Eight: Sample Parses from the Langtech Parser. . . . . . . . . . . . . . . . . . . . . . . . . . . 365
Sample Parses Illustrating Noun Phrases.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 365
Supplement Nine: Sample Parses from the Langtech Parser.. . . . . . . . . . . . . . . . . . . . . . . . . . . 369
Empty Categories: Ø, [u], and [e].. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 369
Supplement Ten: Sample Parses from the Langtech Parser.. . . . . . . . . . . . . . . . . . . . . . . . . . . . 372
Sample Parses Illustrating Verbals. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 372
Outline of Technical Terms. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 381
Residential Grammar (RG) Summary. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 387
References. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 397
Sample Test One. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 403
Sample Test Two. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 414
Sample Test Three. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 432
Sample Test Four. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 450
Index. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 468
vi

INTRODUCTION: LINGUISTICS AND THE JOB MARKET
There is no doubt about it: high school graduates are having a very difficult time finding a job that
pays well and gives them some prospects for advancement. A recent article in the New York Times
(June 6, 2012), summarized the current situation as follows:
“For this generation of young people, the future looks bleak. Only one in six is working full
time. Three out of five live with their parents or other relatives. A large majority – 73 percent
– think they need more education to find a successful career, but only half of those say they
will definitely enroll in the next five years... For this group, finding work that pays a living
wage and offers some sense of security has been elusive.” (More Young Americans out of
High School Are Also out of Work, C. Rampell)
Even among college graduates, the competition for jobs in today’s market is intense. According to
an Associated Press news story on April 25, 2012, half of all new graduates are either jobless or
underemployed:
“In the last year, [graduates] were more likely to be employed as waiters, waitresses,
bartenders and food-service helpers than as engineers, physicists, chemists and mathematicians
combined (100,000 versus 90,000). There were more working in office-related jobs such as
receptionist or payroll clerk than in all computer professional jobs (163,000 versus 100,000).
More also were employed as cashiers, retail clerks and customer representatives than engineers
(125,000 versus 80,000).
“According to government projections released last month, only three of the 30 occupations
with the largest projected number of job openings by 2020 will require a bachelor’s degree or
higher to fill the position – teachers, college professors and accountants. Most job openings
are in professions such as retail sales, fast food and truck driving, jobs which aren’t easily
replaced by computers.” (Recent College Graduates Finding Few Job Prospects, Hope Yen,
Associated Press, www.detroitnews.com/article/20120425)
Despite these numbers, college graduates still fare better than those without degrees according to the
Bureau of Labor Statistics (Yes, A College Education Is Worth the Costs, Rodney K. Smith, USA
Today, 2011, www.usatoday.com/news/opinion/forum/story/2011-12-06/). Unemployment rates
decrease as education increases (Bureau of Labor Statistics, 2010, www.bls.gov):
•14.9% of those without a high school diploma
•10.3% of those with a high school education
•7% of those with an associate degree
•5.4% of those with a bachelor’s degree
•2.4% of those with a professional degree
•1.9% of those with a doctoral degree.

2
Also, educational attainment correlates with income. Here’s the average weekly income for those
who have jobs (Bureau of Labor Statistics, 2010, www.bls.gov):
•$444 for those with less than a high school degree
•$626 for those with a high school degree
•$767 for those with an associate degree
•$1,038 for those with a bachelor’s degree
•$1,550 for those with a doctoral degree
Clearly, making the most out of a college education is crucial. Before choosing a program of study,
students should spend some time surfing the internet and googling such search items as “what
employers want,” “skills needed for the workplace,” “trends in the marketplace,” and so on. There
are many myths and traditions surrounding options like the choice of a major in college. For
example, students often think that it will be impossible to get a job in a specific discipline unless
they study that discipline in depth in college. They cannot, for instance, become a manager unless
they are a business major. Actually, careers often have little to do with specific majors in college.
The Employee Benefits Research Institute, reports that the average tenure for all wage and salary
workers age 25 or older was 5.2 years in 2010, compared with 5.0 years in 1983 (www.ebri.
org/publications, Vol 31, No 12, 2010). That means that in the United States for nearly three
decades workers generally stayed at the same job for about 5 years. As a result, if individuals work
for 50 years, they can expect to have about 10 different jobs during their working careers. Different
sources report slightly different statistics, but surfing the internet clearly shows that frequent job and
career change is the norm in today’s market.
Such change is in stark contrast to the situation in the past when individuals not only kept the same
job for their entire lives but also had the same job as their parents and grandparents whether they be
policemen, farmers, teachers, doctors, coal miners, etc. Nowadays, on the other hand, people can
expect to change not only the company they work for but actually the type of work they do, that is,
change their careers. Thus, one of the first things students should consider in choosing a program
of study is finding one that can provide them with the skills to be adaptable, flexible, and mobile in
a global and unpredictable economy. It is unlikely that the job a student finds after graduation will
be directly linked to the course content of a specific undergraduate major.
Problem Solving Skills Needed for the Workplace
What exactly are the skills that will prepare a student to be adaptable, flexible, and mobile? One way
to answer that question is to look at the kinds of skills that various employers look for. Again,
surfing the internet, one finds that employers across the board look for many of the same skills in
their employees. Chief among these skills is the ability to solve problems. For example, one survey
of 301 executives in Fortune 1000 companies found that 99% of the executives think that the ability
to solve problems and to think critically is either an absolutely essential or very important skill for

college and career readiness (The MetLife Survey of the American Teacher: Preparing Students for
College and Careers, 2011, p. 21).
Other studies also emphasize the importance of creative problem-solving skills. “Employers know
that in business, the chessboard changes daily. As soon as we think all is fine, the economy changes
or the competition makes a surprise move and the company’s own strategy must change,” said Mark
Stevens, author of Your Marketing Sucks (Crown Business, 2003) and CEO of MSCO, a global
marketing firm. “A person who gets locked into a set way of doing things finds it difficult or
impossible to adjust. They are a drag on the business as opposed to an asset for it. [An employee
must know] how to tackle challenges and opportunities in a way no one will find in a textbook.”
(CNN: Top 10 Reasons Employers Want to Hire You, Rachel Zupek, www.CareerBuilder.com,
2011).
A multitude of sources including magazine articles, newspapers, on-line reports, scholarly papers,
and books all suggest that critical thinking and problem solving skills are essential for success in
today’s job market. There is also general agreement on what such skills entail. Two on-line sources
define the necessary skills as follows:
(1) Quintessential Careers: What Do Employers Really Want? Top Skills and Values
Employers Seek from Job-Seekers, Randall Hansen and Katharine Hansen
(www.quintcareers.com, 2012).
a. Analytical/Research Skills. Deals with your ability to assess a situation, seek
multiple perspectives, gather more information if necessary, and identify key issues
that need to be addressed. Highly analytical thinking with demonstrated talent for
identifying, scrutinizing, improving, and streamlining complex work processes.
b. Problem-Solving/Reasoning/Creativity. Involves the ability to find solutions to
problems using your creativity, reasoning, and past experiences along with the
available information and resources. An innovative problem-solver can generate
workable solutions and resolve complaints.
(2) UC Davis, Human Resources: What do Employers Want from Employees? (www.hr.
ucdavis.edu/sdps, 2012).
a. Analytical Thinking – The ability to generate and evaluate a number of alternative
solutions and to make a sound decision regarding a plan of action.
b. Researching – The ability to search for needed data and to use references to obtain
appropriate information.
c. Organizing – The ability to arrange systems and routines to streamline work and
maintain order.
3

4
It is important to note that sources like the above do not stress the need for workers to have extensive
knowledge of particular facts in order to be employable. Actually, nowadays it is rather easy to get
information about almost anything very quicky using the internet. What workers do need instead
is the ability to organize, analyze and integrate the information they find so that they can make viable
proposals to solve problems.
The Information Explosion
In addition to finding a program of study that emphasizes critical thinking and problem solving
skills, students should also consider whether the program will prepare them to deal effectively with
the nature of the modern workplace and the specific problems they will be asked to solve on the job.
First, in today’s workplace, the amount of information that the average corporate worker must digest
is enormous. Consider, for example, just the issue of emails which must be looked at, evaluated, and
often answered before work on a problem can even begin. By 2014, it is projected that the average
number of legitimate corporate emails a worker can expect to receive per day will be 65; the average
number sent will be 39 (Email Statistics Report, 2010, Sara Radicati, The Radicati Group, Inc.,
April, 2010). Adding instant messaging and social networking, it becomes clear that workers must
handle a massive amount of data.
Again, different sources provide different statistics; yet, an undeniable fact is that the internet has
fostered an explosion in the amount of information that is available virtually instantaneously. “Each
year the world produces 800 MB of data per person. It would take approximately 30 feet of shelf
space to hold that amount of information in books. The amount of data produced each year would
fill 37,000 libraries the size of the Library of Congress” (How Much Information Is There? Mark
Shead, 2007; www.productivity501.com). Recent lapses from agencies like Homeland Security, the
CIA, and the FBI have shown alarmingly how easy it is for important communiques to get lost in the
avalanche of data even though such agencies employ experts in surveillance. While national security
is rarely at stake in office emails, the point is clear: to be successful, modern workers need to be able
to digest, sort, organize, and prioritize massive amounts of information efficiently and expertly.
The Problems Facing Today’s Workers
Another factor that students should consider in selecting a program of study involves the nature of
the problems that workers face on a routine basis in the 21st century. Today’s problems are
invariably complex, multifaceted, and subtle. It used to be the case that one could formulate a
problem into a relatively well-defined and stable statement, that one would know when a satisfactory
solution was reached, and that one could clearly evaluate the solution objectively as being good or
bad and right or wrong. That is no longer the case for many problems facing society, government,
and businesses such as the following:

(3) a. Should colleagues be disciplined if they are caught using drugs?
b. How should we deal with crime and violence in our schools?
c. How can we make air travel safe from terrorism?
d. Should people be fired if it is discovered that they are illegal aliens?
e. How can we improve the gas mileage of family cars?
f. Should same sex partners have medical benefits like married couples?
g. How should we improve the language programs in US schools?
As these examples indicate, today’s problems are frequently ill-defined and ambiguous. For
example, in (3a), the question raises other questions like the following: How much drugs, What
kinds of drugs, Are they prescription drugs, Is alcohol included, Is it a first offense, and so on.
Often, today’s problems are associated with strong moral, social, political and professional issues.
For example, it is not possible to talk about preventing violence in schools (3b) or about making air
travel safer (3c) without also considering whether increased surveillance will destroy individual
freedoms or whether certain policies will lead to stereotyping groups on the basis of color or
ethnicity. In some instances such as (3d), there is little consensus about what the problem is, let
alone how to resolve it. Overall, today’s problems are rarely fixed and stable; they involve sets of
complex, interacting issues developing in a dynamic social context. Often, new forms of problems
emerge as a result of trying to understand and solve the original issues (Wicked Problems - Social
Messes, Tom Ritchey, Springer, 2011). For example, proposals about gas mileage (3e) invariably
raise issues about pollution, energy policies, dependence on foreign oil, the dangers of fossil fuels,
the materials from which vehicles are made, and a host of other matters. As a result, in dealing with
today’s problems, there is generally no right answer, and proposed solutions are not good or bad, but
better or worse. To be successful, modern workers must be able to juggle competing proposals to
determine which alternative satisfies the most factors or constituents in the most effective way.
Abstract Thinking
Summarizing, today’s workers need to be good problem solvers and creative thinkers; they need to
be able to organize, analyze, and interpret massive amounts of information and to make proposals
that consider all the stakeholders involved and all the possible consequences that might arise. It is
with those considerations in mind that students should approach the courses they take at university
and the exercises and problems they are asked to complete in those courses. Frequently, students
fail to see the connection between solving a specific problem in a particular course and the problems
they will face outside of the classroom. In fact, there are strategies and procedures that are useful
in attacking all problems, and teaching students how to make such connections is an objective of
education. Students often miss those connections when they focus instead on what appear to them
to be irrelevant facts about useless issues.
The ability to form such connections is vital to success both in college and in the workplace.
Consider the following scenario. A young boy about two years old is walking with his grandfather
in the country near a dirt road when the grandfather says, “Alex, don’t go in the road. You could get
5

6
hurt.” Later that day, the family visits some relatives who live in a suburban subdivision. Alex runs
into the paved street in front of the house to get a ball. The grandfather, very upset, says to Alex, “I
told you not to go into the street.” Alex looks at his grandfather as if to say, “No, you didn’t. You
told me not to go into the road.” At the age of two, children cannot see the connection between a
dirt road in the county and a paved street in a subdivision. Their interpretation of the world is
concrete and literal. It will take a few more years before they are capable of the abstract thinking
required to make such connections.
Abstract thinking is a level of thinking that is removed from concrete, literal, here-and-now
situations. Abstract thinking assesses a variety of specific examples of phenomena and unites them
into a generalized concept that embraces all of the individual instances. Abstract thinkers are able
to put together seemingly disconnected issues and see a larger picture. As an example, consider
another scenario. Two people are applying for a job with an automobile manufacturer. The
interviewer mentions that one of the industry’s most pressing problems is gas mileage; consumers
are concerned about the cost of fuel. The interviewer asks the applicants general questions about
customer satisfaction, options available on new models, electric cars, hybrids, changes needed in the
industry, and so on. During the discussion of options, one applicant volunteers that she is unhappy
with the little “donut” spare tires that come with cars because a person cannot drive very long with
them and must rush to have the flat tire fixed; she believes that cars should come with full spare tires
the way they used to. The other applicant responds by saying that a full spare tire weighs more than
a donut tire. The increased weight will make the car heavier and, therefore, reduce gas mileage.
Who gets the job?
Practice
Students don’t always see the connections between the various exercises that they are assigned to
complete in various courses and the “real world.” But, learning how to solve problems, think
abstractly, analyze data, propose hypotheses, and argue for a particular position – all of the skills
required for today’s workplace – require exactly the kind of practice one gets working on different
exercises in different courses. Practice is essential. Even experienced people practice and train on
a daily basis so they can maintain their expertise and develop professionally. Ballet dancers spend
long hours every day at the barre practicing plies and tendus so they can become proficient. Boxers
train every day; authors write every day; scientists experiment; philosophers ponder; golfers golf.
It’s often said that it takes 10,000 hours of practice to achieve mastery in any skill. Students cannot
expect to become successful thinkers if they don’t have practice solving a variety of different
problems as often as possible. Students generally think that they should be able to solve any problem
in twenty minutes, and, if they can’t, something must be wrong with them or their teachers. That is
incorrect as the above facts about today’s problems clearly indicate. Developing skill in problemsolving
like everything else takes practice. When students don’t see the purpose of taking a
particular course to satisfy a particular general education requirement, they need to try to see the
larger picture. It is exactly that larger picture – the relationship between roads and streets, the

elationship between gas mileage and tire size – that is crucial to adaptability, to flexibility, to
mobility, and to success.
Learning by Doing
Students in the United States generally believe that teachers should supply them with facts,
illustrations and the specific means by which they can successfully complete an assignment. There
is an excellent reason for that belief: teachers in the United States generally supply students with
facts, illustrations and the specific means by which they can successfully complete an assignment.
Unfortunately, the problems that one faces in life and at work do not come with instructions for their
solution. In short, what happens in American classrooms frequently has little bearing on what
happens outside them. When evaluation measures depend on memorization and replication, there
is little motivation for creativity, imagination, and objective thought. Consequently, high school and
college graduates often have great difficulty applying what they have learned in new situations. They
have difficulty with abstract thinking. Complexity often paralyzes them because they have few tools
to break down problems into manageable parts. They have not had enough practice doing so.
Students need to have practice and experience dealing with unknowns, managing problems that have
no clear answers, and evaluating competing approaches to find the best solution for the
circumstances when no single approach is completely satisfactory. As the above discussion on
today’s problems shows, those are the skills that are needed in today’s workplace. A teacher’s
function is not to supply students with the right answers or even with the formula to arrive at the
right answers. Rather, teachers need to act more as guides, helping students figure out how to solve
a problem and discover by themselves the principles that best describe the data under investigation.
In that way, students become better able to cope with new and different problems, which will always
occur. It is counterproductive for students to complain that teachers don’t give them the right
answers or the precise way to find the right answer. After all, teachers will not be following students
around for the rest of their lives pointing out the right answers to problems that arise.
Most students pursue a college degree knowing that it will help them in the job market. They know
that college graduates have more and better job opportunities than people without college degrees.
Furthermore, there is abundant research indicating that “college graduates are healthier, contribute
more to their communities, and raise kids who are better prepared academically” (How Much Is a
College Degree Really Worth, Kim Clark, US News, 2008, usnews.com/education). Having decided
to pursue a college education and chosen a major, an immediate concern for students is finding
courses and programs (concentrations, minors, etc.) which will make them most marketable, that is,
ones which will give them practice in the specific skills employers are looking for. As a result,
students should select classes which will help them become innovative and creative thinkers who
can make proposals and solve problems on their own. In that regard, courses in linguistics are
especially appropriate. All linguistics courses focus on gathering, organizing, and analyzing data,
setting up hypotheses to account for the data, and proposing and defending the most generalized and
7

8
effective solutions. By investigating problems in a variety of languages, students of linguistics get
practice developing just those skills that employers are looking for.
A Linguistic Example: English Syntax
As an illustration, consider two approaches that might be used in teaching some basic facts of
English syntax, which is the study of the way sentences are constructed. This illustration might
appear to be absolutely irrelevant to a student’s life or to the “real world” outside of the classroom,
but we will see that it isn’t. One approach to syntax begins with definitions like those in (4) which
teachers ask their students to memorize.
(4) a. The subject of a verb is the person or thing that performs the action in the verb.
b. The object of a verb is the person or thing that receives the action in the verb.
Given (4), students are then asked to find the subjects and objects in sentences like (5).
(5) The dogs frightened the little girl.
Applying (4) to (5), students identify the dogs as the subject phrase because they are causing the fear,
and the little girl as the object phrase because she is experiencing the fear.
The approach, therefore, seems successful. The students have memorized some formulas, the
definitions in (4), applied them in exercises, and learned a tool that can be reused elsewhere. The
teachers have done their job in supplying the formula for completing the assignment.
The difficulty is that English also has sentences like (6).
(6) The little girl feared the dogs.
In (6), the dogs are still causing the fear and the little girl is still experiencing the fear so it seems,
according to (4), that the dogs is still the subject phrase and the little girl is still the object phrase.
If that is correct, then we cannot say that subjects precede verbs in English and objects follow, which
seems to be the case in most sentences. So something is wrong. Since students do not know how
or why the definitions in (4) were proposed in the first place, they generally have no idea what to do
when the definitions seem to fail as they do in sentences like (6).
Now notice what is going on here at an abstract level. We have some data, which happen to be
sentences in English. We have a hypothesis which describes the data, namely, (4). We get some
further data which the hypothesis can’t handle. Thus, we must revise the hypothesis and look for
another proposal. This is exactly what happens frequently in the workplace. Consider, for example,
another scenario from the auto industry.

Several years ago, Ford Motor Company produced a van called the Mercury Villager. This van
originally only had a sliding door on the passenger side. Marketing analysts at Ford had determined
that parents would not want to have a sliding door on the driver’s side because their children might
get out of the van on the side of traffic. The other major auto manufacturers had vans with sliding
doors on both sides, because their marketing analysts had determined that people would find it too
inconvenient going to one side of the car to deposit items like groceries and then walking to the other
side of the car to get into the driver’s seat. Having sliding doors on both sides of the van was
considered a plus. Ultimately, Ford modified its design and offered a new Villager with sliding
doors on both sides. “A sliding door on the driver’s side of the vehicle [was] the subject of popular
demand from customers” (carpartswholesale.com/cpw/mercury~villager~parts.html). At an abstract
level, this problem has the same characteristics as our grammatical problem. Ford began with data,
namely, its marketing analysis of what it seemed parents wanted in a van. Ford then designed a car
accordingly. But further data indicated that the design did not match all the needs of its customers.
So, Ford began to produce a new van which did match the additional customer information.
Returning to our grammatical example, consider now an alternative approach to describing subjects
and objects which begins by looking at the data, that is, good sentences like those in (5), (6), and (7),
as well as bad sentences like those in (8), where the asterisk means that the sentence is
ungrammatical.
(7) a. They frightened her.
b. She feared them.
(8) a. *Them frightened she.
b. *Her feared they.
Given data like the above, it is clear that words like she and they must be distinguished from words
like her and them. All native speakers of English know this fact unconsciously whether or not they
have studied English grammar in school. They know that sentences like (7) are good and those like
(8) are not; so they say (7), not (8), even though they usually cannot explain why. As a result,
distinguishing the two groups of words is a necessity for a speaker of English, not a convention or
a convenience. The distinction is part of the English language; it is not something that teachers of
grammar made up.
Since the two groups of words exist, suppose we give them each a name. Notice that there is nothing
odd about this: we have names for all kinds of groups (cars, animals, food, laws, sports, etc.).
Suppose we call words like she and they “subject pronouns” and words like her and them “object
pronouns.” With these new names, we can now succinctly state the distinction between subject
phrases and object phrases, also a necessity if one wants to be a speaker of English. Consider (9).
(9) a. Subject phrases are specified by subject pronouns (she, they, etc.)
b. Object phrases are specified by object pronouns (her, them, etc.)
9

10
Given (9), students attempt to replace phrases in sentences like (5) and (6) with pronouns. The result
is always sentences like (7), never (8). Therefore, students, like native speakers, immediately know
what the subjects and objects are: if a phrase can be specified by a subject pronoun, then it is the
subject; if it can be specified by an object pronoun, then it is an object. In addition, students learn
that the principles in (9) are motivated by facts about the English language, specifically the
distribution of the two groups of pronouns.
The approach just illustrated gets to the heart of the matter and the result is worth repeating.
Grammatical facts are not the result of convention, whim or convenience. They are a necessity;
indeed, they are a biological necessity. The sentences of every human language are broken up into
phrases like subject and object because the human brain cannot process unstructured material very
well. Try, for example, to recall the numbers in (10) after just one reading.
(10) 1 - 4 - 9 - 2 - 2 - 0 - 0 -1 - 1 - 8 - 1 - 2
Now try to recall the same numbers in (11).
(11) 1492 - 2001 - 1812
The numbers in (11) are much easier to recall and process because they have structure. In a similar
way, the sentences of human languages must be composed of structured units like subject and object.
Unstructured sentences without phrases are incomprehensible to human brains. Just try processing
the last sentence backwards. In short, languages have grammar, because human biology demands
it. Phrases which teachers of grammar arbitrarily call “subject” and “object” would exist even if
there were no teachers around to name them and describe them.
Again, we need to take stock at this point to see what we have done. We began with a relatively
specific matter, namely, how to identify subjects and objects in English sentences. We then
determined the following: first, that subjects and objects are phrases; second, that phrases are
structures; third, that the sentences of all languages have such structures; fourth, that sentence
structure is an essential property of language; and fifth, that the properties of language are
biologically determined. In short, the study of grammar is essentially a branch of biology. The
transition here has been a bit abrupt to be wholly convincing; in the discussion below, the argument
will be justified in further detail. At this point, however, it is possible to understand the
methodology and to see what the overall objective is. Specifically, we are attempting to find the
most generalized and comprehensive way to account for the data before us. At an abstract level, that
is exactly what today’s workers must do on the job.
The two approaches mentioned above have been called deductive and inductive. In the deductive
approach, one begins with the principle (rule, theory, definition, etc.) and tries to apply it to the data.
We began with (4) and applied it to (5). In the inductive approach, one begins with the data and tries
to discover what the principle (rule, theory, definition, etc.) is. We looked at good sentences like (7)
and bad sentences like (8) and then formulated (9).

Since the principles in (9) are objective and explicit, they are verifiable. Testing them with other
data reveals that they are, in fact, more robust than the definitions in (4), which fail in many cases:
(12) a. The stewardess is cooking the meals. She is cooking them.
b. The meals are cooking. They are cooking.
(13) a. The laundress is ironing the shirts. She is ironing them.
b. The shirts iron easily. They iron easily.
(14) a. The waitress tasted the potatoes. She tasted them.
b. The potatoes tasted fine. They tasted fine.
(15) a. The girl broke the windows. She broke them.
b. The windows broke. They broke.
The inductive approach supplies students with an exercise in problem solving, critical thinking, and
objective analysis. It has the potential of uncovering important generalizations like the principles
in (9) and of helping students understand that such principles are justifiable and, in fact, inevitable
when they are driven by an empirical investigation of the data. As a result, rather than learning
something by rote, students can develop skills for life-long learning, skills that can help even when
there are no teachers available.
In the above illustration, if students focus on the terms “subject” and “object,” they might conclude
that the exercise is irrelevant to their life in the real world, that knowledge of grammatical terms will
not help them on the job. Of course, that is probably correct: very few promotions depend on
knowledge of grammar. But the criticism misses the point. The above illustration is an exercise in
problem solving, in learning to think abstractly, and in formulating robust hypotheses. The solution
to any problem begins with collecting, sorting, and organizing the data, no matter what the data
involve. That step is followed by attempting to find generalizations and connections to account for
the data, principles that will explain the data. In turn, those principles need to be tested on new data
to see if they are robust. When they fail, alternative solutions need to be tried and retested to see if
they are successful. In the end, one produces the best possible proposal. The people who produce
the best proposals are the ones who advance in their careers.
There is a TV commercial which asks the following question: “How can rice production in India
affect wheat output in the U.S., the shipping industry in Norway, and the rubber industry in South
America?” (T. Rowe Price ad; www.youtube.com/watch?v=Ae5BH5KdcI0) The underlying
message of the commercial is that companies, workers, and people in general must be able to make
such connections, because understanding connections that are important and discovering connections
that no one has noticed before are crucial in finding solutions to problems. It is, therefore, essential
that students make every effort to try to understand how the exercises they are asked to complete in
their courses do, in fact, apply to their lives.
11

12
Initially, many students find an inductive approach frustrating; they would prefer that teachers tell
them what to memorize and then test them on their recall. Rather than trying to understand why one
proposal is better than another, they just want to know which one will be on the next test. But
memorization is not learning, and knowledge is not constant. Further, people do not succeed in a
career just by knowing a lot of existing facts. We discover and uncover new phenomena every day
which require changes in existing explanations, theories, practices, and methods. Education must
emphasize the fact that there often are no right answers, only what we know at any given moment.
The purpose of education is to bring students to the point where they can continue to learn on their
own without teachers to guide them.
In short, good teachers strive to make themselves unnecessary by making their students self-reliant.
If a few courses that students take during their education attain that outcome, then students have a
lasting model for life-long learning and for greater adaptability in the workplace. For that to happen,
students need to be given exercises and problems that are new and different from what they have
practiced before. It is useful to remember that the average workers will have ten different jobs
during their working careers. Memorization of a lot of facts is not going to help students cope with
a fluid and variable job market, one that is evolving so fast that it is difficult to keep up with the
changes. The more practice students have with the greater varieties of problems, the broader their
perspectives become. In turn, this broader experience will help them be more adaptable, more
flexible, and more mobile. Although a diploma by itself will give students an edge in securing
employment, they will fare much better over time if that diploma includes courses that give them the
skills and perspectives they need to be responsive to the dynamic nature of today’s job market.
Another Linguistic Example: English Phonology
Let us consider next another linguistic example, this time from phonology, the study of the sounds
of language. Verbs in English come in a variety of tenses such as present tense (walks), past tense
(walked), and future tense (will walk). When people are asked how to form the past tense of regular
English verbs, they generally respond with a reference to written English and say something like,
“Add an ed to the end.” Such a rule cannot be part of the natural grammar of any language.
Children know how to correctly produce the past tense of regular verbs long before they learn to read
and write, let alone spell correctly. Indeed, some adults are illiterate; still, they know how to form
the past tense of regular verbs. Writing is based on convention, and learning to write is optional.
There are hundreds of languages which have never been written down; conversely, there is no natural
human language that exists only in written form. These are facts; they form part of our corpus of data
that must be accounted for. They indicate that no language has any natural grammatical rule based
on writing. Our discussion of English syntax above lead to a hypothesis that the rules of natural
grammar are based on human biology. Language is a product of the human language apparatus,
which includes the brain and the organs of speech and hearing. It is important to investigate whether
this hypothesis can be sustained in other aspects of language like phonology.

Now notice that there are actually three ways to pronounce the past tense of regular English verbs.
First, the past tense is pronounced as the sound [t] in a verb like race (They raced from the house).
Second, it is pronounced as the sound [d] in a verb like raise (They raised their hands) or raze (They
razed the building meaning ‘They demolished the building’). And third, it is pronounced as a
separate syllable with the reduced vowel [] before [d] in a verb like rate (They rated the movies)
or raid (They raided the refrigerator). So, we have the following three possibilities which clearly
show that spelling does not determine the correct form of the past tense:
(16) a. [t] raced and also coped, hiked, laughed, ached, etc.
b. [d] raised, razed and also combed, hugged, loved, aged, etc.
c. [d] rated, raided and also coded, hunted, loaded, aided, etc.
It is also clear from the above examples that the correct past tense for any given regular verb ([t], [d],
or [d]) cannot be random. Young children typically try to change irregular verbs into regular verbs,
saying things like I goed there. Importantly, children make up such forms without ever having heard
them. No matter how inelegant an adult’s speech is, no adult would ever say something like I goed
there yesterday. Further, if go were a regular verb, its past tense would have to be pronounced goed,
and children’s spontaneous production of forms like goed indicates that they know that. All this
indicates clearly that children do not acquire their native language by simply imitating the people
around them; there must be more to language acquisition that imitation.
Second, when speakers coin new verbs, they always make them regular and pick the appropriate past
tense marker from among the three possibilities given in (16). Consider, for example, the verb
material girl in a sentence like Madonna has material girled her way to superstardom. Notice that
the [d] variant in example (16b) is the only possible option. This means that when the verb material
girl was first used, its past tense had to be pronounced with a [d] and not either a [t] or an [d]. It
is worth noting here that people, including children, invent new words all the time and, when they
do, they always make them regular with the expected variations. Consider the following examples:
(17) a. I ketchuped up my French Fries. (A five-year-old child)
b. She two-footed that landing. (A TV commentator on figure skating)
c. We clearanced those dresses. (A department store sales person)
Some new past tense verbs that have recently become common include transitioned, decontented,
googled, texted, friended, etc. The essential fact is that native speakers automatically understand all
these words the first time they hear them.
If the form of regular past tenses is not random and the determining factor is not spelling, then there
is an obvious question: What is it that governs the formation of the past tense of regular verbs? To
answer that question, one must first gather the necessary data to see exactly where each of the three
variants for the past tense occurs. Note that this step is exactly the step one must take on the job
when trying to solve any problem, namely, one must gather and organize the data. Doing so in the
present case yields data like the following (ignore the spelling and listen to the sounds):
13

14
(18) a. [t] occurs in verbs ending in the sounds [p] (hyped), [k] (cracked), [f] (cuffed), [s]
(kissed), etc.
b. [d] occurs in verbs ending in the sounds [b] (robbed), [g] (shrugged), [v] (moved),
[z] (cruised), etc.
c. [d] occurs in verbs ending in the sounds [t] (granted) and [d] (guarded)
Testing and verifying the data in (18) is the next step. One looks at as many different regular verbs
as possible to see if the statements are correct. Such investigation reveals that the statements, in fact,
are correct; however, the statements are no more than a list, and a list does not explain why the
sounds are distributed as they are. In other words, why is [p] in (18a) and not (18b) or (18c)? These
are very important steps in proposing solutions to any problem: one must not only gather, organize,
and verify the data; one must explain why the data fall out the way they do if one is going to put forth
the most robust proposal about the data. Although this is a theoretical example about English
phonology, one can still discuss its effectiveness as a matter of cost. The analysis in (18) is not very
cost-effective theoretically because it reveals no generalizations. It is not a very good way to account
for the data because it is merely a list. We need to look for common features among the items in the
list to see why they are in one list and not the other. We need to look for generalizations. To see
what this means in a situation one is more likely to encounter in business, consider the following
example.
In the business world, cost is often a primary consideration. The choice of one car design over
another will often depend on its cost-effectiveness. It is most cost-effective for a manufacturer to
vary only the exterior of various models and to use the same underlying mechanical and electrical
systems, the same chassis, the same engines, etc. If every vehicle in a manufacturer’s fleet is built
in a completely different way from every other vehicle and there are no common features, costs will
increase. Thus, manufacturers look for designs that can be generalized over many different specific
vehicles. It is not an accident that many vehicles look alike.
Similarly, in designing homes in a subdivision, builders will use the same underlying floor plan and
only vary the exterior elevations because that cuts down on costs. Architects will put bathrooms on
the second floor above bathrooms on the first floor rather than on opposite sides of the house because
that design is more cost effective. If you ask the architect why the upstairs bathroom is where it is
and not somewhere else, the architect can give you the reason and that reason probably has to do
with cost.
In designing vehicles and buildings, engineers and architects will look for ways in which they can
utilize as many of the same features as they can in all their products as a cost-saving measure. They
will try to generalize. Of course, the same is true not only in the automobile and construction
industries, but other businesses as well. At an abstract level, that is exactly what we need to do with
the data in (18). We need to look for generalizations. Specifically, we need to determine why the
data in (18) fall out the way they do. What do the members of each group have in common?

Since language is a product of human biology, we can expect that the formation of the past tense is
governed by a specific set of principles which are directly related to the nature of the human vocal
apparatus, the part of human anatomy concerned with producing sounds. That expectation turns out
to be correct. To see this, consider first the verbs race and raise. Notice that the final sound of the
verb race is [s] and the final sound of the verb raise is [z]. The difference between [s] and [z] is
technical. When speakers make the sound [s], the vocal cords, which are located in the throat and
help to produce different sounds, do not vibrate, which means that they do not produce a buzzing
sound. There is no buzzing sound when one says [sssssss], for example. On the other hand, when
speakers make the sound [z], the vocal cords do vibrate and produce a buzzing sound, which can be
heard when one says [zzzzzzz]. Most native speakers do not consciously realize this difference, but
they instinctively know when they should vibrate their vocals cords and when they shouldn’t.
Actually, speakers can feel the difference between the sounds if they place a hand on their throat
when they say [sssssss] and [zzzzzzz].
Now notice that there is no vocal cord vibration in the sound [t], but that there is vocal cord vibration
in the sound [d]. Again, speakers can actually feel the tension in their throat when they make the
sound [d], and the tension is the same when they make the sound [z]. However, there is no similar
tension in saying either [s] or [t]. Linguists call sounds which involve vibration of the vocal cords
“voiced” sounds, and those which do not involve vibration of the vocal cords “voiceless” sounds.
During the production of words, it is natural to put sounds together that have the same features, that
is, put voiced sounds together and voiceless sounds together. There are lots of English words that
end in the sounds [st] (raced, missed, passed, etc.) and lots that end in [zd] (raised, dazed, posed,
etc.). However, there are no English words that end in either the sounds [sd] or the sounds [zt],
because such words would join voiced and voiceless sounds together.
Therefore, as a result of the nature of the vocal apparatus, a verb like race, which ends in [s]
(voiceless), should have a past tense that is pronounced with a [t] (voiceless), and it does. A verb
like raise, which ends in a [z] (voiced), should have a past tense that is pronounced with a [d]
(voiced), and it does. Using the technical terms, part of the rule for forming the past tense of regular
verbs is as follows:
(19) a. The past tense is pronounced [t], which is voiceless, in regular verbs that end in
voiceless sounds (coped [pt], hiked [kt], laughed [ft], raced [st], etc.).
b. The past tense is pronounced [d], which is voiced, in regular verbs that end in voiced
sounds (rubbed [bd], hugged [gd], loved [vd], raised [zd], razed [zd], etc.).
Consider now the third variant in (16), namely, the past tense [d] in verbs like hunted, rated,
handed, raided, etc. Notice that there are no words in English that end in the sounds [tt] or the
sounds [dd]. Be careful not to think of spelling in such cases. A verb like putt in The golfer does not
putt well ends in the sound [t], not the sounds [tt]. There is a good reason for this: English does not
allow double consonants at the end of any word or syllable. So, the variants in (19) can’t apply when
a verb already ends in [t] or [d]. In such cases, English uses the third form [d], making the past
tense a separate pronounceable syllable. The full set of rules is as follows:
15

16
(20) a. The past tense is pronounced [t], which is voiceless, in regular verbs that end in
voiceless sounds (coped [pt], hiked [kt], laughed [ft], raced [st], etc.).
b. The past tense is pronounced [d], which is voiced, in regular verbs that end in voiced
sounds (rubbed [bd], hugged [gd], loved [vd], raised [zd], razed [zd], etc.).
c. The past tense is pronounced [d] in regular verbs that end in a [t] or a [d] (hunted
[td], rated [td], handed [dd], raided [dd], etc.)
Notice that (20) is much more cost-effective theoretically than (18). Whereas (18) is merely a list
of items without any principle, (20) explains clearly why each item is in the list that it is in. It is that
kind of generalized solution that linguists are looking for. It is also the kind of generalized solution
that employers are looking for from their employees.
With generalizations such as those in (20), we can now begin to answer some fundamental questions
regarding the acquisition of such principles by children, namely, how they mange to acquire them
so quickly and why they make the kinds of mistakes they do. It is clear that the rules for forming the
correct past tense of regular verbs are known unconsciously to all native speakers of English,
including toddlers. Because the principles are based on the nature of the human vocal apparatus,
they are natural and can be acquired relatively quickly. Furthermore, the past tense of irregular verbs
– forms like went for the past tense of go – are unpredictable and, in a sense, unnatural. It takes
children a long time to master such forms. That is why three-year-old children say things like He
hurted me, which they have never heard; they haven’t yet realized that hurt is an irregular verb. Note
that, if it were a regular verb, its past tense would have to be hurted, following (20c) and parallel to
other regular past tenses like hunted, hoisted, handed, hoarded, etc.
Given the above discussion, a child exposed to English must only figure out that the English past
tense is signaled by adding a [t] or [d] type sound to the end of the verb. The distribution of the three
variants ([t], [d], and [d]) is not something that the child must worry about. That distribution
follows largely from the nature of the human vocal apparatus, which makes the distribution of the
variants a matter of necessity, rather than free choice. In short, there is no guesswork involved in
forming the past tense of regular verbs. As a result, we can account for the fact that children acquire
knowledge of the past tense rapidly and without instruction. We also can explain why children
invent forms such as hurted, which they have never heard. The principles prevent hurtt and hurtd.
Most important, children do not need to hear the past tense of every regular verb in order to know
which of the three possible endings is appropriate; from hearing a small number of forms, they can
derive all the rest. These are some of the reasons why children acquire their native language so
rapidly, and why language development can proceed normally even in cases of neglect, abuse,
disease, poverty and other unfavorable circumstances. What makes language acquisition possible
in children is this: the rules – the grammar – of all languages are not haphazard, unpredictable and
random. Rather, they are based on the principles, capacities, and limitations of the human language
apparatus, which all normal children share. As a result, in almost all cases, children can acquire any
language as a native language without even realizing that they are doing it.

Native speakers of English understand the grammar of English completely; yet, their knowledge of
the grammar is generally unconscious. For example, it is safe to say that no one reading this
discussion was consciously aware of the phonetic variations in the past tense ([t], [d], and [d])
unless he or she happened to have taken a course in English grammar taught by someone with
expertise in linguistics. Since the knowledge is unconscious, it follows that parents do not teach
their children the rules of grammar. Quite clearly, parents cannot teach their children a set of rules
if they are not consciously aware of what the rules are. So then, how do children acquire their native
language if no one teaches it to them? The answer is quite simple. The rules of language are based
on principles that are directly related to the nature of the human language apparatus which matures
as children acquire their native language. Children acquire the rules of their native language as a
natural part of their development because the rules, like those in (20), have a natural basis.
Acquiring a native language is like acquiring the skills and coordination for walking. Even without
specific help or prodding from parents, all normal children will eventually walk if they get the proper
nourishment necessary for the physical requirements of the task such as strength and balance.
Similarly, all normal children will eventually acquire a native language if they get the proper
linguistic nourishment, which means only that they need to be exposed to some language or
languages. That is why absence of the ability to acquire a native language as a child is so rare.
Again, as before, notice what is going on here at an abstract level. We have some data, which
happen to be about past tenses in English. We have a hypothesis which describes the data, namely,
that past tenses are formed by adding ed to the end of the verb. We observe some further data which
the hypothesis can’t handle, namely, that children know how to form past tenses before they know
how to spell. Thus, we must revise the hypothesis and look for another proposal. The revision led
to the list of alternatives in (18). We determined that lists like (18) are unsatisfactory; they are not
cost-effective because they reveal no generalizations. What is needed is a set of principles that can
explain the behavior of native speakers, in particular, the fact that they can produce and understand
an unlimited number of new past tenses and can predict what the correct form will be. Accordingly,
we revised (18) to (20) relating the data to human biology and giving a real explanation for why the
examples fall out the way they do. In other words, what we have done with this problem in
phonology is exactly what we did with the problem in syntax, and, at an abstract level, exactly what
Ford engineers did when they modified the design of the Villager.
Linguistics
It is hard to imagine spending one waking moment without language. Whether we are alone or
among other people, whether we dream or daydream, whether we write poetry, follow a recipe, cheer
for the home team, speak or sing, language is involved. All normal children acquire a native
language, no matter where they are born, what the language is or what their home life is like.
Absence of the ability to acquire language is very rare. People who are deaf have language; so do
those who are blind, mute, completely paralyzed, intellectually disabled, or emotionally disturbed.
Language can be disrupted by injury or disease, processed by machines, altered for special occasions,
and exploited for ulterior motives. Despite this extraordinary presence, versatility and variability,
17

18
every human language, whether Old English or Modern Japanese, shares universal features which
are directly related to human biology.
The goal of courses in linguistics is to formulate a theory of language that explains what language
is and how it is acquired and used. In particular, a theory of language must explain how all normal
children can master any of the world’s languages as a native language and can do so without any
specific instruction from parents or care-givers, whereas learning a language as an adult requires
intense study and training. In fact, this problem is huge since it is ultimately concerned with the
biological characteristics of human beings, specifically, with the characteristics of the human
language apparatus which includes the brain and the organs of speech and hearing. Linguistics
courses break down that huge problem into manageable subparts to find satisfactory explanations.
Since many linguistic courses deal with English, which students already know, linguistics is an
excellent discipline for practicing the skills needed for success in the workplace. Students learn how
to collect, organize, and analyze data in English and other languages, make hypotheses to account
for the data, alter hypotheses to accommodate newly discovered data, and evaluate competing
proposals to choose the most viable and robust analysis.
Because language is so pervasive and so peculiarly human, students of linguistics find careers in
many different areas. Some, such as teachers, computer scientists, and speech therapists, use
linguistics directly; others such as market analysts, editors, and advertising executives, use it
indirectly. Still others use their undergraduate major in linguistics as a springboard to careers in law,
education, business, artificial intelligence, medicine, and international relations, as well as graduate
study in linguistics and other fields. A list of career opportunities posted by the Linguistic Society
of America is included as an Addendum to this Introduction. That list and other useful information
can be found at lsadc.org. More job opportunities are available at linguistlist.org/jobs.
Quite naturally, linguistics courses are of direct benefit to language teachers, especially teachers of
ESL (English as a Second Language). A typical ESL class has students who speak different native
languages with varying levels of proficiency in both English and their native language. It is
impossible for ESL teachers to know ahead of time what kinds of questions about English grammar
the students in such mixed classes will ask. The best preparation for ESL teachers is a solid
understanding of English linguistics, that is, a linguistic description of English grammar such as the
description given in this book. This does not mean that ESL teachers should memorize the various
analyses given here. Quite the opposite is the case: a formal linguistic description is not going to be
very helpful to students learning English. What is important is that ESL teachers develop skill in
linguistic analysis, that they learn how to look at languages properly, not as chaotic and whimsical
collections of rules, but as systems rooted in human biology. In this way, ESL teachers can
understand the facts and issues behind their students’ questions and know that there is a real
explanation for those facts. As a result, they will be able to teach English and to answer students’
questions in a way that is truly meaningful.
Many linguistics courses satisfy the general education requirement in formal reasoning because they
emphasize objective methods of analysis, synthesis and evaluation. They are specifically designed

to introduce students to systematic ways of organizing, processing, and analyzing data. The focus
on all LIN courses is on formulating hypotheses explicitly and testing them rigorously. In these
courses, students are asked to solve individual problems in English and a variety of languages as a
means to developing strategies and expertise for solving any kind of problem. Overall, students
learn how to develop explanatory models that can predict the phenomena they observe, that is,
students learn how to propose solutions to problems that are rigorous enough to incorporate the ever
expanding corpus of data. As this entire discussion has indicated, the skills that students acquire in
linguistics are exactly the skills needed in today’s workplace.
Since courses in linguistics typically involve investigation of data in a variety of different languages,
they also help to give students a global, multi-cultural perspective. In today’s world, information
can be almost instantly shared around the globe, as phenomena like the Arab spring testified so
stunningly. Companies and businesses have global interests and must adapt their products and
messages to widely divergent populations. They cannot afford to be egocentric or sociocentric.
Now, more than ever before, it is important for students to understand different cultures and different
ways of looking at the world. Linguistics by its very nature provides students with that global
perspective.
We began this discussion by noting that there are many myths and traditions surrounding options like
the choice of a major in college, for example, that students often think it will be impossible to get
a job in a specific discipline unless they study that discipline in depth in college. Another myth is
that a college education ought to be quantifiable, that students ought to be able to hold up a bunch
of books upon graduation and say, “I am educated because I know everything in these books.” But
expertise in any area is not tangible; it is not something a person can weigh. Becoming a good golfer
or a prima ballerina, for example, involves practice, false starts, and bad days. A good golfer may
be under par one day and over it the next. A ballerina may perform flawlessly on one evening, and
falter on another. Of course, what makes a golfer and a ballerina very good is that they are generally
just that: very good. Over time, we can measure the success of a golfer and a ballerina with such
quantifiable records as scores and box office receipts. But those measures are the effects of the
expertise, not the expertise itself. There are thousands of excellent golfers who have never won a
tournament and thousands of proficient ballerinas who have never made it to the big stage.
The ability to solve problems, to organize and analyze data, to come up with proposals that are
innovative, creative, and robust takes insight, and insight is intangible. It cannot be weighed. The
confidence and security that lead one to perform well on the job are a consequence of practice and
training. Students should not look at their education as something that will give them a store of
enumerable facts or an arsenal of memorized information. Education is about developing intuition,
sensitivity, vision, and perspective, and those attributes are qualities that cannot be directly
measured. Moreover, it is those kinds of attributes that will help students succeed in today’s variable
and unpredictable job market.
19

20
ADDENDUM: CAREER OPPORTUNITIES IN LINGUISTICS
Monica Macaulay and Kristen Syrett (Linguistic Society of America)
Work in the computer industry: Training in linguistics can equip you to work on speech recognition, text-to-speech
synthesis, artificial intelligence, natural language processing, and computer-mediated language learning.
Work in education: People with a background in linguistics and education can develop materials for different
populations, train teachers, design assessments, find effective ways to teach language- related topics in specific
communities, or use the language of a community effectively in instruction. Many applied linguists are involved in
teacher education and educational research.
Teach English as a Second Language (ESL) in the United States or abroad: If you want to teach ESL in the US, you
will probably need additional training in language pedagogy, such as credentials in Teaching English as a Second or
Other Language (TESOL). Many teaching positions abroad require only an undergraduate degree, but at least some
specialized training in the subject will make you a much more effective teacher. Linguistics can give you a valuable
cross-language perspective.
Teach at the university level: If you go on to get a graduate degree in linguistics you might teach in departments such
as Linguistics, Philosophy, Psychology, Speech/Communication Sciences, Anthropology, English, and departments
focused on specific foreign languages.
Work as a translator or interpreter: Skilled translators and interpreters are needed everywhere, from government to
hospitals to courts of law. For this line of work, a high level of proficiency in the relevant language(s) is necessary, and
additional specialized training may be required.
Teach a foreign language: Your students will benefit from your knowledge of language structure and your ability to
make certain aspects of the language especially clear. You will need to be very proficient in the relevant language, and
you may need additional training in language pedagogy.
Work on language documentation or conduct fieldwork: Some agencies and institutes seek linguists to work with
language consultants in order to document, analyze, and preserve languages (many of which are endangered). Some
organizations engage in language-related fieldwork, conducting language surveys, establishing literacy programs, and
translating documents of cultural heritage.
Work in the publishing industry, as a technical writer, or as a journalist: The verbal skills that linguists develop
are ideal for positions in editing, publishing, and writing.
Work for a testing agency: Linguists help prepare and evaluate standardized exams and conduct research on assessment
issues.
Work with dictionaries (lexicography): The development of good dictionaries requires the help of qualified linguistic
consultants. Knowledge of phonology, morphology, historical linguistics, dialectology, and sociolinguistics is key to
becoming a lexicographer.
Become a consultant on language in professions such as law or medicine: The subfield of forensic linguistics
involves studying the language of legal texts, linguistic aspects of evidence, issues of voice identification, and so on. Law
enforcement agencies such as the FBI and police departments, law firms, and the courts hire linguists for these purposes.

Work for an advertising company: Companies that specialize in advertising often do extensive linguistic research on
the associations that people make with particular sounds and classes of sounds and the kind of wording that would appeal
to potential consumers.
Work for the government: The federal government hires linguists for the Foreign Service, the Federal Bureau of
Intelligence (FBI), the National Security Agency (NSA), the Central Intelligence Agency (CIA), the Department of
Defense, the Department of Education, and so on. Similar opportunities may exist at the state level.
Become an actor or train actors: Actors need training in pronunciation, intonation, and different elements of grammar
in order to sound like real speakers of a language or dialect. They may even need to know how to make mistakes to sound
like an authentic nonnative speaker.
http://www.lsadc.org/info/ling-faqs-whymajor.cfm
21

CHAPTER ONE: FUNDAMENTALS OF LINGUISTICS
1.1 GRAMMATICAL CHARACTERIZATION AND GRAMMATICAL REALIZATION
The first language that children acquire during childhood is called their NATIVE LANGUAGE.*
As they mature, children are considered NATIVE SPEAKERS of that language. Knowledge of
one’s native language is quite special. It differs in many ways from the knowledge one gains
studying a language in school, for example, in Spanish class. In particular, native speakers are not
generally aware of what they know about their native language. Also, they usually cannot describe
what they know or explain how they know it. Their knowledge is unconscious. To see this, consider
the following sentences, which differ only in that (1b) contains the word that:
(1) a. Do you think the judge will tell the truth?
b. Do you think that the judge will tell the truth?
Every native speaker of English knows that these two sentences are yes/no questions, that is,
questions that can get either a positive or negative answer. Now suppose that a person does not
know who will tell the truth and wants to find out. In such a case, the person could ask a
WH–question, which is a question that begins with a WH–word like who, what, where, etc. Given
sentences like those in (1), every native speaker unconsciously knows that the results of attempting
to ask such a WH–question will be different. Sentence (2a), a WH–question related to (1a), is
perfectly GRAMMATICAL, meaning it does not violate any rules. On the other hand, sentence
(2b), a WH–question related to (1b) and containing the word that, is UNGRAMMATICAL,
meaning it violates some rule (an asterisk at the beginning of a sentence indicates that it is
ungrammatical).
(2) a. Who do you think will tell the truth?
b. *Who do you think that will tell the truth?
There is a specific principle that rules out sentences like (2b), yet native speakers typically are not
consciously aware of what that principle is. Nor can they explain why the word that can be left out
of (3a) to produce the synonymous (3b), whereas a similar deletion in (4a) produces the
ungrammatical (4b).
(3) a. The lawyers that the judges admire will tell the truth.
b. The lawyers the judges admire will tell the truth.
(4) a. The lawyers that admire the judges will tell the truth.
b. *The lawyers admire the judges will tell the truth.
_______
* Words and phrases in boldface capitals are important concepts. They are found in the index,
which indicates the page(s) on which the concepts are introduced or reviewed. There is also a
glossary of most of these concepts on Moodle in the files section under the folder HANDOUTS.

24
Again, there is a specific principle that accounts for the ungrammaticality of (4b) which native
speakers must know, at least unconsciously, because they all recognize that (4b) is not a well-formed
sentence. About the best that the average native speaker can say about (4b) is that it doesn’t sound
right. The same is true of (2b) and, in fact, all ungrammatical sentences. In general, native speakers
cannot explain why grammatical sentences are grammatical and why ungrammatical sentences are
ungrammatical. Yet they do know the difference between the two.
The principle that predicts the ungrammaticality of (4b) is highly specific. Changing admire to know
in (3) produces the pair of sentences in (5) which mean the same thing.
(5) a. The lawyers that the judges know will tell the truth.
(= The ones who will tell the truth are the lawyers that the judges know.)
b. The lawyers the judges know will tell the truth. (=5a)
On the other hand, changing admire to know in (4) produces the pair of sentences in (6) which mean
very different things.
(6) a. The lawyers that know the judges will tell the truth.
(= The ones who will tell the truth are the lawyers that know the judges.)
b. The lawyers know the judges will tell the truth. (=/ 6a)
Notice also that (6b) looks a lot like (4b); however, (6b) is grammatical, but (4b) is not. Further, (5a)
and (6a) can be expanded as in (7), but a similar expansion of (4a) produces (8) which is not
grammatical.
(7) a. The lawyers that the judges know will tell the truth are from Michigan.
b. The lawyers that know the judges will tell the truth are from Michigan.
(8) *The lawyers that admire the judges will tell the truth are from Michigan.
These examples are not gimmicky or atypical; thousands of pairs of grammatical/ungrammatical
sentences like them can be found in the linguistics literature. Such abundant examples emphasize
that native speakers unconsciously know a vast and detailed array of facts about their language, and
that these facts are largely hidden from introspection and analysis. It is simply incorrect to say that
the facts of language – the grammar – are well known to (even educated) speakers. This means, of
course, that there is a difference between having knowledge of a language and being able to talk
about that knowledge. Linguists refer to the unconscious knowledge that native speakers have of
their native language as their LINGUISTIC COMPETENCE.
Since native speakers do not consciously know what makes a sentence grammatical or
ungrammatical in their native language, it follows that native speakers do not teach language to their
children. While parents do correct children on occasion, it is hardly likely that all native speakers
of English have been told by their parents that sentences like (2b), (4b), and (8) are ungrammatical

and should never be uttered. If speakers required specific instruction of this type, no one would
acquire a native language because there is an infinite number of grammatical and ungrammatical
sentences in every human language. In addition, even when native speakers have expert, fluent
knowledge of grammatical principles, they can’t share that knowledge with their children because
children are unable to comprehend even the most elementary statements about grammar. Most of
one’s native language is acquired before one possesses the cognitive skills to discuss the grammatical
principles which underlie the ability. For example, kindergartners clearly know the difference
between statements and questions before their first day of school, yet none can discuss the difference
in principled terms that reveal the facts behind the difference.
In view of the above, modern linguistics is principally concerned with two broad empirical problems
which have often been referred to as the GRAMMATICAL CHARACTERIZATION PROBLEM
and the GRAMMATICAL REALIZATION PROBLEM. Grammatical characterization entails
describing the linguistic competence of native speakers, that is, discovering and generalizing the
grammatical principles that constitute their unconscious knowledge of their native language. In
grammatical characterization, a linguist describes what the principles are which determine the
grammaticality of examples like (1) through (8).
Grammatical realization, on the other hand, entails accounting for native speakers’ LINGUISTIC
PERFORMANCE, that is, their acquisition and use of their unconscious knowledge. In
grammatical realization, a linguist describes how such principles become part of the linguistic
competence of native speakers, in short, how children achieve mastery of their native language.
Linguists must address both the grammatical characterization problem and the grammatical
realization problem in formulating a theory of human language. Clearly, there are huge obstacles
to doing so. Fundamentally, the linguistic competence of humans is a property of the human mind,
and studying the nature of the human mind is as complex a task as studying the nature of the
universe. The grammatical information that native speakers have unconsciously internalized about
their native language is vast, detailed, and often extraordinarily subtle.
One cannot emphasize enough the difference between the unconscious knowledge that native
speakers have of their native language and the knowledge they learn about their language in school.
For example, in school, children learn that English contains many words called homonyms which
are spelled differently but pronounced the same, such as two, to, and too, all of which are
pronounced [tu]. While this is something children consciously learn in school, it is clear that they
already know the difference between these words even when they don’t know how to spell them.
Consider the following sentences, remembering that children do not know how to read when they
first encounter such sentences in speech:
(9) a. I have to clean shirts.
b. I have two clean shirts.
25

26
When people read the sentences in (9), they can obviously see the difference between to and two. But
even without reading those sentences or having any knowledge of English spelling or homonyms,
it is clear that native speakers, including children, know the difference between to and two. Thus,
they know that they can say (10a) instead of (9a) contracting have and to into hafta, but they cannot
say (10b) instead of (9b).
(10) a. I hafta clean shirts. (where hafta means ‘have to’)
b. *I hafta clean shirts. (where hafta means ‘have two’)
Similarly, in elementary school, children learn to recognize and talk about the distinction between
nouns and verbs. They learn that school is a noun and eat is a verb in sentences like the following:
(11) a. I am going to school. (school is a noun)
b. I am going to eat. (eat is a verb)
Unconsciously, they already know that school is a noun and eat is a verb before the teacher ever
brings the matter up, because they will give an answer like (12b), but never one like (13b).
(12) a. Question: Where are you going to?
b. Answer: School.
(13) a. Question: Where are you going to?
b. Answer: *Eat.
Further, they know that the two uses of the word to in (11a) and (11b) are not the same. As a result,
they know that they cannot contract going to into gonna in (11a) but they can make the contraction
in (11b) as follows:
(14) a. *I’m gonna school.
b. I’m gonna eat.
We know children know the difference because they frequently utter sentences like (14b) but never
utter sentences like (14a). Lastly, children also know that (15a) is ambiguous, while (15b) is not.
(15) a. I am going to work. (ambiguous: work is either a noun or a verb)
b. I’m gonna work. (unambiguous: work is only a verb)
In addition to knowing about two and to, children also know about too. They know, for example,
that (16a) means ‘They came in order to clean the room,’ whereas (16b) means ‘They did so clean
the room.’ That is why the interchange in (17) is grammatical, whereas the one in (18) is not.
(16) a. They came to clean the room. (= They came in order to clean the room.)
b. They did too clean the room. (= They did so clean the room)

(17) a. Question: Why did they come?
b. Answer: To clean the room.
(18) a. Question: Why did they come?
b. Answer: *Too clean the room./*So clean the room.
Children also know that too has a variety of uses like to, so that (19) is ambiguous meaning either
‘They are overly clean’ or ‘They are definitely/emphatically clean.’
(19) They are too clean.
Lastly, children unconsciously know that clean in (20a) is an adjective and clean in (20b) is a verb.
(20) a. They were clean.
b. They did clean.
We know they know this because they will utter only the grammatical sentences in examples like
the following:
(21) a. They were very clean.
b. *They did very clean.
c. They were cleaner.
d. *They did cleaner.
e. *They were clean it.
f. They did clean it.
g. *They were so clean better than us.
h. They did so clean better than us.
In short, when children acquire their native language, they acquire unconscious knowledge of the
parts of speech (nouns and verbs) and of word usage (two, to, and too) long before they go to school
and learn to talk consciously about what they already unconsciously know. They unconsciously
know what a noun is and what a verb is; they just don’t consciously know that a noun is called a
“noun” and a verb is called a “verb.” They will correctly use all three words that are pronounced [tu]
long before they learn to spell them. In fact, although they know the difference between the three
words (two, to, and too) and do not misuse them in speech, many children often misspell them into
adulthood writing sentences like *He is to tall. We can tell what children unconsciously know by
listening to what they say and observing whether or not they understand what is being said to them.
It is the task of linguistics to discover the exact nature of that unconscious knowledge and to
formulate a theory that explains how children acquire that knowledge. As this discussion indicates,
both tasks are formidable.
27

28
1.2 THE BRANCHES OF GRAMMAR AND LINGUISTICS
The crucial question that must be asked in attempting to solve the grammatical characterization
problem is, What is the nature of the unconscious knowledge that all native speakers have? The
question must be answered in each of the areas of grammar that form part of a language: phonetics,
phonology, morphology, syntax, and semantics.
PHONETICS and PHONOLOGY are the two branches of grammar which are concerned with the
study of the sounds of human language. Phonetics deals primarily with the speech sound itself,
including the way in which it is produced, transmitted, and perceived; phonology deals more with
the organization of speech sounds into sound systems.
Since the sounds of language are spelled in a variety of ways both in English and in the other
languages of the world, we must have a unique way of representing each sound regardless of the
spelling. For that purpose, linguists have devised a PHONETIC ALPHABET, which is an
inventory of unambiguous symbols which uniquely represents each individual sound. For example,
the symbol [s] represents the underlined sound in each of the following words: sin, cross, science,
ceiling, and psychology; the symbol [] represents the sound of “t” in writer and the sound of “d” in
rider; the symbol [] represents the underlined vowel sound in hunted, roses, alone, and sofa. A list
of the symbols we will use in this book is given on pages 128 and 129.
One example of the phonological competence native speakers of English have concerns the
pronunciation of [t] and [d] when those sounds occur between two vowels:
(22) a. [t]: wait, write, vote, suit
b. [d]: wade, ride, load, feed
c. []: waiter, writer, voter, suitor; wader, rider, loader, feeder
It is safe to say that native speakers are generally unaware of the way they alter the sound [t] in write
and [d] in ride when they add the [r] suffix. On the other hand, it is clear that they do, in fact,
unconsciously know when and how to make the change. If one makes up words like the following
and asks native speakers how to pronounce them, they will respond in the way indicated:
(23) a. [d]: plouds [plawdz]
b. []: plouder [plawr]
c. [t]: plouts [plawts]
MORPHOLOGY is the branch of grammar which is concerned with the internal structure of words.
The minimal unit of meaning in a word is the MORPHEME, and every word in every language
consists of one or more morphemes. For example, the word write has one morpheme ([rayt])
meaning ‘form letters and words,’ while writer has two morphemes: [rayt] and [r], which signifies
an agent or person performing the action in the verb. The word writers [rayrz] adds a third
morpheme, the “s” signifying plural. Notice that the number of morphemes in a word does not

necessarily equal the number of syllables. The word writers [rayrz] has three morphemes, but the
word spiders [spayrz] has only two. Notice also that native speakers are unconsciously aware of
any modifications that must be made in the pronunciation of morphemes as they appear in different
words. As we observed in the Introduction, the past tense morpheme is pronounced in three different
words depending on the final sound of the verb:
(24) a. [t]: after sounds that are [–VOICED], e.g., kicked, stopped, laughed, etc.
b. [d]: after sounds that are [+VOICED], e.g., plunged, crammed, loved, etc.
c. [d]: after [t] and [d], e.g., hunted, raided, relented, etc.
Again, if one makes up words like the following and asks native speakers how to pronounce their
past tense form, they will respond in the way given, which clearly indicates that they know the rules,
albeit unconsciously:
(25) a. goob the past tense is pronounced [gubd] with the suffix [d]
b. nork the past tense is pronounced [nrkt] with the suffix [t]
c. frist the past tense is pronounced [frstd] with the suffix [d]
SEMANTICS is the branch of linguistics concerned with the meanings of words and sentences.
Native speakers clearly understand the meaning of the major categories of words like nouns and
verbs, but they also understand the meaning of other parts of speech like prepositions (for a brief
review of the parts of speech and other grammatical terms, see the grammar review beginning on
page 309). A preposition is a word that introduces a phrase ending in a noun and typically serves
as a modifier. Examples of prepositions include in, to, into, on, about, within, etc. As an illustration
of native speakers underlying competence about meaning, consider what they know about the
meanings of the prepositions in the following:
(26) a. in: John ran in the street.
b. to: John ran to the street.
c. into: John ran into the street.
d. on: John ran on the street.
e. from: John ran from the street.
f. out of: John ran out of the street.
Clearly, the differences in the meanings of these sentences has to do with the meanings of the
prepositions since everything else in the sentences is the same. Native speakers are unconsciously
aware of those meanings because they will find sentences like the following ungrammatical:
(27) a. *The thief inserted the key out of the lock.
b. *The dentist extracted the tooth into the patient’s mouth.
c. *The teacher stuffed the test papers to her briefcase.
Furthermore, native speakers can supply appropriate endings to sentences like the following:
29

30
(28) a. Mary fell... *under the forest, down the stairs, off the roof, etc.
b. Sue went... into the forest, down the stairs, off the roof, etc.
c. Bill emerged... from the forest, *down the stairs, *off the roof, etc.
SYNTAX is the branch of linguistics that studies the way words are combined to form phrases and
sentences. As we saw above, native speakers’ intuition about the formation of phrases and sentences
is extraordinarily complex and subtle. As another example, consider the various positions in syntax
in which different adverbs can occur in a sentence like The student will read the book. In the
following examples, a “?” in the front of the sentence means that it is questionably grammatical; a
“*” means it is ungrammatical; and a “**” means it is really bad; the comma indicates a pause.
(29) a. Sentence Adverbs: [ultimately]
(1) Ultimately, the student will read the book.
(2) ** The ultimately student will read the book.
(3) The student ultimately will read the book.
(4) The student will ultimately read the book.
(5) * The student will read ultimately the book.
(6) ** The student will read the ultimately book.
(7) The student will read the book, ultimately.
(8) ? The student will read the book ultimately.
b. Manner Adverbs: [thoroughly]
(1) * Thoroughly the student will read the book.
(2) ** The thoroughly student will read the book.
(3) * The student thoroughly will read the book.
(4) The student will thoroughly read the book.
(5) * The student will read thoroughly the book.
(6) ** The student will read the thoroughly book.
(7) The student will read the book thoroughly.
c. Miscellaneous Adverbs: [hardly]
(1) * Hardly the student will read the book.
(2) ** The hardly student will read the book.
(3) ? The student hardly will read the book.
(4) The student will hardly read the book.
(5) * The student will read hardly the book.
(6) ** The student will read the hardly book.
(7) * The student will read the book hardly.

Different speakers may not agree exactly with all of the above judgements (?, *, or **), but it is clear
that the knowledge that they have of their native language is substantially uniform. The task for
linguistics is to account for that unconscious knowledge of syntax.
Turning to the grammatical realization problem, the crucial questions that must be asked are, How
do children acquire their native language, and how do native speakers produce and understand their
native language? The branches of linguistics that are primarily concerned with such questions are
PSYCHOLINGUISTICS, the psychology of language, BIOLINGUISTICS, the biology of
language, and NEUROLINGUISTICS, the neurology of language. Investigators in all these areas
have noted the dramatic difference between native language acquisition, the language one
unconsciously acquires at home as a child, and second language acquisition, the language one
consciously studies in school usually as an adult. Here is a summary:
(30) NATIVE LANGUAGE SECOND LANGUAGE
Speech is the sole input Speech and writing are inputs
Data heard is fragmentary Data studied is detailed and specific
Data heard is haphazard Data studied is organized
Data heard is often defective Data studied is fully grammatical
Data heard is mostly novel Data studied consists of patterns and drills
Acquisition is unconscious Learning is conscious
Acquisition is involuntary Learning is voluntary and deliberate
Virtually no instruction Formal instruction by experts
In addition to the above contrasts, native language acquisition has all of the following characteristics
which the grammatical realization problem must take into account and explain.
First, native language acquisition is rapid. It is completed in a remarkably short period for all normal
children. Generally, children utter their first words around their first birthday. By the time they enter
preschool and kindergarten at age five or six, they have mastered the basic structure of their native
language.
Second, children’s comprehension of language proceeds ahead of their production of language.
Children understand much more than they can express. A typical 12-14 month-old child might
produce about ten words like the following: mommy, daddy, cookie, pops, balloon, milk, truck,
some, up, down, sit, etc. However, the same child is likely to understand about ten times as many
words and phrases like helicopter, plane, Volkswagon, Burt, Bambi, Donald Duck, cheese, Coca-
Cola, hot dog, teeth, toes, belly button, up in high chair, go play piano, give daddy some, don’t spill
it, don’t throw it on the floor, let’s go get changed, brush your hair, etc.
Third, native language acquisition is not dependent on peripheral sensory mechanisms. Language
develops even in cases of blindness, deafness, and mutism. In fact, as the famous example of Helen
Keller attests, children can be both bind and deaf and acquire a native language provided the data
31

32
of the language gets into the child’s brain. In Helen Keller’s case, her teacher taught her the meaning
of words using finger spelling.
Fourth, native language acquisition is not dependent on intelligence. Language develops except in
cases of severe intellectual disability. The only children who fail to acquire a native language are
those who only reach a mental age of about two, that is, children who can barely be toilet trained.
Fifth, native language acquisition is creative. Children do not merely imitate words and phrases they
have heard in their environment. Children say things for which there is no adult model, that is, they
say things like the following which they most likely have never heard anyone say:
(31) a. Allgone itchy. (after being scratched)
b. I do it Kevin’s-self. (= ‘I do it myself’)
c. Do I have a jack sized bed? (since parents have a king sized bed)
d. open (= ‘sharpen’ or ‘inflate’)
e. dirty (for the part of an apple that is not peeled)
f. big water (= ‘lake’)
Children also overgeneralize. They say throwed for threw, mans for men, and so on. It is highly
doubtful that children have heard some say throwed and mans. Interestingly, if throw were a regular
verb, its past tense would have to be throwed; if man were a regular noun, its plural would have to
be mans. Children unconsciously know that. What they do not know when they say throwed and
mans is that those words are irregular.
When children do imitate, they often do so without really understanding what they are saying. For
example, upon telling a child to go to bed, the child might respond, “OK, I do it 12 times.” One
three-year-old child who was frustrated with his toys was heard to say, “I’m having a nervous
breakdown.” Clearly, that child did not understand the meaning of the sentence.
Sixth, native language acquisition is predictable. All children make the same kinds of mistakes
saying things like the following if the language is English:
(32) a. He goed there. (They have not yet learned that go is irregular.)
b. Him goed there. (They have not yet mastered the different uses of he and him.)
c. I goed the truck. (They have not yet learned that go is intransitive.)
Children also use the same strategy. Generally, in second language learning, students are taught and
study a simpler area of grammar before going on to more complex areas. That is not the case in
native language acquisition. The entire grammar is encountered at once. When children are infants,
parents and care-givers do tend to modify their speech using baby-talk. By the time children are two
years old, they are exposed to the entire language generally without restriction. Children will make
attempts at saying all sorts of things. They never worry about being incorrect or sounding stupid.
Sometimes, their pronunciation is so idiosyncratic that only their parents can understand them.

Seventh, the stages of acquisition are uniform. Typical primitive sentences are sentences like the
following:
(33) a. truck [tk] Noun
b. pretty baby [pii bebi] Modifier + Noun
c. daddy coat [dædi ko] Noun + Noun
d. see truck [si tk] Verb + Noun
e. see baby shoe [si bebi su] Verb + Noun + Noun
Psycholinguists have known for many years that the words children know and use at a very early age
tend to fall into two large groups. One group, which have been called PIVOT WORDS, consists
of words that are low in number but high in frequency. This group includes quantifiers (more,
some), determiners (that, a), possessives (my), and adjectives (pretty, big). The other group, called
OPEN WORDS, consists of words that are high in number but low in frequency. This includes
nouns (baby, truck, mommy) and verbs (go, sit, play). In the early stages, pivot words usually don’t
occur together and don’t occur alone. While they form one class initially, over time, children will
differentiate the classes. The differentiation of pivot words can be charted as follows:
(34) a. Time One - around 24-27 months
Pivot One Open
that horsie
a truck
my car
pretty baby
more milk
Pivot One includes quantifiers (more, some), determiners (that, a), possessives (my),
and adjectives (pretty, big).
b. Time Two - about three months later
(Dem) + (Det) + Pivot Two + Open
that a my truck
that the big truck
*that my a truck
*my a that truck
Pivot Two includes possessives, quantifiers, and adjectives. Demonstratives have
been taken out of Pivot One as a separate class.
33

34
c. Time Three - about three months later
(Det) + (Adj) + Open
a big cup
*a my cup
*a more cup
Pivot Three consists only of possessives and quantifiers. Adjectives have been taken
out of Pivot Two and now form a separate class.
The differentiation of pivot words over time is typical of native language acquisition. Children begin
with very large overgeneralized classes of items and gradually differentiate those items into
subclasses as their cognitive and perceptual abilities mature. For example, at a very early age,
children might refer to all animals as doggie. Over time, they will differentiate animals into finer
and finer subcategories. The impetus for this development is children’s maturation, which is both
programmed genetically and influenced environmentally, that is, children will mature in biologically
determined ways in response to environmental stimuli. They are clearly not just imitating and
memorizing. Adults do not refer to a giraffe, for example, as a “big doggie.” There are thousands
of possible combinations of pivot and open words and children generally produce combinations
which they have never heard.
The crucial facts that an adequate theory of language must account for are these: native language
acquisition proceeds in children unconsciously, in predictable ways, and with almost the complete
lack of instruction. In addition, an adequate theory must account for the fact that language does not
require paper and pencil analysis either by adults or children even during the early stages. In fact,
when children are acquiring their native language, they do not know how to read and write. Further,
many adults are illiterate but still have a native language which they produce and understand
effortlessly. A native language is not “figured out” the way one might translate a second, foreign
language. It is immediately generated and understood.
Speaking and understanding are incredibly swift. Consider, for example, what is involved in your
understanding of the sentence that you are now reading, as well as how quickly you are
understanding what this sentence means even though it is very long and complicated and you have
never seen it before in your entire life. You are utilizing all the aspects of grammar of English at
once: you are simultaneously understanding the phonology, the morphology, the syntax, and the
semantics. Further, you have had only a one-time exposure to that other sentence. You are now
reading another totally new sentence which you also have never seen before. The major questions
are, What is there in your brain that makes all that possible and how did it get there?

1.3 THEORIES OF NATIVE LANGUAGE ACQUISITION
There are several popular myths about how children acquire their native language. The first and
perhaps most common is that children acquire their native language because parents and care-givers
teach it to them. At this point, it should be abundantly clear that such is not the case. We have
considered a wide variety of facts about English grammar, and it is safe to say that the average native
speaker is not consciously aware of any of them. Asking English speakers how to form the past
tense of English verbs or how to form a question in English will elicit responses which are largely
inaccurate, if not silly. Clearly, if people do not consciously understand the rules necessary to
produce grammatical English sentences, they cannot teach those rules to their children. Even when
parents do know what the rules are, it is useless to try to teach such rules to children. Children are
acquiring the rules of their native language at a time when they lack the cognitive capacity to discuss
those rules. Children do not compare notes about grammatical rules with their playmates.
Another myth about native language acquisition is that children acquire their language because they
need to or because they derive pleasure from it. It is impossible to substantiate such an idea. The
needs and pleasures of one child are often substantially different from those of another child, yet
native language acquisition is remarkably uniform across children. The fact is that native language
acquisition is involuntary and obligatory. One can choose not to learn to read and write, though one
will have difficulty functioning in society with that choice. But there is no child who wakes up one
day and consciously decides not to acquire the language to which he or she has been accidentally
exposed. We do not find some children who really enjoy acquiring English and, therefore, become
English junkies, accelerate their maturation, and acquire all of English in a few months.
Generally speaking, there have been two broad philosophical and psychological theories about native
language acquisition. One theory is called EMPIRICISM (BEHAVIORISM, NURTURISM).
It was first proposed by the philosophers John Locke (1632-1704), George Berkeley (1685-1753),
and David Hume (1711-1776). The empiricist approach claims that human beings are born with
very little in the way of instinctive behavior, that is, they are born with very little in the way of innate
mechanisms for acquiring knowledge about language or anything else. Most of what they come to
know is a product of their experiences with the physical and social world. Human beings learn by
imitating, associating, repeating, generalizing, and so on.
The other theory is called RATIONALISM (MENTALISM, NATIVISM). It was first proposed
by the philosophers Rene Descartes (1596-1650), Baruch Spinoza (1632-1677), and Gottfried
Wilhelm von Leibniz (1646-1716). The rationalist approach claims that human beings are born with
a rich and detailed system of innate mechanisms which determine the way in which they acquire
knowledge and interpret the data of everyday experience.
Nowadays, most linguists take a rationalist approach to language acquisition acknowledging, at the
same time, that the environment plays a crucial role in activating, validating, and automatizing the
process. It is well-known that children who grow up in loving, nurturing environments fare better
35

36
in all areas of development, including native language acquisition. Conversely, children who are
abused, neglected or traumatized lag behind their more fortunate peers.
1.3.1 NATIVE LANGUAGE ACQUISITION AND MATURATION
There is biological evidence that supports a rationalist view and links native language acquisition
to maturation. First, there is a relationship between sex and acquisition. Females tend to go through
the stages of language acquisition earlier than males reflecting the fact than females mature earlier
and go through puberty earlier than males.
Second, there is an interlocking of speech and motor milestones during development. Around the
age of one year, children will usually say their first words and take their first steps. By eighteen
months, they are uttering two-word sentences and walking. Most pediatricians’ offices have charts
on the wall which outline these milestones. A search on the internet will quickly return similar
charts. What is interesting in this regard is what happens in cases of general retardation such as
those caused by conditions like hypothyroidism. In such instances, there is a general stretching of
a child’s development such that all aspects of development are delayed; however, the synchrony of
speech and motor milestones remains intact. What that means is that milestones like the first word
and first step still emerge together, but do so at a later time in development. That outcome gives
clear support to the hypothesis that language acquisition is governed by maturation.
Third, there is a critical period for native language acquisition during which major neurological
changes occur. Human beings like all vertebrates are bilaterally symmetrical, which means that they
have the same structures on opposite sides of the body (arms, legs, eyes, ears, kidneys, etc.) except
for structures on the midline (spine, nose, tongue, etc.). The human brain is divided into two
hemispheres – the left and the right. Sensation and movement in the body are contralaterally
controlled by the brain, that is, the left side of the brain controls the right side of the body, and the
right side of the brain controls the left side of the body. Language is conspicuously different. The
human brain is asymmetrically specialized for language. In most adult humans, language is under
the dominant control of just one hemisphere, usually the left hemisphere. This is particularly the
case in strongly right-handed people.
1.3.2 CEREBRAL LATERALIZATION
The asymmetric specialization of the brain is called CEREBRAL LATERALIZATION and it is
linked to a critical period for native language acquisition. Specifically, if children are not exposed
to a language before cerebral lateralization is fixed, they will be unable to acquire a native language.
The exact age at which cerebral lateralization is complete is a subject of scholarly debate. A
conservative view would be that it is completed by puberty. There is ample evidence both that
cerebral lateralization occurs and that it is linked to a critical period for native language acquisition.

Evidence for Cerebral Lateralization
First, stimulation of the speech areas of both the left and the right hemispheres of the brain during
neurosurgery produce spontaneous vocalization in young children; however, in most adults, only
stimulation of the speech areas of the left hemisphere produces spontaneous vocalization.
Second, anesthetization of both the left and the right hemispheres of the brain produces a cessation
of linguistic function in children; however, in most adults, only anesthetization of the left hemisphere
produces a cessation of linguistic function.
Third, removal of either the left or the right hemisphere in young children to excise a large invasive
tumor has the same effect on native language acquisition. Generally, the remaining hemisphere takes
over language functions and children recover well. Removal of the left hemisphere of an adult,
however, generally interferes grossly with language function.
Fourth, studies using magnetic resonance imaging (MRI) find that brain activity in the left
hemisphere during the production of speech increases with the age of the child.
Fifth, studies of brain injury show that cerebral lateralization occurs. Trauma or disease can result
in the loss of linguistic function. The condition is called APHASIA. The older individuals are, the
more likely they are to experience aphasia if the left hemisphere is involved in the injury.
Evidence for a Critical Period for Native Language Acquisition
First, children must be exposed to a language before they reach puberty to acquire a native language.
If they are isolated so that they do not hear a language, they may never be able to learn one.
Second, intellectually disabled individuals can make progress in acquiring language up to puberty;
after that, progress is slow and difficult.
Third, the closer to age two that hearing-impaired children are fitted with hearing aids, the more
language they are exposed to and, consequently, the more enhanced their language development is.
Fourth, sudden acquired deafness, as might result from a condition like meningitis, has a profound
effect on language the younger the affected individuals are.
Fifth, the acquisition of a second language becomes more like native language acquisition the
younger the acquirer is. In other words, in America, it would benefit children if they were exposed
to foreign languages in elementary school or before.
All the evidence above points to a biological and maturational basis for native language acquisition.
Thus, regarding theories of language acquisition, we will take a largely rationalist approach. At the
same time, we must account for variation. As we have seen, speakers do not always share the same
37

38
judgements about grammar, which means that the rules that speakers have internalized during their
development do have variations. Despite these differences, the rules must be largely the same since
native speakers of the same language do communicate effectively.
1.4 THE STUDY OF GRAMMAR
In the preceding discussion, the word GRAMMAR was used in several distinct, yet interrelated
senses. This is typical because the term has a variety of applications. Fundamentally and most
commonly, it is used to mean a description of a language, that is, an explicit characterization of the
structure of the sentences in a particular language. A grammatical characterization of English, for
instance, isolates and categorizes the basic linguistic elements found in English sentences, indicates
how these elements are related to each other, and specifies the manner in which they are arranged
in larger units. Terms like noun and verb are examples of labels which grammarians have given to
some of these elements, and a statement like the subject of an English sentence precedes the
predicate is an example of the kind of rule usually found in a grammar of English.
In a second more technical sense, the word grammar is used to mean a theory of language, that is,
a system of hypotheses formulated to account for the various features of human language in general.
For example, one of the outstanding features of all human languages is the fact that each one consists
of a potentially infinite number of sentences. In effect, every human language allows for
CREATIVITY. Although there are severe constraints on individual structures when they are
combined to form complex sentences, every human language is RECURSIVE. This means that
there are structures that can repeat themselves indefinitely. Thus, it is impossible to make a list of
all of the sentences in the English language; additions to any such list can always be made by
combining the sentences one has already thought of with words like and, but, or, etc. There is no
longest English sentence. For example, (35) could, theoretically, be continued without end.
(35) This is the cat that caught the rat that ate the mouse that lived in the house that Jack was in
when he told Harry that Bill said that Mary wanted to know if Sue would be able to tell her
father that Jane hates fairy tales because they make her think of an unfortunate experience
which occurred one day when she was visiting her aunt in Idaho and...
These two uses of the word grammar, i.e., grammar as characterization of a particular language (e.g.,
English grammar) and grammar as a theory of human language generally (UNIVERSAL
GRAMMAR), are basic to all of the other more specialized or expanded uses of the word. In most
schools of modern linguistics, for example, grammar, as a description of a language, includes the
study of sounds and the way sounds are related to meaning. This usage is merely an expansion of
the more common one in which a grammar only describes the classification and arrangement of the
forms of a language, the syntax. Moreover, it is important to realize that, although there is a
distinction between grammar as description and grammar as theory, nevertheless, each of these basic
uses implies the other. On the one hand, a grammatical description is always written within the
framework of some grammatical theory, however loosely formulated, and, on the other hand,
grammatical theories are proposed so that correct grammatical descriptions can be written.

In addition to this, linguists make a distinction between DESCRIPTIVE GRAMMAR and
PRESCRIPTIVE GRAMMAR. In descriptive grammar, no value judgments are attached to the
characterization, that is, the grammar characterizes whatever speakers say. In prescriptive grammar,
on the other hand, value judgments are attached to the characterization, so that prescriptive grammars
prescribe what one should say and why, for example, that one should say It’s I and not It’s me, if one
wishes to sound educated. Throughout the following discussion, we will be concerned with only
descriptive grammar; therefore, our description must provide for utterances like It’s me as well as
utterances like It’s I.
1.5 SPEECH AND WRITING
One of the most common misconceptions about language is the belief that writing is somehow more
important than speech. It is not difficult to imagine how this misconception arose: writing has
always been considered a sign of intelligence and of civilization; it is generally more permanent and
durable than speech; and, through it, the artistic and scientific achievements of humankind have been
recorded and preserved. Although there may be some justification for this cultural superiority of
writing over speech, nevertheless, it is speech and not writing which is basic both to language itself,
as a means of communication, and to linguistics. There are several reasons for this.
First, writing is a relatively recent invention. Even the most ancient examples of writing that have
been preserved are only about seven thousand years old. On the other hand, it seems safe to say that
humans have been speaking since the evolution of human society many hundreds of thousands of
years ago.
Second, writing is much less widespread than speech. Many languages, even today, exist only in
spoken form. Furthermore, there are many extinct languages about which very little is known,
except that they were spoken in a particular area at a particular time, because these languages were
not written down.
Third, it is speech and not writing which human beings are predisposed to learn. The acquisition of
speech in children proceeds automatically and without conscious attention or specific instruction,
unlike the acquisition of writing skills. Acquiring writing ability, on the other hand, is a voluntary,
deliberate choice.
Fourth, and most important, writing is a reflection of speech. All alphabetic writing systems such
as English are based, in one way or another, on the sound systems they represent, that is, a particular
sign or letter always represents some unit of sound or sounds in the spoken language. It is never the
case that the sound system of a language is based on a particular system of writing.
For all of these reasons, writing has only a marginal importance in most language studies. Linguists
are primarily concerned with describing speech and ascertaining how it is acquired and used.
39

40
1.6 LINGUISTIC ARGUMENTATION
Linguistic competence is clearly related to linguistic performance. The facts of grammar are not
arbitrary but, in large part, are due to facts regarding the nature of human beings including their
cognitive abilities and cognitive limitations. Human beings are able to acquire human languages
because they are genetically constituted to do so. It is simply inconceivable that all native speakers
arrive at the same internal understanding of their language on the basis of trial and error, given that
the native acquisition of any language proceeds rapidly and uniformly and without instruction. In
very specific ways, human languages reflect human abilities. For example, the rules of every human
language are based on LINGUISTIC STRUCTURE, that is, on linguistic units. To see this,
consider again examples like the following which contain sentence adverbs:
(36) a. Ultimately, the student will read the book.
b. **The ultimately student will read the book.
c. The student ultimately will read the book.
d. The student will ultimately read the book.
e. * The student will read ultimately the book.
f. **The student will read the ultimately book.
g. The student will read the book, ultimately.
The possible positions that the sentence adverb ultimately can occupy are predictable. This is
because every English sentence consists of structural units. Using traditional terminology, these
units might be labeled and diagramed as follows (AUX=AUXILIARY; DET=DETERMINER):
(37)
One can see from this diagram, that the sentence adverb ultimately can occur among any of the
structural units of a SENTENCE. There are three such units attached directly to the node
SENTENCE so that the diagram (37) asserts that English sentences consist of three units or
constituents: SUBJECT, AUX, and PREDICATE. As a result, there are four possible positions
for the sentence adverb in (36), namely, (36a), (36c), (36d), and (36g). Further, sentences (36b),
(36e), and (36f) are ungrammatical because they violate a structural principle of English, and not for
some arbitrary reason: constituents that belong to one unit (SENTENCE in (37)) cannot be placed
inside of other units, that is, the integrity of units must be maintained. In short, children do not have

to be told that (36b), (36e), and (36f) are ungrammatical: their status is predictable on the basis of
other structural facts about English that they unconsciously know.
Of course, the important question here is how to determine the structural units in the first place, that
is, how to justify a diagram like (37). Such a diagram characterizes English in two ways. First, it
states that English sentences consist of three structural units; second, it states that those units should
have three different labels. Each of these statements must be justified before they can be accepted
as an accurate characterization of English.
1.6.1 DIAGNOSTICS
As we saw in the Introduction, proposals for solving problems must be justified by independent and
objective evidence if they are to be successful. Without such justification, proposals become a
matter of personal opinion and a potential blueprint for failure. The next two sections outline several
diagnostics that linguists use to support their analyses. In the chapters that follow, we will consider
other tests. In using such diagnostics, linguists operate like many other professionals from doctors
to auto mechanics.
Doctors use diagnostics to figure out what is wrong with patients. They will take a patient’s
temperature and blood pressure, run laboratory tests, order X-rays and MRIs, and so on. These
diagnostics give doctors independent and objective measures that help them identify diseases and
other ailments. Auto mechanics work the same way, running various diagnostics to determine why
a car isn’t running properly. There is, thus, nothing particularly individual or novel about the
methodology linguists use. What makes the enterprise somewhat unusual is that the problem
involves the nature of human language, rather than more ordinary concerns like health and vehicles.
But, linguistic analysis is really nothing more than troubleshooting – nothing more than an exercise
in problem solving and finding the most cogent and robust solution possible.
With these remarks in mind, let us look at the question of identifying and labeling the units that make
up an ordinary sentence.
1.6.2 JUSTIFYING STRUCTURE: UNITS
Consider first the question of the number of units. Linguists determine structural units such as the
units of a sentence by attempting to manipulate the words in that sentence in one of three ways:
REFERENCE, OMISSION, and PLACEMENT. Great caution is needed during such
investigations because, as we have seen, the facts of language are largely hidden. It is easy to be
misled. Sometimes various manipulations produce conflicting results. This merely means that the
matter is more hidden than first imagined and that other supporting data, perhaps even from other
languages, must be looked at. For this reason, the more arguments that can be marshaled to support
a particular characterization, the more justified it is.
41

42
The diagram in (37) is a proposal about the syntactic structure of English sentences. Like all
proposals, it must be justified with independent and objective evidence. In an effort to find such
justification, consider the following diagnostics which, in fact, support the units identified in (37),
that is, the SUBJECT (the student), the OBJECT (the book), and the PREDICATE (read the book):
(38) REFERENCE: the elements of sentences can be substituted as units.
In using reference, a linguist will try to see if the putative units can be replaced as single
elements with pronouns and other pro-forms.
a. Who will read the book?
Answer: the students.
Who replaces the SUBJECT the students.
b. What will the students read?
Answer: the book.
What replaces the OBJECT the book.
c. What will the student do?
Answer: read the book.
What...do replaces the PREDICATE read the book.
d. The students said that they will read the book.
They replaces the SUBJECT the students.
e. The students will read the book and we will read it too.
It replaces the OBJECT the book.
Thus, the diagnostic REFERENCE seems to confirm the three units diagramed in (37).
(39) OMISSION: the elements of sentences are omitted as units.
In using omission, a linguist will try to see if the putative units can be omitted in various
sentences.
a. The students said they would read the book if they had to__ _______.
(the PREDICATE read the book is omitted)
b. These students will read the book, and those students will ________ too.
(the PREDICATE read the book is omitted)
Thus, the diagnostic OMISSION seems to confirm the three units diagramed in (37).

(40) PLACEMENT: other elements cannot break up units.
In using placement, a linguist will try to see how items can be moved among the putative
units.
a. Ultimately, the students will read the book.
b. The students ultimately will read the book.
c. The students will ultimately read the book.
d. The students will read the book, ultimately.
e. *The ultimately students will read the book.
Thus, the diagnostic PLACEMENT seems to confirm the three units diagramed in (37).
Notice that the manipulations in (38) through (40) have varying results: sometimes grammatical
sentences are produced; other times, ungrammatical ones. In each case, NATIVE SPEAKER
INTUITION can verify the decisions regarding grammaticality. As we have observed, this does not
mean that native speakers always agree on issues of grammaticality, nor is total agreement crucial.
What is crucial is that, for most of the manipulations, most of the results are agreed upon by most
of the speakers. That makes communication between people possible. Their internalized grammars
of English are substantially the same.
According to the diagnostics, the best characterization of the units in (38) through (40) is, in fact,
(37); that is, the three unit structure diagramed in (37) predicts (38) through (40). There is no
manipulation, for example, which demands that “students will” be analyzed as a unit. Similarly,
there is no question in English that will get the answer “will the” indicating that “will the” is not a
unit. Therefore, most linguists today argue that English sentences have three basic units.
Linguists use the manipulations of reference, omission, and placement to distinguish sentences
which superficially seem identical, but which are, in fact, very different. Consider the following
(A=ambiguous; U=unambiguous):
(41) a. He looked up her street. (A) He looked her street up. (U)
b. He looked up her address. (U) He looked her address up. (U)
c. He looked up her dress. (U) *He looked her dress up.
d. He cleaned up her dress. (U) He cleaned her dress up. (U)
Notice that manipulation of these sentences produces different results, indicating that their structure
is different. In a similar way, physicians, auto mechanics, and other professionals find that some
results on tests suggest one diagnosis, while other results suggest a different diagnosis.
Introductory linguistics students make a number of incorrect assumptions when they attempt to show
that a characterization like (37) is justified. First, they assume that the analysis should be obvious
43

44
or, at least, readily clear after a little thought. Second, they assume that there are unique, formulaic,
infallible, and unambiguous procedures available for them to apply to the problem. Third, they
assume that they do not know what the correct analysis is.
All of these assumptions are wrong. The facts of language are not at all clear, and there are no
procedures available by which one can quickly discover the best characterization of the data. If this
were correct, the job of linguistics would have been completed long ago. In truth, formal
grammatical characterization is only in its infancy. Also, all native speakers, because they are native
speakers, know which sentences, out of an infinite list of sentences, are grammatical and which are
ungrammatical. Native speaker competence demands such knowledge. Therefore, the problem is
to learn how to access the unconscious knowledge one has by asking the appropriate questions.
While there are no steps that students can take which will automatically lead to a solution, there are
diagnostics they can use for help. The trick is to behave like a doctor or an auto mechanic and
remember to use the diagnostics and run the tests.
1.6.3 JUSTIFYING STRUCTURE: LABELS
Consider now the question of the labeling of the units in (37). Labels like SUBJECT, PREDICATE,
and OBJECT are common in traditional grammar.
(42) a. SUBJECT: Person or thing performing the action of the predicate.
b. OBJECT: Person or thing affected by the action of the predicate.
c. PREDICATE: The action performed by the subject.
Observe that these definitions are functional. They state what the units do, not what the units are.
As a result, there are problems. Consider the following examples:
(43) a. The students will burn the money.
b. The students will earn the money.
c. The students will have the money.
It is clear that money is affected when it is burned (43a), but it is hard to see how money is affected
when it is earned (43b). Also, the students do not seem to be doing anything when they just have
the money (43c). Such problems arise frequently with functional definitions.
Also, consider again the following examples discussed in the Introduction (see page 9):
(44) a. The dogs frightened the little girl.
b. The little girl feared the dogs.
Applying (42) to (44a), it seems that the dogs is the subject phrase because they are causing the fear,
and the little girl is the object phrase because she is experiencing the fear. In (44b), the dogs are still

causing the fear and the little girl is still experiencing the fear so it seems, according to (42), that the
dogs is still the subject phrase and the little girl is still the object phrase. But other evidence
conflicts with that hypothesis. Notice what happens when we replace the units with pronouns:
(45) a. They frightened her.
b. She feared them.
The pronoun for the dogs is they in (45a) and them in (45b); the pronoun for the little girl is her in
(45a) and she in (45b). The change in the form of the pronouns suggests that the subjects and objects
have also changed. As a result of these inconsistencies, linguists replace functional labels with
structural labels. Subjects and objects are phrases which typically contain nouns, so they are called
noun phrases (NP); predicates are phrases that typically contain verbs, so they are called verb phrases
(VP). With these labels, traditional diagrams like (37) are replaced with structural diagrams like
the following:
(46)
S = SENTENCE AUX = AUXILIARY
VP = VERB PHRASE NP = NOUN PHRASE
V = VERB N = NOUN
DET = DETERMINER
Note that S in a diagram stands for Sentence; it does not stand for Subject.
Once we have justified phrase structure diagrams like (46), we can write PHRASE STRUCTURE
RULES like the following where the arrow means ‘consists of’; (47a) says ‘a sentence consists of
an NP followed by an AUX followed by a VP’:
(47) a. S NP + AUX + VP
b. VP V + NP
c. NP DET + N
The rules in (47) are incomplete. For example, there is no mention of adverbs or adjectives. NP do
not have to contain a determiner (Students read books), VP do not have to have objects (The students
smiled), and so on. In Chapter Three, we will expand and refine (47).
45

46
1.7 UNIVERSAL GRAMMAR (UG)
The discussion in the last section raises an important question: Why are the rules of all human
languages based on linguistic units as opposed to something else? For example, rather than
permuting a sentence adverb like ultimately among structural units to produce paraphrases like (36a),
(36c), (36d), and (36g), why don’t human languages permute units on the basis of some other
criterion like counting? As it happens, there is no rule in any human language that is based on
counting per se, e.g., there is no rule like (49) based on principle (48).
(48) The rules of human language are based on counting.
(49) Place sentence adverbs optionally after every two words.
Rule (49) might predict that ultimately could occur in every position marked with a # in (50).
(50) The students # said that # they would # read the # book as # soon as # they found # the
time.
The obvious reason that human languages do not contain rules based on principle (48) is that human
beings cannot count and speak at the same time. Try, for example, to reread the present sentence at
normal speed and count the number of words as you reread it. What is the ninth word?
Sometimes languages have rules that position elements initially or finally in sentences. For example,
in a direct question in English, the auxiliary unit (helping verb) comes first as in Will he do it?
Languages even contain rules that position elements in a specific order after an introductory
structural unit. For example, German has a rule that says the verb must be the second structural
element when a sentence begins with a unit other than the subject. But such rules as these are highly
constrained. It is never the case that a language contains a rule for positioning a unit after, say, six
other units. The reason is rooted in human memory constraints. The rules of human language are
based on structural units because human short term memory (STM) is severely constrained.
It is a well known fact that humans can only remember about five to seven bits of information
immediately presented to them, for example, no more than the amount in a local telephone number.
In early childhood, the amount is much less than in adulthood. Gradually, during early development,
as children’s cognitive capacities mature, their ability to operate within STM constraints increases.
But there is an upper limit for all humans. If the information is organized into units or chunks, then
more can be remembered. Read each of the following strings of words at normal speed and then try
to recall them from memory.
(51) a. teachers tests the discuss the may
b. the teachers may discuss the tests

(51b) is much easier to remember than (51a), and the reason is clearly that (51a) has no structure;
it is simply a list of words. (51b), on the other hand, contains units which chunk or organize the
words. We can diagram (51b) as follows:
(52)
As a result of many considerations like these, linguists have concluded that the difference between
grammatical and ungrammatical sentences in human languages is not generally arbitrary. The
ungrammaticality of sentences like (36e) (*The ultimately student will read the book) is due to the
fact that they violate some principle such as (53).
(53) The rules of human language are structure dependent.
Principle (53) is directly related to the nature of human short term memory (STM): if languages did
not have structure dependent rules which organized words into units, sentence length would have
to be very short. Therefore, it follows that the nature of language is directly related to the nature of
human beings, in particular, to human mental abilities and limitations. Notice that there is no a
priori reason why humans should have evolved the capacity (53) rather than (48). Computers can,
in fact, deal with rules based on principles like (48) and routinely do so. If extraterrestrial beings
were to land on earth speaking languages that contained principle (48), it is clear that humans would
not be able to understand them without paper and pencil analysis.
Modern linguistic theory has uncovered many principles like (53) which help to define the nature
of language and, therefore, the nature of human beings. In linguistics, a principle like (53) is called
a LINGUISTIC UNIVERSAL, and the entire set of such linguistic universals is called
UNIVERSAL GRAMMAR (UG) or CORE GRAMMAR.
UG defines what can and cannot form part of a human language; therefore, it forms part of a
human’s genetic endowment. That means that children do not have to learn principles like (53); they
are born with tacit knowledge of the fact that the rules in the language to which they have been
accidentally exposed are based on (53) and not on (48). The reason is clear: human beings must
operate within their capacities and limitations. They cannot spend the afternoon underwater, for
example, without some artificial device. They cannot outrun cougars. They cannot understand
languages that contain rules based on counting. It does not take much experience, if any, for children
to learn that they have these limitations. They are part of the biological makeup of all humans.
47

48
UG contains two types of universals: FORMAL UNIVERSALS and SUBSTANTIVE
UNIVERSALS. A formal universal is a stipulation on the organization of grammar and on the form
and functioning of grammatical rules. (53) is an example of a formal linguistic universal. It
stipulates how the rules of natural language must work, in particular, they must be based on
structure. The rules of language cannot be based on counting.
The second type of linguistic universals, substantive universals, are stipulations on the elements or
features which can occur in natural language. For example, the sounds of language are distinguished
by a number of features like the following:
(54) a. [m] [+CONSONANTAL, +VOICED, +NASAL]
b. [b] [+CONSONANTAL, +VOICED, –NASAL]
c. [p] [+CONSONANTAL, –VOICED, –NASAL]
The letter m signifies the sound [m], which is really an abbreviation for a number of phonological
features. Specifically, [m] is a consonant ([+CONSONANTAL]) which means that it is formed by
temporarily obstructing the air as it leaves the vocal apparatus; the obstruction occurs while the vocal
cords are vibrating ([+VOICED]); and the air exits through the nose in articulating [m], that is, it is
a nasal sound ([+NASAL]). Changing any of these features changes [m] to some other sound. For
example, if the air exits through the mouth rather than the nose ([–NASAL]), then [b] is produced.
If the air exits orally without the vocal cords vibrating ([–VOICED]), then [p] is produced. The
features given in (54), therefore, define the phonological structure of some English consonant
sounds. The number and types of these features is not arbitrary. For example, all sounds are either
[+NASAL] or [–NASAL]; the air exits through the nose or the mouth.
Air could potentially escape the human head from the lungs via three routes: the mouth, the nose,
and the ears. In fact, when the tympanic membrane (eardrum) is not intact, people can blow smoke
through all these orifices. Despite this, no human language contains a phonological feature
opposition [+AURAL]/[–AURAL] meaning involving the ears/not involving the ears. The reason
for this is transparent: humans cannot use their ears for the production of sounds. There is no a
priori reason why the ear should have evolved in the way it did; conceivably, evolution could have
produced an aural canal with an unimpeded connection to the outside. This would have made it
possible to expel air through the ears normally; however, that did not occur.
On the other hand, a separation of the nasal and oral passages did occur in the evolution of
mammals. From the point of view of UG, human languages utilized this separation globally, so that
most natural languages contain a phonological feature opposition [+NASAL]/[–NASAL]. This
feature opposition is possible because humans have a velum which can be lowered to close off some
of the oral passage and allow nasal modification of the air stream. Thus, we have a clear example
of how natural languages are constrained by human anatomy.
Like formal universals, substantive universals are part of the genetic endowment. They form part
of the tacit linguistic competence that all humans innately possess, quite simply because they are the

direct reflection of human beings’ capacities and limitations. Children don’t have to learn that the
sounds of their native language won’t require modification through their ears, just as they don’t have
to learn that the rules of their native language are not based on counting. They are born with the
intuitive knowledge that [+AURAL] and [–AURAL] are not possible phonological features of
human languages, just as they are born with intuitive knowledge of principle (53). Further, since all
normal children can breathe either through their nose or their mouth, children intuitively know that
their native language might contain the features [+NASAL] and [–NASAL].
Notice however that every language does not have to contain the same set of universals. Not every
language has the same rules and the same features. For example, English does not use the feature
opposition [±NASAL] to distinguish its vowels, but French does. UG simply defines the limits of
what can be part of a human language because of the capacities and limitations of the human
language apparatus (the brain and the organs of speech and hearing). Therefore, a native speaker’s
tacit linguistic competence consists of two kinds of knowledge: (i) innate knowledge or UG, which
is part of a human being’s genetic endowment and consists of the kinds of rules and features which
any language might contain; and (ii) acquired knowledge, which consists of the particular rules and
features that occur in the native language(s) to which the speaker has been exposed as a child.
We find examples of substantive universals, as expected, in syntax and semantics. Consider the
following syntactic features [±PREHEAD] and [±POSTHEAD] which indicate where in a phrase
a category can occur relative to the head of the phrase, which is the focal element:
(55) Syntax: [±PREHEAD] and [±POSTHEAD]
PREHEAD HEAD POSTHEAD
all those new books about syntax available
all those available new books about syntax
a. [+PREHEAD, –POSTHEAD]: those and new
(1) those new books
(2) *books new those, *books those new
b. [–PREHEAD, +POSTHEAD]: about syntax
(1) those books about syntax
(2) *those about syntax books
c. [+PREHEAD, +POSTHEAD]: available
(1) those available books
(2) those books available
49

50
While languages do have syntactic features like [±PREHEAD] and [±POSTHEAD], they do not, as
we have seen, have features based on counting like *[±EVEN-NUMBERED-WORD].
Substantive semantic features, as we will see in detail in Chapter Four, are based on human
perceptual capacities, especially vision. Consider the following:
(56) Semantics: [±DISJUNCTIVE] and [±CONJUNCTIVE].
a. [±DISJUNCTIVE].
(1) [+DISJUNCTIVE]: emphasizing separation
(2) [–DISJUNCTIVE]: not emphasizing separation
b. [±CONJUNCTIVE].
(1) [+CONJUNCTIVE]: emphasizing joining
(2) [–CONJUNCTIVE]: not emphasizing joining
c. DISJUNCTIVE CONJUNCTIVE
(1) from + –
(2) to – +
(3) into – –
(4) out of + +
Using the features [±DISJUNCTIVE] and [±CONJUNCTIVE], we can relate the meanings of
prepositions directly to human perception. Such features are also used to define the meaning of
nouns and verbs:
(57) a. [+DISJUNCTIVE, –CONJUNCTIVE]
Nouns: freedom, deprivation, separation, lack, empty, exit, abandonment, etc.
Verbs: free, deprive, separate, lack, empty, exit, abandon, etc.
b. [–DISJUNCTIVE, +CONJUNCTIVE]
Nouns: union, possession, supply, gift, entrance, penetration, etc.
Verbs: unite, possess, supply, give, enter, penetrate, etc.
Features like [±DISJUNCTIVE] and [±CONJUNCTIVE] are defined in terms of visual capacities
which all humans have. We do not find possible features based on capacities that human do not
have. For example, a robotic eye can compute the exact area of an object like a circle or a square
by looking at it. Humans cannot do that. We have words in language like small, tiny, large and
huge, but, for obvious reasons, we do not have words or semantic features that mean, for instance,
‘three square inches.’

The essential point is that substantive universals are directly related to the human language apparatus
and, therefore, can potentially be part of the grammar of any natural language. There is a pool of
phonological feature like [±NASAL] and [±VOICED] that are based on the human vocal apparatus,
a pool of semantic features like [±DISJUNCTIVE] and [±CONJUNCTIVE] that are based on human
perceptual abilities, and a pool of syntactic features like [±PREHEAD] and [±POSTHEAD] that are
based on human cognitive abilities (specifically, limitations on STM). Languages must select their
features from these three pools. What makes languages and dialects different is that they select
different features and distribute them differently.
1.8 LEVELS OF ADEQUACY
There are several levels of adequacy that a linguist can attain in grammatical characterization. The
first level, called OBSERVATIONAL ADEQUACY, is attained when the facts are noted. The
second level, called DESCRIPTIVE ADEQUACY, is attained when the facts are described with
generalized principles or rules. The third, highest level, called EXPLANATORY ADEQUACY,
is attained when those principles are related to the nature of the language apparatus of humans. As
a linguist moves from observational to explanatory adequacy, an ever higher level of generality is
reached. At the highest level, linguists begin to relate individual languages like English to UG and,
therefore, to offer a real explanation for linguistic phenomena. In attempting to attain explanatory
adequacy, a linguist is also able to separate out what is arbitrary and language specific, from what
is predictable in terms of UG. As a result, attaining explanatory adequacy in grammatical
characterization helps to solve the grammatical realization problem. Our discussion of the data in
(73) in the previous section was an illustration of this. As a further example, this time from
phonology, consider the following data.
At the level of observational adequacy, we observe that, in many languages, there are four major
points of articulation for distinguishing consonant sounds. Looking again at nasal sounds, we find
the following:
(58) a. [m] articulated at the extreme front of the mouth
b. [n] articulated against the alveolar ridge behind the upper teeth
c. [ñ] articulated against the palate or roof of the mouth
d. [] articulated at the extreme back of the mouth
The four nasals occur in the following words:
(59) a. [m] whim, met, simmer
b. [n] win, net, sinner
c. [ñ] Spanish words like mañana (‘tomorrow’)
d. [] wing, singer
51

52
At the level of descriptive adequacy, many phonological descriptions of these sounds classify the
four nasals as follows:
(60) a. [m] is bilabial [+BILABIAL]; produced with the two lips
b. [n] is alveolar [+ALVEOLAR]; produced by raising the tip of the tongue to
the alveolar ridge, i.e., the ridge behind the teeth
c. [ñ] is palatal [+PALATAL]; produced by raising the front part of the tongue
to a point on the hard palate just behind the alveolar ridge
d. [] is velar [+VELAR]; produced by raising the back of the tongue toward
the soft palate or velum
Notice that this description requires four phonological features and a considerable amount of
unnecessary redundancy. For example, in addition to saying that [n] is [+ALVEOLAR], one can say
it is [–BILABIAL], [–PALATAL], and [–VELAR]. To correct this, and reach a higher level of
descriptive adequacy, we can reduce the number of oppositions from four to two by reducing the
number of features from four to two as follows:
(61) a. [+ANTERIOR] articulated in the front of the mouth (at or in front of the alveolar
ridge)
b. [–ANTERIOR] not articulated in the front of the mouth (behind the alveolar ridge)
c. [+CORONAL] articulated with the blade of the tongue (the portion immediately
behind the tip) raised from its neutral position
d. [–CORONAL] not articulated with the blade of the tongue (the portion
immediately behind the tip) raised from its neutral position
With these two features, we describe nasals much more efficiently as follows, reaching a higher level
of descriptive adequacy than (60), which requires four features:
(62) a. [m] [+ANTERIOR, –CORONAL]
b. [n] [+ANTERIOR, +CORONAL]
c. [ñ] [–ANTERIOR, +CORONAL]
d. [] [–ANTERIOR, –CORONAL]
These sounds do not have equal frequency or distribution in English. The closest approximation to
a palatal nasal in English occurs in a word like onion. The velar nasal, though it can occur medially
and finally, cannot occur initially. The three primary positions in (62), namely, (62a), (62b), and

(62d), are also the positions for the six stop consonants in English. Consider the following (recall
that [+VOICED] means that the vocal cords are vibrating and [–VOICED] means they are not):
(63) a. [p] [+ANTERIOR, –CORONAL, –VOICED]
[b] [+ANTERIOR, –CORONAL, +VOICED]
b. [t] [+ANTERIOR, +CORONAL, –VOICED]
[d] [+ANTERIOR, +CORONAL, +VOICED]
c. [k] [–ANTERIOR, –CORONAL, –VOICED]
[g] [–ANTERIOR, –CORONAL, +VOICED]
Notice the succinctness of this method of characterization. As we noted in previous sections, a letter
like p is a representation for the sound [p], which itself is an abbreviation for a constellation of
phonological features. Further, all those features relate to the structure of the human vocal apparatus
so that they comprise a NATURAL CLASS of sounds. As it turns out, it is precisely those natural
classes that are operative in phonological processes, which is why they are termed “natural.” For
example, consider the following words:
(64) a. impossible, imbalance, immeasurable
b. intangible, indiscrete, innumerable
c. incongruent, ingratitude
Observe that the prefix meaning ‘not’ is pronounced in three ways:
(65) a. [im] before words beginning with sounds that are [+ANTERIOR, –CORONAL],
namely, [p, b, m]
b. [in] before words beginning with sounds that are [+ANTERIOR, +CORONAL],
namely, [t, d, n]
c. [i] before words beginning with sounds that are [–ANTERIOR, –CORONAL],
namely, [k, g]
This is, of course, not arbitrary. Sounds assimilate to each other, that is, they tend to become like
each other, when they are adjacent. We can formulate the following rule for English:
(66) If a prefix ends in a nasal, assimilate that nasal to the following consonant sound.
(66) accounts for the occurrence of the bilabial nasal [m] before the bilabial stops [p] and [b], of the
alveolar nasal [n] before the alveolar stops [t] and [d], and of the velar nasal [] before the velar
stops [k] and [g]. Notice that this distribution corresponds to the natural classes formed by
combinations of the features [±ANTERIOR] and [±CORONAL] in (62) and (63). We find
53

54
verification for this distribution in the occurrences of the Latin prefix con– and the Greek prefix syn–
both meaning ‘with’ or ‘together’:
(67) a. [m] compatible, combustible, community
b. [n] continue, condense, connect
c. [] conclude, congruent
(68) a. [m] sympathy, symbiosis, symmetry
b. [n] syntax, syndrome, synnema
c. [] synchrony, syngamy
We also find other cases of assimilation in the phonology of English. For example, note the
alternation of [p] and [b] in the root of inscription, scripture as opposed to scribe, scribble; the
alternation of [s] and [z] to indicate the plural of taps, bets, picks as opposed to tags, beds, pigs.
Similar examples are very common in the languages of the world. In view of this, we can formulate
the following principle for UG, thereby reaching the level of explanatory adequacy:
(69) a. Assimilate [+NASAL] consonants to the place of articulation of a following stop,
that is, to the features [ANTERIOR] and [CORONAL].
b. Assimilate adjacent [–NASAL] consonants in voicing.
Of course, (69) is directly related to the nature of the human vocal apparatus. It is more natural to
make adjacent features similar because it is easier to articulate them when they are similar. Children
engage in such simplification from the earliest stages of language acquisition when their vocal
apparatus is still not completely formed. For example, they say things such as gig kig for big pig
assimilating the stops to the same place of articulation because it is too difficult for them to articulate
a syllable like big which begins with an anterior stop and ends with a nonanterior one; it is easier to
make them both the same, so children say gig.
With regard to levels of adequacy, the above characterization proceeds as follows. If it stops at
simply noting the data in (64), then only observational adequacy would be attained. If it stops at the
rule (66), which describes (62) in generalized terms applicable to other data (cf. (67)), then
descriptive adequacy would be attained. If the characterization continues to principle (69) and
relates (69) to the structure of the human vocal apparatus, then explanatory adequacy has been
attained.
The significance of the highest level is that one can separate what is expected to occur in natural
language from what is not expected to occur. Also, at the level of explanatory adequacy, the real
nature of apparent exceptions is often clarified. To see this, notice that the rule (66) is not always
followed, even in English. Consider the following data involving the prefix non– meaning ‘not,’
which is also of Latin origin:

(70) a. [n] nonperson, nonbiodegradable, nonman
b. [n] nontaxable, nondelivery, nonnative
c. [n] noncombat, nongaseous
Notice that non is always pronounced with a final [n]. We must say, therefore, that, if non is a
prefix, it is an exception to (66). However, the matter does not stop there. The prefixes in– and con–
are of Latin origin, and, in Latin, they also function as prefixes, though they are related to
independent words (in and cum). Similarly, syn–, which is of Greek origin, functions as a prefix in
Greek, though it is related to an independent Greek word (syn). The element non, on the other hand,
never functions as a prefix in Latin; it occurs only as a separate word meaning ‘not’ and in a few
compound words. Even in English, the merger of non with the following word is not as complete
as the merger with in– or con– or syn–. Notice the slight pause and shift of articulatory posture after
saying non in the examples of (70). Such hesitation and adjustment does not occur in the examples
in (64), (67) and (68). This indicates that our characterization has missed something. Even in
English, it is probably wrong to consider non a prefix, though it looks like one. If we treat words
like nonman as compounds like snowman or manhole, then non is not a prefix and, accordingly, is
not subject to rule (66). Thus, an apparent exception, after closer examination, is seen not to be an
exception at all.
The significance of the above discussion is that it illuminates the nature of linguistic processes and
clearly indicates that languages are not governed solely by arbitrary conventions. As a further
example, consider the following.
The past tense in English verbs is variously pronounced [t] as in kicked, [d] as in hugged, and [d]
as in hunted (the symbol [] stands for the sound of “a” in sofa). The occurrence of each of these
variants is predictable in terms of the phonetic environment in which the past tense occurs. If it
occurs after a voiceless consonant, it is realized phonetically as [t], e.g., cooked, hoped, and laughed;
if it occurs after a voiced consonant or after a vowel, it is realized phonetically as [d], e.g., begged,
rubbed, and loved; and, if it occurs after [t] or [d] themselves, it is realized phonetically as [d], e.g.,
prodded, pointed, and suggested. We can state the following generalizations ([æ] = “a”in bat).
(71) a. [t] after [–VOICED] (backed is pronounced [bækt])
b. [d] after [+VOICED] (bagged is pronounced [bægd])
c. [d] after [+STOP, +ANTERIOR, +CORONAL], i.e, after [t, d] (batted is
pronounced [bætd])
This distribution is, of course, largely an assimilatory process (and notice, incidentally, how
irrelevant and unrevealing the spelling of the past tense –ed is to the linguistic analysis). The
exception in distribution is apparently the last environment (71c) where [t, d] do not completely
assimilate in words endings in [t, d] (pointed and prodded). However, other facts of English indicate
that this distribution is, in fact, not exceptional. English has no final long sounds; there is no
difference between but and butt from a phonological point of view. Furthermore, final stops in
English are unreleased, that is, they are articulated without full explosion (compare the p in pit with
55

56
the p’s in tiptop). Therefore, in order to keep the past tense suffix as a recognizable and
pronounceable entity in a word, English must give it the status of a separate syllable in words like
pointed and prodded where it would otherwise not be possible (*pointt and *prodd).
It is clear from the above examples that the correct formation of the past tense in English is not
arbitrary for regular verbs (that the past tense of go is went is irregular and unpredictable and
therefore marked). Of course, it is completely accidental that English should choose the alveolar
stop consonant series [t]/[d]/[d] to signal past tense. However, given the choice of alveolar stops,
the distribution of these three sounds is not arbitrary. As we have seen, the nature of the human
vocal apparatus makes certain combinations of sounds difficult or impossible to articulate. For
example, it is not possible to articulate a long unreleased final stop. If the past tense is signaled by
an alveolar stop, and that suffix is added to a verb whose root already ends in an alveolar stop, then
there is no choice but to affix the [d] option. The past tense of hunt cannot be [huntt] or [huntd],
both of which end in consonant clusters that are impossible for English. It must be [huntd]. For
this reason, children often say hurted and even goed or goded (rather than went), although they have
never heard such forms. If these verbs were regular, the past tense would be exactly what children
invent. In language acquisition children immediately overapply a rule. Later, they begin the process
of making finer and finer distinctions, including learning the marked exceptions. In short, for a good
many verb forms, the past tense is completely predictable given the choice of alveolar stop and the
constraints on human articulation. Children do not have to learn the past tense of every verb
separately, only of the irregular verbs.
This analysis offers a real explanation for the rapidity with which children acquire knowledge of the
past tense in English. The explanation is given in terms of UG, which is simply a reflection of the
human language apparatus. Not only does this description of past tense explain the rapidity of
acquisition, it also explains why children regularly utter forms they never could have heard like
hurted and goed and goded. Children do not know yet that these are irregular verbs. They have
realized that English arbitrarily signals past tense by the suffixation of an alveolar stop to the verb
root. Their intuitive linguistic competence dictates that the distribution of the various alveolar stops
must obey the constraints on articulation. Therefore, children say hurted and goed. In fact, irregular
forms take children a long time to master, precisely because they violate general principles.
1.9 RELATING LINGUISTIC COMPETENCE AND LINGUISTIC PERFORMANCE
In the specification of UG, the close association between grammatical characterization and
grammatical realization becomes clear. Formal and substantive universals, which appear as the rules
and features in the grammars of human languages, are grounded in the linguistic capacities of
humans. Therefore, the discovery of a universal in the course of grammatical characterization can
often lead to a hypothesis about the way humans process information. An example of this is the
postulated formal universal (72a) which, it might be argued, leads to the perceptual strategy (72b).

(72) a. The clause is the major structural unit in human language; that is, the maximal
specification of the syntactic rules of language is the clause.
b. The clause is the major perceptual unit in human language; that is, if a string of
words can be interpreted as a clause, then it must be interpreted as one.
(72a) claims that the major syntactic unit found in the characterization of human languages is the
CLAUSE, which is a string of words that typically contains one subject and one tensed (present,
past, future) verb. (72b) claims that ordinary language use centers around clause structures, so that
a string of words that could possibly be a clause is probably a clause. Properly, (72a) is part of
grammatical characterization, the result of attempting to characterize the structure of English and
various other languages; (72b) is part of grammatical realization. We can see the operation of (72)
in the following data.
(73) a. The coach thought that John succeeded.
b. The coach thought John succeeded.
c. That John succeeded pleased the coach.
d. * John succeeded pleased the coach.
Notice that the word that can be left out of (73a) to produce (73b). We saw the same operation in
(3). However, if that is left out of (73c), the ungrammatical (73d) results (cf. (4)).
Faced with data like (73), a linguist can assume that the ungrammaticality of (73d) is either arbitrary
or principled. We have argued above that languages would be unlearnable under the conditions
which attend ordinary acquisition if the distinction between grammatical and ungrammatical
sentences were always arbitrary. Recall that linguistic knowledge is unconscious; accordingly,
children are not taught language by their parents. They must discover the grammar of their native
language without the benefit of instruction. Furthermore, since children can learn ANY human
language and do so effortlessly and rapidly, the facts of language acquisition demand that linguists
assume that some process other than trial and error guides children. Data like (73) lead linguists to
assume that (73d) is ruled out by some principle of UG, not by some arbitrary language convention.
Data like (73) from many different languages suggest that the principle entails something like (72).
Students new to linguistic analysis typically believe that a sentence like (73d) is ruled out because
it doesn’t make any sense, that is, it doesn’t mean anything. But this is incorrect as the
grammaticality of (73c) clearly shows (cf. John’s success pleased the coach). What is wrong with
(73d) is not that its meaning is somehow bizarre; rather, it is syntactically ill–formed.
Observe that, in (73d), John succeeded is a clause; it has both a subject (John) and a tensed verb
(succeeded). Principle (72b) demands that it be interpreted as a clause. Since the string of words
John succeeded already has a tensed verb, the clausal unit is fulfilled and pleased the coach, which
contains another tensed verb (pleased), cannot be part of the same clause. The function of that in
(73c) is, in fact, to separate the two clauses. Consider the following:
57

58
(74) a. [Something] pleased the coach.
b. [That [John] succeeded] pleased the coach.
c. [John] succeeded and pleased the coach.
The bracketed elements in (74) are subjects. In (74a), the sentence consists of one simple clause
with one subject (something) and one tensed verb (pleased). In (74b), on the other hand, the
sentence consists of two clauses, where the subject of the main clause is itself another clause, John
succeeded. The function of that in (74b) is to subordinate one clause to the other, to separate them.
In (74c), the word and performs a similar function of separating two clauses, only in this case it
coordinates the two tensed verbs (succeeded and pleased), rather than subordinating one to the other.
(73d) eliminates the that, so the two clauses are no longer separated; accordingly, the sentence is
ungrammatical. It contains two tensed verbs that are not separated by either a subordinator such as
that (cf. (74b)) or a coordinator like and (cf. (74c)). More importantly, its ungrammaticality is not
arbitrary: it is predictable from (72). Therefore, children do not have to learn not to say (73d), any
more than they have to learn not to try to look through solid objects. They are born with intuitive
knowledge of the principles in (72), which evidently derive from some as yet unknown fact(s) about
human brain function. These principles predict that all sentences like (73d) will be ungrammatical
in any human language. This includes (73d), as well as (75d) and any others which violate (72).
(75) a. The player who(m) the coach picked won.
b. The player the coach picked won.
c. The player who won did a backflip.
d. *The player won did a backflip.
e. The player won and did a backflip.
Despite these remarks, it is always possible that a language might allow some particular construction
under highly specified conditions even when there is a general constraint against it. After all, it is
a well–known fact that languages do have exceptions. In such cases, it is useful to distinguish what
is expected in language on the basis of universal principles from what is unexpected.
Linguists call rules and structures that are expected UNMARKED, and those that are unexpected,
because they violate general principles, MARKED. The more tightly constrained a principle is by
the nature of human beings, the less likely it is to have exceptions and the more highly marked such
exceptions become. For example, if the equivalent of (75d) were found to be grammatical in some
language, it would be highly marked. As one would expect, marked constructions take children
much longer to acquire than unmarked ones.
In the unmarked case, we expect that languages will follow the principles of UG. For example,
consider the following formal universals using examples from phonology:

(76) The feature system of language must permit cross-categorial reference.
a. [+NASAL]: m, n,
b. [–NASAL]: p, b, t, d, k, g
c. [+ANTERIOR]: m, p, b, n, t, d
d. [–ANTERIOR]: , k, g
e. [+CORONAL]: t, d, n
f. [–CORONAL]: m, p, b, , k, g
(77) The feature system must allow precise identification.
a. [+NASAL, +VOICED, +ANTERIOR, –CORONAL]: m
b. [+NASAL, +VOICED, +ANTERIOR, +CORONAL]: n
c. [+NASAL, +VOICED, –ANTERIOR, +CORONAL]: ñ
d. [+NASAL, +VOICED, –ANTERIOR, –CORONAL]:
Notice in (76) that different sounds belong to different classes at the same time. Thus, [m] belongs
to the natural class of “nasal sounds,” “anterior sounds,” “noncoronal sounds.” This cross-categorial
reference is the unmarked case in language. It is more efficient to define elements in terms of
overlapping features than to define each element with a unique set of features. As we saw above,
phonological processes operate on natural classes. In the unmarked case, if a process applies to one
member of a natural class, it will also apply to the other members. In this way, speakers can have
experience with only one member and generalize to the other members. Such generalization
accounts in part for how rapidly children acquire their native language. Unmarked cases have
maximum predictive value.
Of course, there are deviations from the expected cases. For example, (77) lists [ñ] as an expected
palatal nasal sound for English. However, the four nasals listed are not equal to each other in
distribution and frequency. The palatal nasal [ñ] never occurs in contrastive environments like the
other nasals: [sm] (some), [sn] (sun/son), and [s] (sung). In fact, [ñ] is really [ny] in English
(onion, union, etc.). Further, [] never occurs initially, whereas both [m] and [n] do. While the
formal universals in (76) and (77) make predictions about expected cases, they are not without
marked deviations.
What we find in syntax and phonology, we also see in semantics. Consider, for example, the
following formal semantic universal:
(78) The color words of a language must divide the spectrum into contiguous segments (red,
orange, yellow, green, blue, indigo, violet).
Given (78), a language might have a word that conflates red and orange or one that conflates indigo
and violet, but it cannot have a word that conflates orange and blue.
59

60
1.10 THE GENERALITY OF LINGUISTIC UNIVERSALS
We have seen abundant examples of how language operates in predictable ways. There are general
principles underlying grammar which are rooted in the nature, capacity, and limitations of the human
language apparatus. Moreover, the principles underlying language are very general and derive from
global characteristics of human cognitive ability, characteristics which operate in many domains in
addition to language. To see this, we will compare the way humans process language to the way they
process visual stimuli.
First, consider how the brain interprets ambiguity. As we have seen, languages have ambiguous
sentences like the following which can have either of the meanings indicated:
(79) This baby is too young to stand up.
a. This baby is too young to stand up by himself.
b. This baby is too young to be stood up.
The only way for a speaker to express both meanings of This baby is too young to stand up would
be to say something like This baby is too young either to stand up by himself or to be stood up by
someone else. A speaker cannot utter This baby is too young to stand up and assume that people will
understand both possible meanings. In other words, the human brain can only process one meaning
of an ambiguous sentence at a time.
Visual perception works the same way. A Necker cube like the one below is an ambiguous figure
which can be seen in two ways. Stare directly at the cube and watch what happens.
As with the interpretation of an ambiguous sentence, the visual interpretation of an ambiguous figure
is unconscious and involuntary. One cannot not look at a Necker cube and see both orientations
simultaneously. The human brain can only process one orientation at a time. Moreover, one cannot
consciously choose which orientation to see. Thus, in this example, ambiguity in language is under
more general constraints than just linguistic universals. The human brain interprets all ambiguities
in the same way. Thus, in this case, language reflects general mental capacities.

As a second example of the generality of linguistic phenomena, consider the following sentence:
(80) The farmer looked under the fence, and the farmer looked over the fence.
Speakers probably interpret (80) to mean that the farmer gazed under the fence and then the farmer
gazed over the fence.
The second clause in (80) is actually ambiguous. The farmer looked over the fence can mean either
that the farmer gazed above the fence or that the farmer inspected the fence.
On the other hand, in the context of (80), it is unlikely that speakers would assume the second
meaning applies, that is, that the farmer inspected the fence. We see the same prejudice in visual
images. Note that two Necker cubes placed side by side are likely to be interpreted in the same
orientation.
Again, language and vision are under higher generalized constraints applicable to all phenomena the
human brain is presented with. The interpretation is unconscious, involuntary, and uniform. It is
impossible to look at the above cubes and order one’s brain to see the cube on the left in one
orientation and the cube on the right in the other orientation, and then command one’s brain to
switch the perception back and forth at will.
In this example of ambiguity in context, both language and vision operate alike indicating that there
is some higher mental capacity which dictates the way the brain processes both linguistic stimuli and
visual stimuli.
61

62
As a third example of the generality of linguistic phenomena, consider ellipsis which is the omission
of items. In English, it is possible to omit constituents from sentences when they can be understood
from their occurrence elsewhere. A sentence like (81a) can become (81b), but not (81c).
(81) a. John likes apples, Sue likes oranges, and Bill likes pears.
b. John likes apples, Sue, oranges, and Bill, pears.
c. *John, apples, Sue, oranges, and Bill likes pears.
Visual images can also include ellipses. Everyone will recognize the following figures as a butterfly,
a hexagon (stop sign), the letter “A”, and an ellipse.
Items that are left out of sentences are subject to the following condition:
(82) THE RECOVERABILITY OF DELETION CONDITION (RDC): All gaps in
sentences must be recoverable.
The RDC means that items cannot be left out of sentences unless they can be recovered from items
elsewhere in the sentence. Thus, we do not find sentences like the following:
(83) a. *John apples.
b. *Sue consumed.
A condition like the RDC also holds in visual perception. One cannot leave so much out of an image
that it becomes unidentifiable. For example, it seems doubtful that people would identify the
following image as definitely a butterfly:

As a fourth example of the generality of linguistic phenomena, consider the contextual sensitivity
of phrases. The sentence (84) is ambiguous depending on whether the chickens are going to eat or
someone is going to eat the chickens, that is, whether the chickens is the subject of the verb eat or
its object.
(84) The chickens are ready to eat.
In (85a), the chickens must be the subject of lay, while in (85b), the chickens must be the object of
cook.
(85) a. The chickens are ready to lay eggs.
b. The chickens are ready to cook.
Of course, in the world of cartoons in which chickens can perform all sorts of human activities, (85b)
might be ambiguous like (84). The point is that the phrase the chickens are ready will have different
interpretations depending on the context in which it occurs. The same is true in visual perception.
The three horizontal lines below are of equal length, but the top and bottom lines seem shorter than
the middle line because of the context in which they occur.
Again, as before, the interpretation of the images is unconscious. Contextual information is
automatically processed. As a result, in context all the following ambiguous sentences would
probably not be recognized as ambiguous.
(86) a. Flying planes can be dangerous.
b. The thief hit the lady with the hat.
c. John left the house messy.
d. The students are revolting.
e. Mary saw the bus stop.
f. They sent her baby clothes.
63

64
As a fifth example of the generality of linguistic phenomena, consider the issue of complexity. It
is possible to create sentences that are beyond the ability of the human brain to process even though
they are grammatical, that is, they don’t violate any rules in the grammar. Consider the following
sentences which gradually increase in complexity to the point that the last two examples are
completely uninterpretable.
(87) a. People who believe that the world is flat are nuts.
b. Bill said that to him that the world is flat is idiotic.
c. Bill said that that the world is flat is ridiculous.
d. That that the world is flat is idiotic is surprising to Bill.
e. Who is that that the world is flat is idiotic surprising to?
In the area of visual perception, we find the same constraint. Some images like the following are
beyond the ability of the brain to process even though a magnifying glass will show that it is drawn
correctly, that is, it is “grammatical.”

The existence of examples like (87d) and (87e) means that we must make a distinction between
sentences which are bad because they violate a rule in the grammar and those which are bad because
the brain cannot process them. Linguists call the former “ungrammatical” and the latter
“unacceptable.” Here is a synopsis with examples.
(88) GRAMMATICAL/UNGRAMMATICAL
a. Grammatical:
She is mixing up her children.
She is mixing her children up.
b. Ungrammatical:
*She her children mixing up is.
(89) ACCEPTABLE/UNACCEPTABLE
a. Acceptable:
He passed up the chance to look up the name.
b. Unacceptable:
*He passed the chance to look the name up up.
(90) IMPORTANT POINTS
a. Acceptability is a technical linguistic concept that has nothing to do with
appropriateness.
b. Acceptability is only an issue when the question of brain capacities arises.
c. None of the following sentences involve acceptability:
(91) EXAMPLES
(1) John mixed up the recipe.
(2) John screwed up the recipe.
(3) John f***ed up the recipe.
a. Grammatical but Unacceptable: They attributed the painting [that was stolen from the
Louvre last night despite the intensive security throughout the building] to Picasso.
b. Ungrammatical but Acceptable: They attributed to Picasso the painting [that was
stolen from the Louvre last night despite the intensive security throughout the
building].
65

66
c. Ungrammatical and Unacceptable: They attributed to Picasso [a Spanish-born French
painter, sculptor, graphic artist and ceramist, considered by many to be the greatest
artist of the 20th century] the painting [that was stolen from the Louvre last night
despite the intensive security throughout the building].
Returning to the main topic, as a sixth example of the generality of linguistic phenomena, consider
the matter of incongruity. The following sentences make no sense because the parts of the sentences
are incongruous:
(92) a. Bill convinced Mary that the world is round, but she didn’t believe him.
b. Now that the surgeon has removed all of his vital organs, John feels better than he
ever has.
c. The ice sculpture that melted in the sun just shattered into bits.
d. Both of my parents are married to my grandmother.
We also see incongruity in visual perception. The following figure also makes no sense because the
parts are incongruous:
As before we have a match between linguistic stimuli (sentences) and visual stimuli (images). Some
higher cortical constraints are operating in both language and vision which dictate how the stimuli
must be interpreted. None of the six examples are the results of conscious decisions. The human
brain automatically imposes an interpretation on the stimuli. The interpretation of all the visual
stimuli above is the same for all human beings regardless of their native language because all human
beings have the same visual apparatus. Similarly, the linguistic stimuli given above in English can
be replaced by similar examples in all human languages.

1.11 LINGUISTIC UNIVERSALS AND HUMAN MENTAL ABILITIES
As we have seen, human languages are governed by the constraints, capacities, and limitations of
the human brain. We have already seen the effects of STM limitations. The sentence in (93a) can
be paraphrases as (93b).
(93) a. John called up a girl
b. John called a girl up.
Most, though perhaps not all, speakers will find (94b) to be an acceptable paraphrase of (94a).
(94) a. John called up a girl he used to go with.
b. ?John called a girl he used to go with up.
On the other hand, all native speakers will find (95b) to be unacceptable, although it is technically
grammatical (violates no rules).
(95) a. John called up a girl he used to go with back in Idaho before he left for college.
b. *John called a girl he used to go with back in Idaho before he left for college up.
The reason that (95b) is unacceptable is because the preposition up is separated from called by a
phrase that is very long and internally complex. By the time one gets to up, one has forgotten called.
In addition to memory limitations illustrated above, there are limitations on how complex sentences
can be. The easiest sentences for the brain to process are those which contain multiple branching
constructions.
(96) MULTIPLE BRANCHING CONSTRUCTION: one with no internal structure, that is,
no internal dependency relationships. Multiple branching constructions simply involve a
string of concatenated elements.
a. I saw John, Bill, Mary, Sue,...
b. [John] + [Bill] + [Mary] + [Sue] + ...
Sentences with multiple branching constructions can theoretically go on indefinitely. They are
constrained only by things such as boredom and the need to sleep:
(97) This morning I woke up at 7:00, got out of bed, went to the bathroom, took a shower, got
dressed, ate breakfast, brushed my teeth, checked my email, let out the cat,....
Another type of construction that is very easy for the human brain to process is a right branching
construction as follows:
67

68
(98) RIGHT BRANCHING CONSTRUCTION: one whose internal dependency
relationships branch to the right, that is, the most deeply embedded constituent is the one
furthest to the right.
a. This is the cat that saw the rat that ate the cheese...
b. [this is [the cat that caught [the rat that ate the cheese] ] ]
What makes a right branching construction so easy to process is that each proposition is completed
before the next proposition is begun. Nothing needs to be kept in STM for very long. On the other
hand, left branching constructions like the following quickly lead to unacceptability:
(99) LEFT BRANCHING CONSTRUCTION: one whose internal dependency relationships
branch to the left, that is, the most deeply embedded constituent is the one furthest to the
left.
a. Mary’s brother’s wife’s dog is named Spot.
b. [ [ [ [Mary’s] brother’s] wife’s] dog]
Nested constructions like (93b) are fairly easy for the brain to process.
(100) NESTED CONSTRUCTION: phrase á and phrase â form a nested construction if phrase
â falls totally within phrase á , and if there is some nonnull element to to the left and to the
right of phrase â within á.
a. John called Sue up yesterday.
b. John [called [Sue] up] yesterday.
As we saw in (95b), when the phrase inside the nest is long and complex, the nested construction
becomes unacceptable.
By far, the most difficult constructions for the human brain to process are center-embedded
constructions.
(101) CENTER (SELF)-EMBEDDED CONSTRUCTION: phrase â is center-embedded in
á if â is nested in á and, furthermore, â is a phrase of the same type as á.
a. [1] Something upset Ann.
[2] That something annoyed Sue upset Ann.
[3] *That that something pleased John annoyed Sue upset Ann

. [1] The man killed the cat.
[2] The man that the woman loves killed the cat.
[3] *The man that the woman that the boy annoyed loves killed the cat.
c. [1] The boy cried.
[2] When the cat died, the boy cried.
[3] *Because when the cat died the boy cried, the girl left.
Center-embedded constructions are unacceptable in all languages. The following examples are from
German, which places verbs at the end of certain clauses. An English sentence like (102a) translates
as (102b) in German, with all the verbs piled up (a word-for-word translation is given in (102c)).
Such a sentence is unacceptable in German.
(102) a. When the man saw the woman who grabbed the dog because the boy said that it bit
the girl who was sick, he laughed.
b. *Als der Mann die Frau die den Hund weil der Knabe dass er das Mädchen das krank
war biss sagte packte sah, lachte er.
c. When the man the woman who the dog because the boy that it the girl that sick was
bit said grabbed saw, laughed he.
Of course, it could be the case that an English or German speaker might want to express the
sentiments in such unacceptable sentences. In both language a paraphrase exists which contains
right branching constructions as follows:
(103) a. It upset Ann that it annoyed Sue that something pleased John. (=101a3)
b. The cat was killed by the man who was loved by the woman who was annoyed by the
boy. (=101b3)
c. The girl left because the boy cried when the cat died. (=101c3)
d. Als der Mann die Frau sah, die den Hund packte, weil der Knabe sagte, dass er das
Mädchen das krank war biss, lachte er. (=102b)
69

70
1.12 LANGUAGE VARIATION
All humans languages are based on UG, a specification of the kinds of rules and features that can
exist in human languages. As we have seen, this specification directly reflects the nature of the
human language apparatus, which includes the organs of speech and hearing and the brain. The
existence of UG accounts for the fact that any human being can acquire any human language as a
native language with equal facility.
The principles in UG are available to virtually all humans; in fact, there are only three groups of
children who fail to acquire language normally. The first group includes those children who have
been kept severely isolated: one cannot acquire a language if one is not exposed to a language. The
second group includes the profoundly disabled intellectually, children who fail to attain a mental age
above two. Lastly, psychotic children show severe limitations in native language acquisition, usually
because they do not attend to language when they are exposed to it. Since all of these groups are
extreme cases, absence of the ability to acquire language is very rare in the overall population.
It is important to note also that language can develop quite normally even in cases of severe sensory
and motor deprivation. Children who are deaf, mute, and blind can acquire language, as the
accomplishments of Helen Keller attest.
Given the nature of children who fail to acquire a native language and the presence of language in
deaf, mute, and blind children, it is clear that the major component of UG is neurological. There is
something about the nature and functioning of the human brain that makes human language possible.
For this reason, general principles of grammar are very difficult to discover. It is well to keep this
in mind, as we explore the mysteries and intricacies of grammar in later chapters.
In the search for the grammatical principles which underlie all human languages, it is easy to be
mislead. There are many superficial differences among the world’s languages, and even among
dialects of the same language. Cultural and ethnic diversity among different groups of people add
even more superficial complexities. Students new to linguistics are quick to focus on these
superficial variations and tend to overestimate their importance. For example, because different
languages sound so different, students often think that some are easier to pronounce than others.
This may be true when one acquires a second language as an adult, but it is not true for native
language acquisition: all other things being equal, no language is harder for a child to pronounce
than another. Also, although different languages organize objects in the world differently, there is
no organization that cannot be acquired by every child. The fact that nomadic tribes living in deserts
have many words for different kinds of wind and sand simply tells us that wind and sand are an
important feature of desert life. That is hardly surprising news.
Despite the enormous variety among the world’s languages and the people who speak them, any
child could potentially acquire any language as a native language. This fact is of crucial significance
in modern linguistic theory. While there is currently much debate over what the principles of UG

are and what grammatical model best expresses them, virtually all linguists agree that the grammars
of human languages reflect innate, biologically determined characteristics of human beings.
1.13 LANGUAGE AND CULTURE
Culture and language are totally different objects of inquiry. Culture is conventional, subject to
whim and fancy, and alterable by consensus. For example, what is considered appropriate attire on
an airplane or in a concert hall today would not have been accepted twenty years ago. The number
of taboo words and taboo subjects has dropped precipitously in the same period. It is possible to
decide whether or not we can burn the flag, address elders by their first names, eat meat on Friday,
or publicly breast-feed an infant. These aspects of culture, like all others, endure or change at the
whim of the society that institutionalizes them.
The structure of language, on the other hand, is not conventional, not subject to whim and fancy, and
not alterable by decree. Consider, for example, the following differences between statements and
direct questions:
(104) Statements Direct Questions
John is going. Is John going?
John has gone. Has John gone?
John will go. Will John go?
Clearly, direct questions in English are formed by inverting the order of the subject and the auxiliary
or helping verb (verbs like is, has, and will). If a sentence lacks an auxiliary, then an appropriate
form of the verb do must be inserted and the tense must be taken off the main verb and put on the
form of do:
(105) Statements Direct Questions Ungrammatical Direct Questions
John went. Did John go? *Do John went? *Went John?
John snores. Does John snore? *Do John snores? *Snores John?
It takes children a relatively long time to work out the details of forming direct questions in English,
and the insertion of do causes people learning English as a second language considerable difficulty.
So, why not get rid of these processes? After all, in indirect questions, these processes don’t apply.
Consider the following indirect questions:
(106) a. Ask Mary if John is going.
b. Ask Mary if John has gone.
c. Ask Mary if John will go.
d. Ask Mary if John went.
e. Ask Mary if John snores.
71

72
Why not form direct questions in the same way, i.e., simply prefix statements with a word like if as
many other languages do? For example, let us decree that all the following will be grammatical
direct questions from now on:
(107) a. If John is going? (Replaces Is John going?)
b. If John has gone? (Replaces Has John gone?)
c. If John will go? (Replaces Will John go?)
d. If John went? (Replaces Did John go?)
e. If John snores? (Replaces Does John snore?)
Think of the aggravation that people would be spared. Of course, the idea is absurd. We cannot
change how we form direct questions in English by legislation, mutual consent, or even conspiracy.
Language reflects culture in only the most superficial ways. For example, specific kinship
terminology is influenced by culture. In ancient Rome, there were two different words for uncle,
avunculus (mother’s brother) and patruus (father’s brother) because the distinction was important
in Roman society. In the evolution of Latin to the modern Romance Languages, e.g., French, the
distinction was lost because it was no longer relevant. Such subtleties and changes in usage are
common in languages, as are the societal conventions that dictate what is proper (whether to say I
need to use the bathroom or I gotta take a leak, for example). Thus, the fact that a language has two
words for uncle is entirely conventional and based on societal concerns, whereas the structure of
languages conforms to the capacities and limitations of human cognitive ability, which is entirely
independent of social conventions. It is not an accident, for example, that the sentences of every
human language can be broken up into phrases. The fact is that the human brain cannot process
unstructured material very well so sentences must be organized according to structural principles that
all languages share. The manipulation of phrases is also dictated by such structural principles.
Notice that it is phrases that invert to form direct questions in English, not words:
(108) Statements Direct Questions
John is going. Is John going?
That man is going. Is that man going?
The man laughing is going. Is the man laughing going?
The man who is laughing is going? Is the man who is laughing going?
The man who is laughing is going? *Is the man who laughing is going?
Every native speaker of English knows that the italicized sequences in (108) are phrases. Of course,
their knowledge is generally unconscious as we have seen. If asked to identify the subjects in those
sentences they might even get the answer wrong. But, at an unconscious level, they know exactly
what all the subjects are because they know it is possible to replace all of them with the pronoun he.
Further, they know they can replace the man laughing with he, but not just the man. The
ungrammaticality of *He laughing is going has nothing to do with English per se; it is
ungrammatical because it violates the structural principles on which all human languages are based.

The structural principles that apply when the phrases in (108) are replaced with pronouns operate
in other uses of pronouns. Consider the following sentences and notice that the first them in (109b)
refers to the children while the second them refers to the matches; further, the interpretation of the
pronouns parallels (109c), which is grammatical, not (109d), which is ungrammatical.
(109) a. The children should not play with the matches.
b. Why did you give them them?
c. Why did you give the children the matches?
d. *Why did you give the matches the children?
Perhaps more interesting is the fact that the reduction of the word them to [m] in rapid speech also
follows general principles. Note that (110a) and (110b) are grammatical, but (110c) is not.
(110) a. Why did you give m them?
b. Why did you give m m?
c. *Why did you give them m?
Grammatical principles dictate how pronouns are interpreted even when the interpretation varies in
sentences that look almost identical like the following:
(111) a. John appeared to Bill to behave himself. (himself = John)
b. John appealed to Bill to behave himself. (himself = Bill)
The referent for himself in these sentences is corroborated in the following:
(112) a. *She appeared to him to behave himself. (himself = she)
b. She appealed to him to behave himself. (himself = him)
Given data like the above, it is clear that cultural matters do not underlie the principles that apply in
the formation of questions, in the replacement of phrases with pronouns, in the reduction of words
and phrases during rapid ordinary speech, and in grammatical processes generally. Rather, such
principles are based on grammatical structure, which, in turn, is based on the nature of the human
language apparatus.
73

74
1.14 TYPES OF LANGUAGES
It is natural to wonder where human linguistic ability came from. Quite often, people assume that
human language is just a complex version of animal communication systems. That is not the case.
At a fundamental level, human language has no significant structural relationship to other types of
communication. To see this, we begin with the following classification of types of languages.
(113) a. Natural Languages.
[1] Animal Languages.
[2] Human Languages.
[a] Vocal/auditory languages like English, Russian, Chinese, etc.
[b] Gestural/visual languages like American Sign Language (ASL), British
Sign Language, Chinese Sign Language, etc.
b. Artificial Languages.
[1] Nonnatural Human Languages like Esperanto, Klingon, Romulan, etc.
[2] Computer Languages.
[a] Programming languages like Python, Lisp, Cobol, etc.
[b] Word processors like Word Perfect, Word, Open Office, etc.
Each of these types of languages has special characteristics. Beginning with animal languages, we
find that they can all be subsumed under two classes which we will simply call Type One and Type
Two. They have the following characteristics:
(114) Type One.
a. Description: the system consists of a fixed number of signals each one of which is
associated with a specific range of external conditions or internal states.
b. Example: Howling Monkeys
[1] A voluminous roar signifies defense of group territory.
[2] A rapid grunting (male) signifies a disturbance.
[3] A terrible bark (female) signifies a disturbance.
[4] A purr is used to foster mother/young relationship.
[5] A series of cries (usually three) signifies a falling infant.

c. Example: Honeybees
(115) Type Two.
[1] Round Dance: used when food is fairly close to the hive.
The forager bee travels in a small circle on a honeycomb cell.
[2] Sickle Dance: used when food is neither close nor far away from the hive.
The forager bee does a crescent-shaped dance on a honeycomb cell.
[3] Waggle Dance: used when food is far away from the hive.
The forager bee does a dance in the shape of a figure eight with a vigorous
wiggling of the body.
a. Description: the system consists of a fixed number of linguistic dimensions each one
of which is associated with a nonlinguistic dimension such that selection of a point
along one indicates a corresponding point along the other.
b. Example: The Waggle Dance of Honeybees
In addition to indicating that the food is far away from the hive, the waggle dance
indicates the direction of the food source, the sun being used as a reference point. A
waggle dance proceeding on the vertical honeycomb upward means that the food is
in the direction of the sun; if downward, it means the food is in the direction opposite
the sun; if 60 degrees left of straight up, it means the food is 60 degrees to the left of
the sun; etc.
In each of the above examples, observe that there is a one-to-one correspondence between the signal
and the message. Each signal means only one thing and there is only one way to signal each
message. Humans use such signals also, which have only one meaning. Example include waving
hello and goodbye, holding up a hand to signal stop, sticking out a thumb on the side of the road to
hitch a ride, moaning more loudly as pain increases, and so on. These are all forms of
communication, and they are also quite specific. Sticking out a middle finger on the side of the road
does not convey the same thing as sticking out a thumb.
Even walking can be communicative. If you look out the window and see a woman walking swiftly
and glancing repeatedly over her shoulder, you would probably think the woman is trying to run
away from someone or something. Here, we could also add facial expressions and body movements,
which often indicate attitudes, emotional states, reactions, and so on.
Human language is not like any of these examples. Every human language consists of an infinite
number of signals each one of which involves the pairing of a superficial sound structure with an
underlying semantic structure. There is no one-to-one correspondence between sound and meaning
in any natural human language. All human languages have both ambiguity and paraphrase.
75

76
(116) AMBIGUITY (one sound, more than one meaning).
a. Lexical level (the level of the word).
[1] I found a bug. (bug = ‘insect’ or ‘computer problem’)
[2] They make good models. (make = ‘build’ or ‘serve as’)
b. Phrasal level.
[1] He looked over the fence. (= ‘gazed above the fence’ or ‘inspected the fence’)
[2] He hit the lady with the hat. (= ‘he used the hat’ or ‘the lady has the hat’)
c. Sentential level.
[1] Visiting relatives can be boring. (= ‘visiting is boring’ or ‘relatives are boring’)
[2] The chickens are ready to eat. (= ‘chickens eat’ or ‘someone eats chickens’)
(117) PARAPHRASE (one meaning, more then one sound).
a. The nurse gave the shots to the students.
The nurse gave the students the shots.
The shots were given to the students by the nurse.
The students were given the shots by the nurse.
b. John called up a girl.
John called a girl up.
Importantly, ambiguity and paraphrase even occur in the speech of children. In (118), we have
examples of ambiguous sentences from children; in (119), examples of paraphrase.
(118) a. Milk. = ‘That’s milk’; ‘I want milk’; ‘the milk spilled’; etc.
b. Mommy sock. = ‘That is mommy’s sock’; ‘Mommy, put on my sock’; etc.
(119) a. Get it baby up.
b. Get it me up.
c. Get baby up.
Thus, child language cannot be equated with “primitive” animal-type languages. Fundamentally,
human languages are different from animal systems even at the earliest stages of acquisition.
In addition to the above, as we have seen, human languages typically contain sentences that look very
similar on the surface, but have to have very different interpretations like the following:
(120) a. John is eager to please.
b. John is easy to please.

(121) a. Bill appeared to John to behave himself.
b. Bill appealed to John to behave himself.
The essential point is that communication is useful for animals. It fosters group cohesion, warns
members of dangers, establishes hierarchies, attracts mates, and so on. We would, therefore, expect
evolution to be fertile and produce a wide variety of communication systems in the animal world,
each constrained by the animal’s abilities and needs.
Locomotion is also useful. It allows an animal to move about to find food or avoid danger. Here
too we expect evolution to be fertile and produce a wide variety of ways to move about. The fact
that two animals move in the same way does not mean that their movement is necessarily related.
Both birds and bees fly, for example, but the mechanisms by which they fly are completely different.
Birds fly using special muscles attached to their wings; bees fly by vibrating their thorax. The
structure of the bird wing is totally different from the structure of the bee wing. Thus, no biologist
would say that the flight of birds evolved from the flight of bees, or vice versa.
Similarly, the structure of human language is completely different from the structure of animal
languages. It therefore makes no sense to say that one evolved from the other. The appearance of
multiple means of communication is an example of what biologists call CONVERGENT
EVOLUTION – the appearance of the same biological trait or behavior in species that are unrelated.
Different communication systems and different forms of locomotion are directly related to the nature
and characteristics of the organisms that exhibit them. Birds can fly because they have wings;
humans can’t fly because they don’t have wings. While this seems so obvious as to hardly need to
be stated at all, some people do think that animal communication systems are primitive forms of
human languages. They are not.
Because human language is directly related to the nature of human beings, its structure is not subject
to whim, fancy, popular consensus or even legislation. Native speakers of English, for example,
cannot decide collectively to change the way sentences are structured. Legislators cannot pass laws
about how sentences should be structured. This does not mean that languages don’t change over
time. They do. Nor does it mean that certain forms of speech can’t be viewed as appropriate while
others are viewed as objectionable. There are cultural constraints which society can change as it sees
fit, and such constraints are completely different from constraints on linguistic structure like the
internal organization of phrases. Consider, for example, the various ways in which one might
comment on the death of someone’s wife:
(122) I’m very sorry to hear that your wife
a. passed away. [probably appropriate]
b. died. [perhaps too blunt]
c. kicked the bucket. [rude]
d. went to heaven. [childish]
77

78
Or consider which of the following might be appropriate for a nurse to ask:
(123) Would you mind
a. leaving a specimen in that container. [probably appropriate]
b. leaving a urine specimen in that container. [a bit too clinical]
c. taking a leak. [rude]
d. going peepee in that paper cup. [childish]
Turning to computer languages, specifically word processors, we note that such languages also have
characteristics that make them completely different from natural human languages:
(124) a. They are based on mathematical operations like counting.
b. Their memories are flawless and expandable.
c. They are equipped with special operations like search, copy, paste, etc.
d. They are controlled by arbitrary commands.
Significantly, in a word processor, anything can be a unit for the purposes of movement, insertion,
deletion, or substitution of the text. For example, suppose one has written the sentence The student
will read the book. Using a mouse, one can perform all the following operations:
(125) a. Move Will between read and the: The student read Will the book.
b. Insert s’ teacher after student: The students’ teacher will read the book.
c. Delete h, stu, and ent: Ted will read the book.
d. Substitute sha for wi: The student shall read the book.
In short, anything can be a unit in the operations one can perform in a word processor. As we have
seen, that is not the case in human language. The operations in human language (Reference,
Omission, and Placement) are structure-dependent. The difference, of course, derives from the fact
that human beings are not computers. Human languages are based on the nature of the human
language apparatus. Computer languages are based on the computer’s capacities.
1.15 THE EVOLUTION OF HUMAN LANGUAGE
There are basically two opposing views about the evolution of human language. One view claims
that there has been a straight line evolution of language from primitive organisms to human beings
with a continual growth in the capacity for language. The contrasting view holds that human
language is species specific to human beings, meaning that the structures and functions which have
made human language possible emerged after homo sapiens became a separate species roughly six
million years ago. The pioneering work on language and biology is Eric Lenneberg’s 1967 book The
Biological Foundations of Language. In this section and later chapters, we will examine
Lenneberg’s arguments incorporating the research that his book stimulated.

Theories which assume a straight line evolution fall into two groups, which we will call
CONTINUITY THEORY I and CONTINUITY THEORY II.
CONTINUITY THEORY I claims there has been a straight line evolution of language with only
quantitative differences. This theory links the development of linguistic competence with
intelligence. The idea is that human languages differ from animal languages only in having more
features, sounds, constituents, categories, words, and so on. As we have seen, that claim cannot be
maintained. Human language is fundamentally different from animal systems. In particular, all
animal languages have a one-to-one correspondence between the signal and the message, whereas
all human languages lack that characteristic having both ambiguity and paraphrase. The differences
are attested even in the speech of children which contains ambiguity and paraphrase.
Furthermore, if this theory were correct, one would expect differences in linguistic competence
between geniuses and people with intellectual deficits. That is not the case. The only human beings
who lack the capacity to acquire a native language are those who only reach a mental age of about
two-years-old. Very smart people may have larger vocabularies and may use language in more
subtle ways than people who are less smart, but they do not have a greater ability to acquire a native
language. It is interesting that the capacity for language is available to the broadest range of human
beings. Absence of the capacity for language is extremely rare.
It is also important to note that it makes little sense to try to measure intelligence across species.
Animals are as smart as they need to be to survive and reproduce. It is somewhat odd to say that,
for example, a horse is not as smart as a spider or vice versa.
CONTINUITY THEORY II claims there has been a straight line evolution of language with
qualitative differences and a stepwise accumulation of skills. Support for this theory general comes
from citing examples of different linguistic abilities from various species in the animal kingdom.
Such random citation is not good biology. The taxonomic classification of animals is based on
shared features in individual groups, what biologists call PHYLOGENETIC relationships. Two
animals are said to belong to the same group (species, genus, family, order, class, phylum, etc.)
because of the characteristics they share. One cannot legitimately support this continuity theory by
citing examples from all over the animal kingdom without regard to phylogenetic relationships.
Further, the difficulty with such a theory is that, if it were true, we would expect to see the closest
thing to human language in our closest relatives, that is, the other primates. Again, that is not the
case.
The DISCONTINUITY THEORY claims that human language is species specific to human beings,
like bipedal locomotion. It is the result of anatomical and neurological specializations that have
occurred after homo sapiens became a separate species. Support for this theory is found in the
examination of changes in both anatomy and physiology that have occurred in human evolution.
79

80
Neurological Changes
Let us consider first the neurological changes. As we have noted, the human brain is divided into
two cerebral hemispheres. Each hemisphere is further divided into four lobes (PARIETAL,
OCCIPITAL, TEMPORAL and FRONTAL) and the different areas in each hemisphere are
designated by numbers originally proposed by the German anatomist Korbinian Brodmann.
1,2,3: Primary sensory cortex
7: Secondary sensory area
4,6: Primary motor cortex
9: Motor area for the hand (Exner’s center)
44: Motor area for the face (Broca’s area)
40: Supramarginal gyrus
39: Angular gyrus
17: Primary visual area
18,19: Secondary visual area
41,42: Primary auditory area
22: Secondary auditory area (Wernike’s area)

As we noted in our discussion of cerebral lateralization, the two hemispheres of the brain exhibit
contralateral control, with the left side of the brain controlling he right side of the body, and the right
side of the brain controlling the left side of the body. Also, recall that for most people, language is
under the dominant control of the left hemisphere.
The parietal lobes are concerned with sensation. Damage to Brodmann area 1, 2, and 3, the primary
sensory cortex, interferes with sensation. Objects held in the hand may feel too hot, too cold, too
heavy, too light, and so on. In testing for damage, one must isolate the function. In the present case,
patients need to be blindfolded so that they do not use vision in the identification. Care must also
be taken to avoid sounds and smells that might give a clue to identification. Damage to Brodmann
area 7, the secondary sensory area interferes with the identification of objects. The object may feel
normal, but the patient may be unable to name what it is.
The occipital lobes are concerned with vision. Damage to Brodman area 17, the primary visual area,
in both hemispheres would render a patient cortically blind. Damage to Brodmann areas 18 and 19,
the secondary visual area interfere with the identification of objects by sight. A patient will see a
ladder but might not be able to name it or know what its function is.
The temporal lobes are concerned with hearing and memory. Damage to Brodmann areas 41 and
42 interferes with the gross identification of sounds, which will be heard as too high, too low, too
loud, too soft, and so on. Damage to Brodmann area 22, especially the posterior portion in the left
hemisphere (for most people), which is known as Wernike’s area, produces receptive aphasia.
Patients cannot comprehend spoken language.
The frontal lobes are concerned with movement and what are generally called higher cortical
functions such as attention, creativity, organization, prioritization, and the like. The posterior portion
of the frontal lobes, Brodmann areas 4 and 6 are the primary motor cortex, responsible for the control
of bodily movements. Also in the frontal lobes in the left hemisphere (for most people) is Brodmann
area 44, called Broca’s area. Damage to Broca’s area produces an expressive aphasia. Patients are
unable to express their thoughts. They know what they want to say but cannot form their thoughts
into meaningful sentences.
In roughly the center of the hemispheres are two large association areas, the angular gyrus
(Brodmann area 39) and the supramarginal gyrus (Brodmann area 40). These areas integrate stimuli
from the surrounding areas. In the left hemisphere (for most people), they are of particular
importance for language. First, damage to these areas can produce global aphasia, patients losing
the ability to speak, understand, write, and read. Second, these areas are not present in the brains of
subhuman primates. Third, these areas are the last to develop in childhood.
The development of these association areas of the brain are crucial for the linguistic abilities of
human beings. Notice that the ordinary use of language crucially involves the association of visual
and auditory stimuli and linking those stimuli with movements of the vocal apparatus and thought.
Making such general associations between such diverse modalities is not something that subhuman
81

82
animals can do; therefore, it is important to investigate the significance of this human advancement
more closely.
Stimuli that affect animals can be grouped into two classes, LIMBIC and NONLIMBIC. Limbic
stimuli are concerned with survival and include stimuli that involve eating, drinking, sleeping,
reproduction, self-preservation, etc. Nonlimbic stimuli are not concerned with survival. They
involve seeing, hearing, speaking, etc. It is relatively easy to get an animal to make an association
between one nonlimbic stimulus and one limbic stimulus. Ivan Pavlov, a famous Russian
physiologist, showed that one could, for example, relatively easily teach a dog to make an association
between a nonlimbic stimulus like a bell and a limbic stimulus like food such that the animal would
begin to salivate as soon as the bell was heard, even before the food was presented.
It is also fairly easy to teach an animal to make an association between two nonlimbic stimuli of the
same type, e.g., two visual stimuli. This has been done successfully with chimpanzees and apes, so
that the animals can associate visual representations of objects with actual objects. In the case of
Sarah, a chimp, symbols of different colors and shapes represented objects. Sarah was able to
respond appropriately to different arrangements of the symbols expressing different situations.
What is very difficult to teach an animal to do is to make an association between two different types
of nonlimbic stimuli, such as a visual stimulus and an auditory stimulus. That ability is a
prerequisite to the ordinary use of language. It’s exactly what children must do when someone holds
up a spoon and says the word spoon. Apparently, humans alone can easily acquire this ability
because they have the two association areas that subserve the ability, the angular gyrus and the
supramarginal gyrus, which integrate visual and auditory inputs as well as other information.
In addition to the changes that have occurred in neuroanatomy, mentioned above, there has also been
an asymmetrical development of the language areas of the left hemisphere in human beings,
including the primary auditory area, Wernike’s area, and Broca’s area. This asymmetry is present
at birth and can even be detected in a five-month-old fetus, indicating that humans are genetically
predisposed to acquire a human language.
Since all these neurological developments are peculiar to human beings, it is clear that the human
brain is specialized for encoding and decoding language. The increase in the relative size of the
language areas has necessitated a decrease in the other areas of the brain. For example,
proportionately, the primary visual area in the human brain is smaller that the corresponding area of
other primates. Since the skull is a confined space, any development or increase in one area must
be offset by a decrease elsewhere. As a result, as a groups, humans do not see or hear as well as
other animals. We require more light, and sounds must be louder to be detected.
On the other hand, human language has become possible because of all these neurological changes.
Language as we know it would not be available to humans without the emergence of cerebral
lateralization, the localization of language to specific areas, and the cross-modal associations needed
for integrating multiple stimuli.

Changes in the Vocal Apparatus
Turning now to the vocal apparatus, the major change that has occurred in human evolution is the
lowering of the position of the larynx (voicebox) in the throat. In subhuman primates, as well as
newborn humans, the larynx is situation high in the throat. Gradually during the first few years of
life, the larynx becomes lower in the throat. Since a lowered larynx increases the chances of choking
to death and exposes more of the throat to infections which can be fatal, the question is, Why does
the relative position of the larynx change as it does. The answer seems to be that the lowered
position is advantageous for speech. It opens up a large area in the back of the throat, the pharynx,
thereby increasing the range of sounds that can be produced by the vocal apparatus.
It is important to note that human languages are not optimally efficient systems. They contain
significant redundancies. For example, in English, the formation of a WH-question involves
positioning the WH-phrase at the front of the sentence and inverting the positions of the subject NP
and the auxiliary. Thus, a question like (126a) becomes (126b).
(126) a. John will buy WHAT?
b. WHAT will John buy?
In addition to the changes in position of three constituents (the WH-phrase, the subject, and the
auxiliary), WH-questions involve a change in the intonation pattern of a sentence, as well as
(perhaps) a quizzical look on the face and a shrug of the shoulders. Why do we have all this
redundancy when (126a) is by itself a fine question? The answer seems to lie in the nature of the
human perceptual system. An optimally efficient system would require very careful attention: if just
one feature of the system were missed, there could be a lack of communication. Redundancy allows
for a perceptual system that is less alert. If one feature is missed, others will suffice to ensure that
communication is achieved.
A lowered larynx allows for the creation of many more phonological features that would otherwise
be possible. Neanderthal Man, for example, couldn’t produce the sounds [i], [u], [a], [k], [g], and
[], because his vocal cavity was considerably smaller. If a hypothetical language has, say, eight
sounds, it would only take three features to unambiguously distinguish those eight sounds. As an
example, consider the following:
(127) Optimally efficient system (the tilde "˜" above the vowel means that the vowel is nasal):
m b g e˜ õ e o
CONSONANT + + + + – – – –
NASAL + + – – + + – –
ANTERIOR + – + – + – + –
83

84
We do not find such optimally efficient languages that have the minimal number of features
necessary to distinguish all sounds. The phonological component of human languages are about 50%
efficient. That means that there are about twice as many features as are optimally needed. The
lowering of the larynx allows for this increase in the number of phonological features which, in turn,
makes more sounds available for human languages. Apparently, the pressures of natural selection
favored the lowered position of the larynx for language over the dangers that position created.
Other changes in the vocal apparatus include the development of the alveolar ridge, a bony ridge
between the upper teeth and the palate. Also, the teeth of other primates are uneven, while those of
humans are even, forming an unbroken palisade around the oral cavity. These developments add
positions in the oral cavity against which new sounds can be made. For example, English [t] and [d]
are articulated with the tongue touching the alveolar ridge.
There has also been an increased prominence in the muscles of the lips and complexity of the
muscles associated with articulation. Among primates, only humans can round their lips to form an
airtight closure followed by a rapid explosive opening as is necessary to make the sounds [p], [b],
[m], [w], etc. The increased mobility of the lips also makes possible sounds such as [f] and [v].
The human tongue itself is more mobile and flexible than the tongue of other primates making
possible a very wide variety of different sounds. This includes not only vowels but also interdental
sounds like[] and [ð].
These anatomical changes in the shape of the vocal apparatus have been accompanied by functional
changes. Humans are the only species that use of the tongue and pharyngeal wall to modify
phonations. Further, humans are the only species that make use of the nasal and oral separation for
the production of [+NASAL] and [–NASAL] sounds. We do not find, for example, cows that make
a distinction between [mu] and [bu].
Changes in the Auditory Apparatus
There have only been slight anatomical changes in the auditory apparatus of humans during the
course of evolution. One obvious change is that the ears cannot be moved to locate the source of
sounds except in a very few people who can wiggle their ears when parties get dull. Still, even such
talented humans do not move their ears to locate sounds; they turn their heads.
The human ear is most sensitive to sounds within the frequency range of 250-4000 cycles per second,
which is the range of normal speech. It is also most sensitive to sounds at a 50-65 decibel range,
which is the range of loudness for normal conversation.

1.16 SUMMARY
As we have seen, there is a high level of abstractness associated with the characterization of
linguistic competence; principle (53) – that the rules of human language are structure dependent –
is hardly a subject of discussion among five–year–olds. Trying to explain it even to an educated
adult is not easy. Also, there is a large measure of creativity associated with linguistic performance;
most of the sentences children experience are novel. Despite this, many psychologists and linguists
believed, until quite recently, that the linguistic abilities of human beings could be attributed to a
complex interaction of elementary learning mechanisms, such as imitation, conditioning, association,
and the like. Biologists have also shared this belief and attempted to document the separate skills
needed to explain human language by citing antecedents found elsewhere in the animal kingdom.
However, it has now become quite clear that the acquisition and use of language by ordinary human
beings cannot be explained as the product of simple learning mechanisms, each with a precedent
elsewhere in the animal kingdom. This is substantiated by a number of observations which we
summarize here.
First, human beings are capable of learning any of the world’s languages. Obviously, a child is not
predestined to acquire a specific language but learns the language of the community into which he
has been placed by chance.
Second, from the inception of speech, a child does not merely imitate his parents or respond to
stimuli but produces entirely novel sound sequences, most of which he has not specifically heard.
One of the outstanding characteristics of both language acquisition and use is their productivity.
Children say things like I goed there, which they clearly have never heard.
Third, all normal children learn their language with remarkable rapidity and uniformity; and, even
nonnormal children, such as those who suffer from severe sensory deprivation, acquire linguistic
proficiency. A human can be deaf, mute, and blind and still acquire a human language.
Fourth, the data on the basis of which children learn their language is of a highly fragmentary sort;
many of the sentences they hear deviate considerably from perfect well–formedness. Despite this,
all children learn to distinguish those utterances which are grammatical sentences from those which
are not. And all children, despite their particular home environment, arrive at essentially the same
internalized grammar.
Fifth, while the members of the linguistic community do correct children on occasion, the level of
instruction never approaches the order of complexity and abstractness which must be assumed to
account for the normal use of language.
Sixth, there is no one–to–one correspondence between sound and meaning in any human language,
whereas all known animal communication systems have very restricted one–to–one correspondences
between signals and meaning. Human languages all have sound sequences that mean more than one
85

86
thing, i.e., ambiguities like (128); they also have multiple sound sequences that can mean the same
thing, i.e., paraphrases like (129).
(128) a. It’s too hot to eat.
b. Each of the sisters said that they took out their brothers yesterday.
(129) a. The boy called up the girl.
b. The boy called the girl up.
On the basis of these observations and many others, it appears that language acquisition is possible
in children because all human languages, despite their superficial diversity, are of a universally
well–defined type and because all children have an innate mental capacity to learn this particular type
of communication system. Consequently, the essential problem for linguistics is to construct a
theory of human language which is sufficiently rich to account for the diversity among the world’s
languages, but which, at the same time, is constrained enough to account for the facts of ordinary
language acquisition. The formal representation of such a theory entails a grammar of linguistic
universals or UG, the investigation of which necessarily links the study of language structure to the
study of human psychology and biology, and ultimately, to the study of the nature of human beings.
The innateness hypothesis of language acquisition claims only that children are able to learn their
native language because human languages all reflect innate human abilities.
A beginning student of linguistics must realize that it is of crucial importance to study the structure
of sentences with a view toward describing what can and cannot be said in a language, rather than
prescribing what should and should not be said, and to approach the problem of grammatical
characterization with a view toward discovering the underlying principles upon which languages are
based. The approach in theoretical linguistics is consequently very different from the approach taken
in most language classes in the schools, including language arts classes and English classes, even
though some of the same terminology is used.
Theoretical linguists do not generally do grammatical analysis to help improve a teacher’s ability to
teach languages or a student’s ability to learn languages. Theoretical linguists deal with native
speakers who already completely know their native language. Theoretical linguists do grammatical
analysis in an effort to characterize native linguistic competence which they believe is directly
related to the nature of human beings. Thus, it is crucially important to try to find out what exactly
causes some sentences to be grammatical and others to be ungrammatical.
This approach to the study of language seeks to explain linguistic phenomena, not merely observe
and describe their existence. As we saw above in the characterization of the past tense in English,
it is possible to attain the level of explanatory adequacy by relating the distribution of [t]/[d]/[d]
in regular English verbs to principles of UG which are based on the structure and function of the
human vocal apparatus. As a result, this characterization of the past tense attains the level of
explanatory adequacy because it relates linguistic phenomena to the nature of human beings.
Further, it offers an explanation for why children learn the past tense so rapidly despite their lack of

instruction and why they systematically produce forms they never could have heard. Actually, the
characterization predicts that children will say hurted and goed. Children assume, quite naturally,
that hurt and go are regular verbs and behave exactly like all the other verbs they have acquired. It
is, therefore, not an accident that they invent these forms which they have never heard. Of course,
explanation leading to prediction is precisely what all science is about.
To summarize, the significance of the above investigation and discussion is that
we explain why children say things they never could have heard like He hurted me;
we understand why children learn their native language so rapidly and without specific
instruction;
we can link language development with neurological development and explain why learning a
language as an adult is more difficult because the brain is fully developed and innate mechanisms
are no longer operative;
we understand why all normal children, regardless of racial, ethnic, cultural, social, or economic
heritage, can learn any human language as their native language and will do so with equal facility;
we understand that no language or dialect is superior to any other language;
we understand that the differences among languages are not biologically based;
we see that, despite superficial differences, all languages and dialects share universal
characteristics;
we learn that there are limits to the ways that languages can differ from each other;
we discover that language is not linked to intelligence, that people who speak so-called “inferior”
dialects do not have deficient mental capacity.
87

88
EXERCISES FOR CHAPTER ONE
1. What must a descriptive grammar of English say about the word up in the following
sentences? Can you justify your opinions?
a. The price has been upped.
b. They can’t handle the ups and downs of life.
c. He stood up his date.
d. Bob has to be the upest guy on the team.
e. He lives up the street.
f. Let’s up the ante.
g. Sue gets up her children at six.
h. It’s up for grabs.
I. The surf is up.
j. (1) He put up the money.
(2) He put the money up.
(3) He put the money up there.
(4) He put the money up in that case.
2. All of the following sentences are ambiguous. How can you disambiguate them?
a. The thief hit the lady with the hat.
b. The repair man will look up the street.
c. John left the house messy.
d. The prisoner of war spoke foolishly.
e. John slipped on the boots in the kitchen.

f. They hired Spanish teachers.
g. They are all finished.
h. They do so love a good meal.
I. We made them idols.
j. They made idols.
3. The following are grammatical English sentences. Why are they so difficult to understand?
a. Spain was where district and provincial administrators were concerned a country of
corrupt officials.
b. The horse raced past the barn fell.
c. Where Al had had had Sue had had had had had had had had the teacher’s approval.
d. Americans Italians and Germans really like.
e. That that is is that that is not is not that it it is.
4. The following excerpts are from songs, games, magazines, and various other sources. What
makes them special? Turn them into sentences that are unremarkable.
a. They need to adore me so Christian Dior me; it’s vital you sell me, so Machiavell’ me.
b. Freedom from incrustations of grime is contiguous to rectitude.
c. Like my mother is a total space cadet, she like makes me do the dishes and clean the cat
box, I am sure, that’s like gross, barf out!
d. I have hear of you and your classes on languages for foreigner people. I am very
interesting about take some courses on your school. I have the necessity to learn English
for development very well in my profession in Mexico City. Gratefully beforehand any
information.
e. Nowadays, the only sexuality about which journalese is coy tends to be homosexuality,
and that is adequately covered by he has no close female friends or he is not about to
settle down.
89

90
5. Find the words which express the main idea in each of the following sentences. If you can,
do this by identifying the subject and the main verb.
a. When the dancer faces the audience and with one foot describes a rond de jambe à terre
en dedans so that the foot describes on the ground a semicircle from back to front, the
head should be inclined to the same side as the foot that makes the movement.
b. To be born, or at any rate bred, in a handbag, whether it had handles or not, seems to me
to display a contempt for the ordinary decencies of family life that reminds one of the
worst excesses of the French Revolution.
c. It was from these missions the Jesuit fathers carried the word of God to the high and
undiscovered plateau to those indians still existing in their natural state and received in
return martyrdom.
d. When in the Course of human Events, it becomes necessary for one People to dissolve
the Political Bands which have connected them with another, and to assume among the
Powers of the Earth, the separate and equal Station to which the Laws of Nature and of
Nature’s God entitle them, a decent Respect to the Opinions of Mankind requires that
they should declare the causes which impel them to the Separation.
e. In vertebrates, mechanisms for maintaining body temperature by adjusting heat
production and heat loss, a feat lower animals, who are at the mercy of their
environment, cannot achieve, have evolved.
f. In looking forward to the moment which is intended to terminate the career of my public
life, my feelings do not permit me to suspend the deep acknowledgment of that debt of
gratitude which I owe to my beloved country for the many honors it has conferred upon
me.
g. As we are now entering upon a book in which the course of our history will oblige us
to relate some matters of a more strange and surprising kind than any which have
hitherto occurred, it may not be amiss in the prologomenous or introductory chapter to
say something of that species of writing which is called the marvelous.
6. Why do linguists make a distinction between linguistic competence and linguistic
performance? Why is competence the focus of linguistic description? Should it be?
7. Itemize some of the differences between native language acquisition and second (foreign)
language acquisition. Consider characteristics of the learner and the circumstances of learning.

8. Here are two typical definitions from traditional grammar:
The SUBJECT of a sentence is the phrase which specifies the topic of the sentence.
The DIRECT OBJECT of a verb specifies the person or thing directly affected by the action
described in the verb.
What problems arise when these definitions are applied to the subjects and direct objects in
the following sentences?
a. John burned the money.
b. John earned the money.
c. Sue suffered a blow to the head.
d. This book reads well.
e. Bill received a promotion.
9. Here are some further definitions from traditional grammar of three parts of speech:
A VERB is a word that expresses an action, occurrence, or state of being.
A NOUN is the name of a person, place, or thing.
An ADJECTIVE is a word that describes a noun.
Using these definitions, find the verbs, nouns and adjectives in the following sentences:
a. They won on their final at–bat.
b. Do–it–yourself–ers get things done.
c. He Fred Astaired her around the ballroom.
d. The River is not your average save–the–farm film.
e. All carry–on luggage must be stowed.
f. These French fries are so ketchuped–up, they’re soggy.
91

92
Can you think of other ways of defining the parts of speech? Try to find criteria like the
following:
A VERB is a word that can carry time distinctions (play, played).
A NOUN is a word that can carry number distinctions (book, books).
An ADJECTIVE is a word that can be compared (tough, tougher, toughest).
10. Each of the following pairs of unpunctuated sentences consists of a statement and a yes/no
question, which is a question to which one can get a yes or no answer. As a native speaker of
English you know which member of each pair is the yes/no question. How do you recognize
a yes/no question? Formulate a rule for forming a yes/no question.
a. (1) will that man go
(2) that man will go
b. (1) the little girl has gone
(2) has the little girl gone
c. (1) the boy with the red hair isn’t going
(2) isn’t the boy with the red hair going
d. (1) all of them went
(2) did all of them go
e. (1) could there be more than one right answer
(2) there could be more than one right answer
Having formulated a rule for forming yes/no questions in English, consider how you might
teach a four–year–old child that rule.

CHAPTER TWO: PHONETICS AND PHONOLOGY
PHONETICS and PHONOLOGY are the two branches of grammar which are concerned with the
study of the sounds of human language. The distinction between the two fields is as follows:
phonetics deals primarily with the speech sound itself, including the way in which it is produced,
transmitted, and perceived; phonology deals more with the organization of speech sounds into sound
systems. The difference is similar to the difference between studying the materials out of which
buildings are constructed (e.g., bricks, concrete, steel, etc.) and studying the way in which a building
is constructed out of these materials (e.g., by alternating a layer of brick with a layer of concrete, by
encasing the steel in concrete, etc.). PHONETICIANS study the raw materials, the sounds;
PHONOLOGISTS study the systems formed from these sounds.
2.1 PHONETICS
PHONETICIANS describe and classify the sounds of human language in the following three ways:
(i) in terms of the way in which they are produced by the vocal apparatus; (ii) in terms of the physical
properties of the sound wave emanating from the speaker; and (iii) in terms of the effect the sound
wave has on the various parts of the ear. These three methods of description and classification are
known respectively as ARTICULATORY phonetics, ACOUSTIC phonetics, and AUDITORY
phonetics. By far, the most common of the three is articulatory phonetics; hence, in the discussions
which follow, the terminology of articulatory phonetics will be employed.
The diagram in Figure One represents a cross–section of the human head, showing the parts of the
vocal apparatus. In the production of speech sounds, the most important of these parts are the vocal
cords. These are two elastic membranes located in the larynx or Adam’s Apple.
In normal breathing, air is forced from the lungs, up the trachea, through the vocal cords, and out of
the mouth or the nose or both. If the air is not obstructed in any way by the parts of the vocal
apparatus, then no sound will be produced. Conversely, when one or more of the parts of the vocal
apparatus form an obstacle in the path of the air, then a sound results.
2.1.1 VOWELS
There are two basic types of sound segments in human speech: VOWELS and CONSONANTS.
Vowels are typically produced when the air is modified by the vibration of the vocal cords and when
the tongue is held in specific positions in the mouth. The vibration of the vocal cords is known as
VOICING. Sounds articulated with the vocal cords vibrating are [+VOICED]; those without, are
[–VOICED]. Voicing is a feature of all English vowels, whereas the position of the tongue is what
distinguishes one vowel from another. For example, the sound of a in father is produced when the
tongue is in a low position in the back of the mouth, and the sound of i in machine is produced when
the tongue is in high position in the front of the mouth. In the production of both vowels, the vocal
cords are vibrating.

94
FIGURE ONE: THE VOCAL APPARATUS
A traditional chart of some vowel sounds specified by the relative positions of the tongue in the
mouth is given in Figure Two, which is the kind of representation one is likely to find in various online
sources. The symbols in brackets are those used by many phoneticians all over the world. Each
PHONETIC SYMBOL is independent of specific languages and spelling conventions. Thus, the
symbol [i] represents the sound of i in the English word machine, of eo in the English word people,
of ie in the German word sieben, of ie in the French word vie, of i in the Spanish word hijo, etc.
The vowels in Figure Two are distinguished by three horizontal tongue positions from the front of
the mouth to the back (FRONT, CENTRAL, and BACK) and three vertical tongue heights (HIGH,
MID, and LOW). In addition, vowels produced with considerable muscular effort (TENSE) are

distinguished from those that don’t involve such effort (LAX). If these differences are considered
features, then Figure Two amounts to a claim that eight difference feature contrasts are necessary to
distinguish the twelve English vowels: [±FRONT], [±CENTRAL], [±BACK], [±HIGH], [±MID],
[±LOW], [±TENSE], and [±LAX]. That claim must be examined to see if it is justified.
FIGURE TWO: ENGLISH VOWELS
FRONT CENTRAL BACK
[i] as in beat [u] as in boot TENSE
[] as in bit [] as in put LAX
[e] as in bait [o] as in boat TENSE
[] as in bet
[] as in
butted
[] as in bought
LAX
[a] as in pot TENSE
[æ] as in bat [] as in but LAX
HIGH
MID
LOW
In the Introduction above, we talked about the importance of providing solutions to problems that
are the most cost-effective. Also, in Chapter One, we discussed the importance of attaining
descriptive adequacy, that is, in describing language with the most generalized principles possible
to cover all the data. These two matters, cost-effectiveness and descriptive adequacy, are directly
related. Employers are looking for employees who can provide them with solutions to problems that
satisfy the most factors or constituents in the most effective way (see page 5). Linguists are looking
for descriptions of language that account for all the data in the most generalized way. Given Figure
Two, the question is this: is it necessary to have eight contrasting features to describe the twelve
English vowels. Is that the most cost-effective description? The answer is, No.
The features [±HIGH], [±FRONT], etc. in Figure Two contain a number of redundancies, that is,
unnecessary duplications. For example, there seems to be no difference between [–BACK] and
[+FRONT], [+BACK] and [–FRONT], [+TENSE] and [–LAX], [–TENSE] and [+LAX], and so on.
Furthermore, it seems excessive to have three horizontal positions, three vertical positions, as well
as the features TENSE and LAX, to describe the twelve English vowels (note the six empty cells in
Figure Two). In an attempt to reach the highest possible level of descriptive adequacy, linguists seek
to remove such unnecessary redundancies and excesses from their descriptions. Therefore, it is more
common today to replace the classification in Figure Two with the one in Figure Three.
Using the terminology from the Introduction (see page 14), Figure Three is more cost-effective than
Figure Two because it uses a fewer number of features to distinguish all the vowel sounds. In Figure
Three, we have eliminated the feature contrasts [±FRONT], [±CENTRAL], and [±MID], reducing
the number of features to describe tongue position from six to three. In addition, we have added the
95

96
feature [±ROUND] to distinguish rounded vowels from unrounded vowels, specifically, the vowel
[] from the vowel [].
FIGURE THREE: DISTINCTIVE FEATURES FOR ENGLISH VOWELS
i e æ u o a
HIGH + + – – – + + – – – – –
LOW – – – – + – – – – + + –
BACK – – – – – + + + + + + +
ROUND – – – – – + + + + – – –
TENSE + – + – – + – + – + – –
There is more involved in choosing the feature inventory in Figure Three over Figure Two than just
simplicity. The number of features is, of course, important. But more important is whether or not
the feature inventory will be sufficient to describe all the data. In linguistics, that means all the
vowels in all the languages of the world. It is, therefore, an empirical issue. We must take the above
feature inventory and test it against other data in other languages to see if it is satisfactory. Notice
that is exactly what has to be done in the workplace. For example, if a manufacturer decides to make
a change in the number of options available for a product (size, shape, color, weight, etc.), the effects
of that change must be studied against all the relevant issues that could be impacted by the change
(customer satisfaction, cost, distribution, marketing, etc.). Such extensive study minimizes
unforseen consequences which can be very costly to a business. Theoretical changes work in the
same way, whether one is dealing with linguistics, psychology, physics, or any other discipline. For
just this reason, linguistic theory, like all other theories in all other disciplines, is a constantly
evolving theory in which changes are proposed and subsequently modified, extended, abandoned,
and so on. It is important for students to see that the choice of the features in Figure Three above
those in Figure Two is an entirely empirical issue that must be tested and justified; therefore, we will
return to this matter at several points below.
The technical definitions for the distinctive features of English vowels are as follows:
• HIGH – NONHIGH — High sounds are produced by raising the body of the tongue above
the level that it occupies in neutral position; nonhigh sounds are not produced with such a
raising of the body of the tongue.
• LOW – NONLOW — Low sounds are produced by lowering the body of the tongue below
the level that it occupies in neutral position; nonlow sounds are produced without such a
lowering of the body of the tongue.

• BACK – NONBACK — Back sounds are produced by retracting the body of the tongue from
the neutral position; nonback sounds are produced without such a retraction.
• ROUND – NONROUND — Rounded sounds are produced with a narrowing of the lips;
nonrounded sounds are produced without such narrowing.
• TENSE – NONTENSE (LAX) — Tense sounds are produced with a deliberate, accurate, and
maximally distinct articulatory gesture that involves considerable muscular effort; lax sounds
are produced rapidly and somewhat indistinctly.
In addition to the unitary sounds charted above in Figure Three, English contains a number of
DIPHTHONGS, which are vowels followed by a GLIDE ([w] or [y]). The true diphthongs are [y]
as in soy and soil; [ay] as in sigh, nice, guy, lie, lye, sign, aisle, and choir; and [aw] as in cow, house
and doubt. In most dialects, the [–LOW, +TENSE] vowels of English are also articulated with a
glide, particularly in word final position: [siy] see, [sey] say, [suw] sue, [sow] sew.
The glide portion of these examples is not equivalent phonetically to a [w] or [y] as these sounds
would appear in words like wet ([wt]) and yet ([yt]). Therefore, other texts may represent the
above examples as follows: [] as in soy, [a] as in sigh, [a] as in cow, [i] as in see, [e] as in
say, [u] as in sue, and [o]as in sew. For simplicity, in this book, all glides are represented as [w]
and [y], e.g., [wt] wet, [kaw] cow, [suw] sue, [sow] sew; [yt] yet, [sy] soy, [siy] see, [sey] say.
2.1.2 CONSONANTS
Consonants are produced in a variety of different ways. They are classified in terms of the place in
the oral cavity (Figure One) where they are articulated and also the manner of their articulation. A
chart of some consonants specified according to this method of classification, as well as the features
[±CORONAL] and [±ANTERIOR] discussed in Chapter One, is given in Figure Four.
2.1.2.1 STOPS
Some consonants involve the complete stoppage of the air flow at some point in the vocal tract.
h
Such consonants are known appropriately as STOPS. These include [p ] (the sound of p in pan),
h h
[t ] (the sound of t in tan), and [k ] (the sound of c in can and con and k in kin and Ken).
The point in the vocal tract at which the stoppage of air occurs is used to distinguish one stop
h
consonant from another. [p ], which is produced by stopping the air with both lips, is called a
h
BILABIAL stop; [t ], which is produced by placing the tongue against the alveolar ridge (see Figure
h
One), is called an ALVEOLAR stop; and [k ], which is formed with the tongue against the velum,
is called a VELAR stop.
97

98
MANNER OF
ARTICULATION
STOP
[–VOICED]
[–VOICED]
[+VOICED]
FRICATIVE
[–VOICED]
[+VOICED]
AFFRICATE
[–VOICED]
[+VOICED]
NASAL
FIGURE FOUR: ENGLISH CONSONANTS, LIQUIDS, AND GLIDES
PLACE OF ARTICULATION
ANTERIOR NONANTERIOR
NONCORONAL CORONAL NONCORONAL
BILABIAL
h p (pun)
p (spun)
b (bun)
LABIO–
DENTAL
f (fie)
v (vie)
INTER–
DENTAL
(thigh)
ð (thy)
ALVEOLAR
h t (ton)
t (stun)
d (done)
s (sue)
z (zoo)
(ALVEO)
PALATAL
š (shoe)
ž (beige)
(chew)
(Jew)
VELAR GLOTTAL
h k (come)
k (scum)
g (gum)
[+VOICED] m (sum) n (sun) (sung)
LIQUID
[+VOICED]
[+VOICED]
GLIDE
[–VOICED]
[+VOICED]
l (lay)
(cater)
(whey)
w (way) y (yea)
r (ray) (ail)
(cotton)
h (hay)

The superscript “h” in the transcription of stops indicates a puff of air called ASPIRATION that is
emitted when the stops are released; this puff of air is missing when these stops follow [s]. The
distinction is apparent by placing a tissues in front of the mouth and noticing first that the tissue
h h h
moves when saying pan [p æn], tan [t æn], and can [k æn] and second that the tissue does not move
when saying span [spæn], Stan [stæn], and scan [scæn]. Compare similar pairs like pit/spit, pin/spin,
tall/stall, cool/school, etc.
h h h
In the production of [p ], [p], [t ], [t], [k ], and [k], the vocal cords do not vibrate; hence, these
sounds are called VOICELESS ([–VOICED]) stops. Other stops, for example, [b] as in bet, [d] as
in debt, and [g] as in get, are formed by a complete stoppage of the air flow with an accompanying
vibration of the vocal cords. These stops are called VOICED ([+VOICED]). The symbol []
represents a glottal stop that appears in a few words like button [bn] and cotton [kan] or, in
some dialects, a word like bottle [bal].
2.1.2.2 FRICATIVES
A second group of consonants, called FRICATIVES, is formed by a closure in the vocal tract,
which, though not complete as in the articulation of stops, is sufficiently constricted to cause
turbulence in the air flow, thereby producing a hissing sound. This group includes the following
sounds: [f] (wafer), [v] (waiver), [] (ether), [ð] (either), [s] (racer), [z] (razor), [š] (mesher), an
[ž] (measure).
Like stops, fricatives are classified according to the point in the vocal tract at which the obstruction
in the air flow occurs: [f] and [v] are produced by a constriction formed with the lower lip and upper
teeth; they are called LABIODENTAL fricatives; [] and [ð], produced by a constriction behind
the upper teeth, are INTERDENTAL; [s] and [z], produced by a constriction against the alveolar
ridge, are ALVEOLAR; [š] and [ž], produced by a constriction against the palate, are PALATAL.
The sound [ž] occurs initially only in the word genre which was borrowed from French. Though
relatively uncommon in English, [ž] does appear medially (vision, measure, usual, Asian, etc.) and
finally (beige, massage, camouflage, etc.).
2.1.2.3 AFFRICATES
Closely allied to the articulation of stops and fricatives is a group of consonants called
AFFRICATES. These consonants combine a complete closure at some point in the vocal tract, as
in the articulation of stops, with a turbulent release that produces a hissing sound, as in the
articulation of fricatives. Examples of these consonants are [], the underlined sounds in chunk,
lecher, and search, and [], the underlined sounds in junk, ledger, and surge. These are the only
affricates in English, but there are several more in the other languages of the world.
99

100
2.1.2.4 NASALS
A fourth group of consonants, called NASALS, are produced by a closure in the mouth and a
simultaneous lowering of the velum so that the air escapes through the nasal passage (see Figure
One). English has three nasal consonants. They are distinguished from each other by the point in
the mouth where the closure is made. [m], the sound of m in Kim, is a bilabial nasal; [n], the sound
of n in kin, is an alveolar nasal; and [], the sound of ng in king is a velar nasal. Each of the English
nasals is a voiced consonant. The velar nasal [] never occurs initially in a word or syllable.
2.1.3 LIQUIDS AND GLIDES
In addition to consonants and vowels, languages contain two other classes of sounds, LIQUIDS and
GLIDES.
2.1.3.1 LIQUIDS
LIQUIDS include [l], the sound of l in low, and [r], the sound of r in row. [l] is produced with a
closure made by the center of the tongue at the alveolar ridge, in such a way that the air is allowed
to pass out of the mouth along the sides of the tongue. Since the air passes along the sides of the
tongue, l–sounds are often referred to as LATERAL consonants. In contrast to the articulation of
[l], [r] usually involves a closure, or near–closure, made with the sides of the tongue, so that the air
escapes from the mouth over the central portion of the tongue. A liquid sound can also be produced
by making the tongue FLAP against the alveolar ridge. This sound, represented as [] or [D], is
h
heard instead of [t ] or [d] in words like water and ladder.
2.1.3.2 GLIDES
The last group of sounds includes [y], the sound of y in yell, and [w], the sound of w in well. These
consonants are either called GLIDES or, because of their similarity to the vowels [i] and [u],
SEMIVOWELS. Typically, they are produced like vowels since they do not involve a complete
stoppage of the air flow, but they are distributed like consonants since they occur before and after
vowels, as in the word wow.
The four groups of sounds (vowels, consonants, liquids and glides) are distinguished by the
following distinctive features:
FIGURE FIVE: THE MAJOR PHONOLOGICAL CATEGORIES
FEATURES CONSONANTS VOWELS GLIDES LIQUIDS
CONSONANTAL + – – +
VOCALIC – + – +

The definitions for the remaining distinctive features are as follows:
• SONORANT – NONSONORANT (OBSTRUENT) — Sonorants are sounds produced with
a vocal tract cavity configuration in which spontaneous voicing is possible; obstruents are
produced with a cavity configuration that makes spontaneous voicing impossible. [Note:
sonorants are usually hummable.]
• CONSONANTAL – NONCONSONANTAL — Consonantal sounds are produced with a
sustained vocal tract constriction at least equal to that required to produce fricatives;
nonconsonantal sounds are produced without such a constriction.
• VOCALIC – NONVOCALIC — Vocalic sounds are produced with an oral cavity in which
the most radical constriction does not exceed that found in the high vowels [i] and [u], and
with the vocal cords positioned so as to allow spontaneous voicing; in producing nonvocalic
sounds one or both of these conditions are not satisfied.
• CONTINUANT – NONCONTINUANT (STOP) — In the production of continuant sounds,
the primary constriction in the vocal tract is not narrowed to the point where the air flow past
the constriction is blocked; in stops the air flow through the mouth is effectively blocked.
• NASAL – NONNASAL — Nasal sounds are produced with a lowered velum which allows
the air to escape through the nose; nonnasal sounds are produced with a raised velum so that
the air from the lungs can escape only through the mouth.
• SIBILANT – NONSIBILANT — Sibilant sounds are produced by forcing the air through a
narrow opening produced by a groove in the midline of the tongue. Typically, sibilant sounds
have a hissing quality; nonsibilant sounds do not have this quality.
• CORONAL – NONCORONAL — Coronal sounds are produced with the blade of the
tongue (the portion immediately behind the tip) raised from its neutral position; noncoronal
sounds are produced with some other articulator than the blade of the tongue.
• ANTERIOR – NONANTERIOR — Anterior sounds are produced with an obstruction that
is located in front of the palato–alveolar region of the mouth; nonanterior sounds are produced
without such an obstruction.
• VOICED – NONVOICED (VOICELESS)— Voiced sounds are produced with the vocal
cords vibrating; voiceless sounds are produced without such vibration.
Figure Six categorizes English consonants, liquids, and glides using these nine distinctive features;
it also includes the four relevant features discussed above with vowels. Again, in terms of costeffectiveness,
Figure Six is preferred to Figure Four because it removes redundancies and
contradictions and, at the same time, appears to be adequate to describe the data investigated so far.
101

102
FIGURE SIX: DISTINCTIVE FEATURES FOR ENGLISH CONSONANTS, LIQUIDS & GLIDES
FEATURES
STOPS NASALS FRICATIVES AF* LIQUIDS GLIDES
p b t d k g m n f v ð s z š ž è l ³ r w y h
SONORANT – – – – – – – + + + – – – – – – – – – – + + + + + + + +
CONSONANTAL + + + + + + + + + + + + + + + + + + + + + + + + – – – –
VOCALIC – – – – – – – – – – – – – – – – – – – – + + + + – – – –
CONTINUANT – – – – – – – – – – + + + + + + + + – – + + – + + + + +
NASAL – – – – – – – + + + – – – – – – – – – – – – – – – – – –
SIBILANT – – – – – – – – – – – – – – + + + + + + – – – – – – – –
ANTERIOR + + + + – – – + + – + + + + + + – – – – + – + – + + – –
CORONAL – – + + – – – – + – – – + + + + + + + + + – + + – – + –
VOICED – + – + – + – + + + – + – + – + – + – + + + + + – + + –
HIGH – – – – + + – – – + – – – – – – + + + + – – – – + + + –
LOW – – – – – – + – – – – – – – – – – – – – – – – – – – – +
BACK – – – – + + + – – + – – – – – – – – – – – – – – + + – +
ROUND – – – – – – – – – – – – – – – – – – – – – – – – + + – –
*AFFRICATES

2.1.4 REVIEW OF ISSUES
At this point, it is important to examine the choices we have made in the feature inventories above.
For vowels, we noted that Figure Two involved eight feature oppositions: [±FRONT],
[±CENTRAL], [±BACK], [±HIGH], [±MID], [±LOW], [±TENSE], and [±LAX]. Figure Three
involved five: [±HIGH], [±LOW], [±BACK], [±TENSE], and [±ROUND]. We determined that the
reduced number of features in Figure Three was preferable. The argument was largely about
simplicity, but other factors are also relevant. For example, the feature opposition [±ROUND] must
be part of the inventory to account for phenomena that exist in the world’s languages. There is no
way to know this ahead of time; one must test the hypothesis by examining new data. In Turkish,
for example, an unrounded vowel in a word cannot be followed by a rounded vowel. The
phenomenon, called VOWEL HARMONY, requires that only certain combinations of vowels in
a word are possible. Clearly, we cannot account for the data in Turkish without the feature
[±ROUND]. Thus, quite apart from issues of simplicity, the empirical evidence demands that our
feature inventory includes the opposition [±ROUND]. Even in English the feature is important:
only vowels that are [+BACK, –LOW] are [+ROUND] ([u, , o, ]).
The chart of consonants in Figure Four specifies two sets of features to indicate place of articulation.
One set includes only the two oppositions [±ANTERIOR] and [±CORONAL]. The other includes
seven oppositions: [±BILABIAL], [±LABIODENTAL], [±INTERDENTAL], [±ALVEOLAR],
[±PALATAL], [±VELAR], and [±GLOTTAL]. On the basis of simplicity alone, we would naturally
choose the set with two as opposed to the one with seven. But again, this is an empirical issue. In
our discussion of levels of adequacy in Chapter One (page 51 ff.), we saw that the data favor the first
set with just [±ANTERIOR] and [±CORONAL]. We observed that the prefix meaning ‘not’ is
pronounced in three ways in English ([m] as in impossible, [n] as in intangible, and [] as in
incurable):
(1) a. [m] before sounds that are [+ANTERIOR, –CORONAL], namely, [p, b, m]
b. [n] before sounds that are [+ANTERIOR, +CORONAL], namely, [t, d, n]
c. [] before sounds that are [–ANTERIOR, –CORONAL], namely, [k, g]
We also saw that the distribution is not arbitrary. Sounds assimilate to each other, that is, they tend
to become like each other when they are adjacent. The rules in (1) account for the occurrence of the
bilabial nasal [m] before bilabial sounds, of the alveolar nasal [n] before alveolar sounds, and of the
velar nasal [] before velar sounds. The distribution is succinctly accounted for with the features
[±ANTERIOR] and [±CORONAL]. In Latin, the roots of nouns undergo different rules depending
on whether they end in a consonant that is [+CORONAL] ([t] and [d]) or [–CORONAL] ([p], [b],
[k], and [g]). In many places below, we will find additional empirical support for the two feature
oppositions [±ANTERIOR] and [±CORONAL].
The chart of consonants in Figure Four involves a set of six feature oppositions for a manner of
articulation: [±STOP], [±FRICATIVE], [±AFFRICATE], [±NASAL], [±LIQUID], and [±GLIDE].
The chart of consonants in Figure Six also involves a set of six feature oppositions for manner of
103

104
articulation, but they are different oppositions: [±SONORANT], [±CONSONANTAL],
[±VOCALIC], [±CONTINUANT], [±NASAL], and [±SIBILANT]. In this case, sheer number is
not the issue. So, we need to see which set is more empirically justified. The data indicate that the
second set is. First, notice that the second set includes the oppositions [±CONSONANTAL] and
[±VOCALIC], which allowed us to generalize the differences between the four major types of
sounds in natural languages, namely, vowels, consonants, liquids and glides (see Figure Five).
Second, the inventory in Figure Six also includes the opposition [±SIBILANT]. This opposition
operates in a number of phonological processes in English. For example, the plural of nouns is
indicated by the suffix [z] only in words that end with a sound that is [+SIBILANT], i. e., [s, z, š,
ž, , ] as in bases, mazes, sashes, garages, arches, judges, etc. The data indicate that [±SIBILANT]
must be included in the inventory. Thus, in terms of both simplicity and descriptive adequacy, we
have elected to use the feature oppositions in Figure Three for vowels and Figure Six for consonants,
liquids, and glides. We have eliminated redundancies and excesses and included additional feature
oppositions that are empirically necessary like [±ROUND] and [±SIBILANT]. At the same time,
the features we have chosen seem to be the minimum set required to account for all the data.
As we have seen, the kinds of considerations made in linguistics must also be made in many other
areas. Let us consider an example from the retail business. It is generally thought that a store must
provide customers with several different varieties of a product in several different sizes and shapes
in order to satisfy the needs of as many different customers as possible. Thus, in the average
supermarket, one finds, for example, many different brands of mayonnaise in containers of many
different sizes and shapes. Jim Senegal, the founder of Costco, had a very different marketing
strategy. He reasoned that cutting down on the number of choices would help customers make their
selection more easily, and, in the end, they would purchase the items Costco sells. The result is that
Costco only offers about 4,000 different items compared to the 16,000-150,000 items stocked by
many competitors. Walmart offers over 100,000 different items. Senegal’s strategy has worked.
“In the last 27 years Costco has become the fourth largest retailer in the United States and the eight
largest in the world.” (Costco Case Study, 2011; http://themarketplaceoflife.blogspot.com/2011/01/)
Senegal’s strategy was risky since is went against the conventional wisdom in retail about the
number of choices customers want. Note that there are redundancies and excesses when a
supermarket stocks many different brands of mayonnaise in containers of many different sizes and
shapes. The questions are: If the supermarket gets rid of the redundancies (different brands with
exactly the same ingredients) and the excesses (multiple shapes, sizes, and quantities), will the
resulting selection satisfy as many customers? Will customers shop elsewhere? Or will customers
simply make a choice from the selections they are given? Several studies have shown that customers
report greater satisfaction when their original options are limited. The Costco business model is a
case in point. It includes turnover on a very limited number of selections for each type of product,
e.g., only one size and brand of mayonnaise. That strategy combined with the addition of other
innovations – volume purchasing, efficient distribution practices, and reduced handling of
merchandise – have helped to make Costco so successful. It is not an accident that Costco carries
only a very limited number of items.

For many students, the study of language in a linguistics course appears to involve exotic issues that
have no relationship to events in the real world. As we have seen, this is not the case. The questions
we have considered in this Chapters involve redundancy, consistency, adequacy, generality,
efficiency, and cost, which are exactly the same questions that arise in real world problems such as
the selection of brands and packaging of mayonnaise. When students can see this larger picture, they
can also see that only the data are different. In the Introduction, we noted that it is exactly the larger
picture – the ability to make connections at an abstract level between seemingly unrelated matters
– that is crucial to adaptability, to flexibility, to mobility, and to success.
2.1.5 SUPRASEGMENTALS
There are, of course, many more sounds in the world’s languages than the ones that have been
described above. In addition, human speech is characterized by several SUPRASEGMENTAL
features, that is, vocal modification executed along with the pronunciation of vowels and consonants.
Two of these suprasegmental features are PITCH, i.e., the relative tone at which individual sound
segments or groups of segments are produced, and STRESS, i.e., the relative emphasis (or accent)
given to individual sound segments or groups of segments. Features like pitch and stress are as
important in speech as vowels and consonants. Often, they serve to distinguish words from each
other. For example, permit must be pronounced with heavy stress on the first syllable [p´rmt] if
it is used as a noun, but with heavy stress on the second syllable [prm´t] if it is used as a verb.
Notice that this variation of stress is an example of how the sound system of a language is directly
related to the classes of forms in a language. It is impossible to describe this variation without
reference to whether permit is used as a noun or as a verb. There are many similar cases in English,
e.g., contract, pervert, present, and so on. Consider also pairs of examples like the following:
(2) a. Look at the black bird. (white house, long horn, light weight, light year, etc.)
b. Look at the blackbird. (White House, longhorn, lightweight, light-year, etc.)
One cannot describe the stress given to sequences like black+bird, unless one considers whether the
sequence is used as a compound noun or as an adjective–noun construction. Examples such as these
attest to the interrelationship of sound and form in the grammar of a language.
2.1.6 SYLLABLES
There are as many syllables in a word as there are vowel sounds. Note the following paying attention
to sound, not spelling:
(3) a. One syllable: I my mine strengths
b. Two syllables: enough pygmy mining strengthens
c. Three syllables: Michigan Missouri Tennessee strengthening
105

106
As these examples show, a syllable must contain a vowel. It also can begin and end with no
consonants or one or more consonants. Using the symbol “C” for consonant and “V” for vowel, the
following are examples of one syllable words that begin and end with one consonant, a CVC pattern:
(4) CVC: pin seam gain suit zoom beaut
We could simply repeat Cs to indicate multiple consonants, but since the number of successive
consonants can be fairly large (cf. strengths [strgs]), linguists often use a superscript to indicate
the most number of consonants and subscript to indicate the least number of consonants. In the
following notations, the superscript on the symbol “C” indicates the maximum number of consonants
(n=any number greater than 1); the subscript indicates the minimum number (0=none).
(5) a. All one syllable words: # C V C #
b. All two syllable words: # C V C V C #
c. All three syllable words: # C V C V C V C #
n n
n
0
n
0
n
n
0
n
0
n
0
n
0 0 0 0
Here are some examples of one, two, and three syllable words using this notation:
(6) a. All one syllable words:
I:
# n C0 V
a
n C 0
y
#
I’m: a ym
my: m a y
mine: m a yn
streams: str i mz
strengths: str gs
Flint: fl nt
b. All two syllable words:
enough
n # C0 V
i
n C0 n
V
n C 0
f
#
pygmy: p gm i
mining: m a yn
strengthens: str g nz
Utah: y u ta
Milford: m lf rd
c. All three syllable words:
Michigan:
n # C0 m
V
n C0 š
V
n C0 g
V
n C 0
n
#
Missouri: m z u r i
Tennessee: t n s i
instructive: nstr kt v
strengthening: str g n
fiasco: f i y æ sk o

The last syllable is called the ULTIMA; the next to last syllable, the PENULT; the third to last
syllable ANTEPENULT. Words with accents on each of these syllables are given in the following
examples where the X indicates the position of the accent; the exact phonetic value for unaccented
vowels may vary from speaker to speaker ([], [], [], etc.):
(7) a. Accent on the ANTEPENULT: X
Michigan: m ´ š g n
buffalo: b ´ f l o
asterisks: æ´ st r sks
America: m ´ r k
asparagus: sp æ´ r g s
b. Accent on the PENULT: X
pygmy: p ´ gm i
Utah: y ú t a
Missouri: m z ú r i
aroma: r ó m
verenda: v r æ´ nd
chianti: k i y á nt i
c. Accent on the ULTIMA: X
streams: str í mz
enough: i n ´ f
Tennessee: t n s í
Illinois: l n ´y
For practice, there are additional examples of stress placement in English words on pages 133 and
134. The answers to those practice examples are on pages 144 and 145.
2.2 PHONOLOGY
With a complete phonetic system for the identification and classification of speech sounds available,
phonologists can begin their work. Essentially, their task is to discover the ways in which the sounds
of language are systematized. This includes finding out (i) which sounds, out of all the sounds a
speaker of some language utters, are the linguistically significant sounds for that language, and (ii)
which rules govern the organization and distribution of these sounds with respect to each other.
In analyzing English, for example, phonologists observe that all English speakers pronounce p in
h
several different ways. As we have seen, the sound [p ] is followed by a perceptible puff of air,
called ASPIRATION; this occurs in the articulation of a word like pin. In a word like spin, on the
other hand, the p is not aspirated. As before, we use a superscript “h” for aspiration: the aspirated,
h
voiceless, bilabial stop in English is [p ], and the unaspirated, voiceless, bilabial stop is [p].
107

108
In examining English, phonologists also observe that there are no pairs of words which are
distinguished by aspiration, in the way that there are many pairs which are distinguished by other
h
phonetic features, such as [–VOICED] ([p ]) and [+VOICED] ([b]), e.g., plank and blank, tap and
tab, rapid and rabid, and so on. Lastly, they observe that [p] occurs only after the sound [s] and that
h [p ] occurs everywhere else. Thus, in spot, spin, spoke, and Spain, the p is unaspirated, but in pot,
pin, poke, and pain, the p is aspirated.
h
These observations are important because they reveal that the distinction between [p] and [p ], i.e.,
unaspirated versus aspirated, is not a significant distinction in the phonology of English, unlike the
h
distinction between [p ] and [b], i.e., voiceless versus voiced. Thus, while it is possible to predict
h h
when [p], as opposed to [p ], will occur in English words, it is not possible to predict when [p ], as
opposed to [b], will occur. This distinction is crucial in phonological analysis.
Phonologists often call the distinctive classes of sounds in a language the PHONEMES of that
language, and refer to the positional variants of phonetically similar sounds as the ALLOPHONES
of a particular phoneme. To distinguish phonemes from allophones, the former are placed in slanted
h
lines, e.g., /p/ and /b/, and the latter retain their phonetic symbolization, e.g., [p] and [p ]. The
relationship between a phoneme and its allophonic variants is expressed in the form of a rule
h
included in the grammar of a language. For example, the relationship between /p/, [p], and [p ] in
English is expressed in two rules as follows:
(8) a. The phoneme /p/ is realized phonetically as its allophone [p] when it occurs in the
environment after [s], e.g., in the word spot.
h
b. The phoneme /p/ is realized phonetically as its allophone [p ] in most other
environments, e.g., in the word pot.
It is important to realize that sounds which are phonemic, i.e., distinctive, in one language may be
h
allophonic, i.e., nondistinctive, in another language. For example, while [p ] and [p] are allophones
of the same phoneme /p/ in English, they are two different phonemes in Chinese. To put it
h
differently, while [p ] and [p] do not distinguish pairs of words in English, they do distinguish pairs
h
of words in Chinese. The Chinese word pai, if pronounced with a [p ], means ‘branch,’ and, if
pronounced with a [p], means ‘weeds.’ It is very difficult for English speakers to master this
distinction when they attempt to learn Chinese; very often, they confuse the two sounds. Conversely,
speakers of many Asian languages have great difficulty in distinguishing the phonemes /l/ and /r/ in
English, because these two sounds are not phonemic in their native language. Without considerable
practice, a native speaker of Japanese, for example, might say rorripop when she means lollipop.
When English speakers hear rorripop, they find it hard to believe that the Japanese speaker can’t
easily hear the difference between what she’s said and lollipop. Similarly, a Chinese speaker will
find it hard to believe that an American cannot easily distinguish between the distinct sounds [p ]
h
and [p].

Continuing with their analysis of English, phonologists observe that the other voiceless stop
phonemes in English, namely, /t/ and /k/, show the same allophonic variations as /p/. Thus, in stop,
stub, skin, school, the t and k sounds are unaspirated, but in top, tub, kin, and cool, they are aspirated.
This is not unexpected, since /p/, /t/, and /k/ form a NATURAL CLASS of sounds, which is a class
that shares the same set of distinctive features. In this case, the three sounds form the natural class
of [+CONSONANT, –CONTINUANT, –NASAL, –VOICED]. Furthermore, the same natural class
has another allophonic variation. In final position or at the end of a syllable, they become
unreleased. Notice that the final sounds in cap, pat, and tack are not fully exploded in the way the
initial sounds of the same words are; similarly, the p sound at the end of the first syllable in captain
and the k sound at the end of the second syllable in refracted are also not fully exploded. We will
h
use the symbol "°" to indicate lack of release. For example, we have [kæp°] (cap) and [kæp°t v]
(captive). The environment in which "°" occurs is predictable by rule: voiceless stops are unreleased
in final position of a syllable or word. Since the feature RELEASED is predictable in English, it
is not distinctive.
In Chapter One, we discussed the classification of several distinctive features for English, such as
[±NASAL], [±VOICED], [±ANTERIOR], and [±CORONAL]. We now have two nondistinctive
features in English: [± ASPIRATION] and [±RELEASE]. Summarizing, we have contrasts between
distinctive and nondistinctive features like the following:
(9) DISTINCTIVE FEATURES.
a. [± VOICED]
[1] [–VOICED] [p], [t], [k]
[2] [+VOICED] [b], [d], [g]
b. [± ANTERIOR]
[1] [–ANTERIOR] [k], [g]
[2] [+ANTERIOR] [p], [b], [t], [d]
c. [± CORONAL]
[1] [–CORONAL] [p], [b], [k], [g]
[2] [+CORONAL] [t], [d]
(10) NONDISTINCTIVE FEATURES.
a. [± ASPIRATED]
[1] [–ASPIRATED] little breathiness
[a] [p] spade [sped]
[b] [t] steam [stim]
[c] [k] school [skul]
109

110
[2] [+ASPIRATED] considerable breathiness
b. [± RELEASED]
[a] h [p ] paid h [p ed]
[b] h [t ] teem h [t im]
[c] h [k ] cool h [k ul]
[1] [–RELEASED] no explosion
[a] [p°] mop [map°]
[b] [t°] rate [ret°]
[c] [k°] nook [nk°]
[2] [+RELEASED] explosion
[a]
h
[p ] paid
h
[p ed]
[b]
h
[t ] teem
h
[t im]
[c] h [k ] cool h [k ul]
Phonologists discover the distinctive features of a language by looking for a pair of words
distinguished by one and only one sound. Such a pair is called a MINIMAL PAIR. If two sounds
occur in a minimal pair, then they occur in identical environments. This means that the two sounds
are distinctive and distinguish words. (11) contains examples of minimal pairs; (12) does not.
(11) MINIMAL PAIRS Good Examples (SPELLING IRRELEVANT):
a. mace/base [mes]/[bes]
b. ether/either [ir]/[iðr]
c. known/sewn/shone [non]/[son]/[šon]
h
d. phlegm/Clem [flm]/[k lm]
e. who’d/hood [hud]/hd]
f. bird/heard [brd]/[hrd]
(12) Not Examples (SPELLING IRRELEVANT):
a. buff/bluff [bf]/[blf]
b. choose/loose [uz]/[lus]
c. though/tough [ðo]/[tf]
d. chef/chief [šf]/[if]
e. great/heart [gret]/[hart]
f. beard/heard [bird]/[hrd]

As we have had occasion to notice many times in this book, an important objective of linguistics is
the construction of formal grammars. A formal grammar is simply one that is perfectly explicit and
testable. Precision is essential in linguistics, as it is in all other sciences. Without precision,
hypothetical principles and rules cannot be evaluated and tested with confidence. If hypotheses
cannot be tested, then substantive conclusions cannot be drawn. Rules such as those in (8) are too
informally written to satisfy scientific criteria. Therefore, linguists usually replace them with a
SLASH–DASH notation as follows:
FORMAL PHONOLOGICAL RULES:
(13) a. [+STOP, –VOICED] [–ASPIRATED] / [s] ___
a voiceless stop is unaspirated after [s]
111
b. [+STOP, –VOICED] [–RELEASED] / {+, #}
a voiceless stop is unreleased before a syllable boundary or a
word boundary
Each symbol in this notation has a precise meaning. As a result the rules are explicit and testable.
The slash–dash notation, which will be used throughout this text, observes the following
conventions:
THE SLASH–DASH NOTATION:
(14) a. means ‘is’
b. (...) items in parentheses are optional
c. {...} items in curly brackets or braces are alternatives
d. / means ‘in the environment of’
e. ___ means ‘before’ or ‘after’ or ‘between’
f. / ___x means ‘before x’
g. / x___ means ‘after x’
h. / x___y means ‘between x and y’
i. # means ‘word boundary’
j. #___ means ‘after a word boundary,’ that is, ‘at beginning of word’
k. ___# means ‘before a word boundary,’ that is, ‘at end of word’
l. + means ‘syllable boundary’ or ‘morpheme boundary’
m. +___ means ‘after a syllable or morpheme boundary’
n. ___+ means ‘before a syllable or morpheme boundary’
Summarizing the theoretical apparatus developed so far, we say that phonological representations
occur in pairs: a phonemic or abstract representation coupled with a phonetic or real representation.
The two representations are related by phonological rules.

112
The necessity for both levels of representation can be illustrated further in the following way (see
Figure Seven below). Suppose one English speaker utters the sentence If you hit me, I’ll hit you.
What the speaker actually says in the second part of this sentence is [ayl hu]. The verb, therefore,
appears to be hitch [h]. But, although the listener has heard [h], what the listener understands
is /ht/ (hit). The real or superficial or phonetic representation, the one actually spoken and heard,
is [h]; the abstract or underlying or phonemic representation, the one intended and understood, is
/ht/. The relationship between the two is accounted for by phonological rules like those in (13),
specifically, (15a) and (15b).
(15) a. [t] [] / ___ [y]
b. [y] Ø / []___
(15a) changes a [t] to a [] before [y]; (15b) deletes a [y] after []. The symbol "Ø" means ‘zero’;
so (15b) says that a [y] becomes zero (is deleted) after []. These rules represent the change of the
sound [t] to []. The derivation from underlying form to surface form proceeds as follows:
(16) UNDERLYING FORM: /ht yu/
APPLICATION OF RULE (15a): h yu
APPLICATION OF RULE (15b): h u
SUPERFICIAL FORM: [hu]
This derivation is intended to represent the native speaker’s knowledge (competence) of the intended
meaning of what is actually uttered, namely, [hu].
To an introductory linguistics student, a derivation like (16) often appears to be the result of
hocus–pocus. This is not the case. The derivation proceeds from facts: in the example cited, there
is a discrepancy between what is uttered and heard ([hu]), on the one hand, and what is intended
and understood (/ht/), on the other. Further, all native speakers of English know this unconsciously;
it is part of their linguistic competence. If this were not the case, the person hearing [hu] in the
context given, would not understand the intended meaning. These facts are givens; there is nothing
that a linguist can do but attempt to describe and explain them. It is not the job of linguistics to alter
the way people communicate (a hopeless task, incidentally). The postulation of an abstract
underlying form, related by rule to a real superficial form, is merely an effort to describe observed
phenomena. Again, linguistic theory proceeds in the appropriately scientific manner. Notice that
the example here is not isolated; in ordinary speech, the spoken form rarely matches the intended
meaning word for word:
(17) Spoken Form Intended Meaning Written Form
a. [uwango] /du yu want tu go/ Do you want to go?
b. [iydn] /it yr dnr/ Eat your dinner.
c. [lvmvlft] /l f ðm hæv lft/ All of them have left.
d. [kwigwawt] /kd wi go awt/ Could we go out?

FIGURE SEVEN: PHONOLOGICAL AND PHONETIC REPRESENTATIONS
PHONEME: ALLOPHONE:
Abstract Real
Distinctive Features
[±VOICED]
[±ANTERIOR]
Nondistinctive Features
[±ASPIRATED]
[±RELEASED]
Not predictable by rule Predictable by rule
Nonredundant Redundant
Found in minimal pairs Not found in minimal pairs
Found in overlapping distribution
(in the same environments)
Found in complementary distribution
(not in the same environments)
Underlying (deep) Superficial (surface)
Slanted line notation
/spat/ spot
Square bracket notation
[spat°] spot
EXAMPLES: EXAMPLES:
Will he hit me?
/wl hi ht mi/
Will he hit you?
/wl hi ht yu/
Willie hit you?
/wli ht yu/
Will he hitch you up?
/wl hi h yu p/
Willie hitched you up?
/wli h+d yu p/
Will he hit me?
[wli ht° mi]
Will he hit you?
[wili hu]
Willie hit you?
[wli hu]
Will he hitch you up?
[wli hu p°]
Willie hitched you up?
[wli ht°u p°]
113

114
2.3 PHONOTACTICS
Further investigation of English phonology reveals that there are general restrictions on the
sequences of sounds that are permissible in the language. There are ACCIDENTAL GAPS,
sequences of sounds that just don’t happen to occur, and, NONACCIDENTAL GAPS, sequences
of sounds that are not possible. The permissible sequences are specified in the PHONOTACTIC
RULES of the language, that is the rules concerned with the sequential arrangements of sounds in
larger units. Consider the following:
(18) ACCIDENTAL GAP (possible words): [bld], [stæno], [sprim]
(19) NONACCIDENTAL GAP (impossible words): [ftd], [zdæno], [spnim]
In analyzing the PHONOTACTICS of English, phonologists would observe that an English word
cannot end with two nasal consonants even if it is spelled with two. We see this in a root such as
/kalmn/ which loses the second nasal in the word column [kalm], but retains it before the suffix
–ist in the word columnist [kalmnst]. Compare also autumn/autumnal and hymn/hymnal. The loss
of the second nasal at the end of a word is accounted for by the following phonotactic rule:
(20) [+NASAL] Ø / [+NASAL]___#
In addition to (20), the phonotactic rules of English do not permit other consonants to occur before
nasals at the beginning of a word. Notice that both consonants are pronounced only when they occur
in separate syllables:
(21) a. mnemonic [nmánk] ‘helping the memory’
amnesia [æmn´ža] ‘loss of memory’
b. pneumonia [nmónya] ‘respiratory disease’
apnea [æ´ pniya] ‘temporary loss of breathing’
c. know [nó] ‘grasp in the mind’
acknowledge [æknál] ‘recognize’
The data in (21) indicate that English also has the following rule:
(22) [+CONSONANT] Ø / #____[+NASAL]
The presence of rules like (22) is what makes some words retaining their foreign spelling look so
strange as compared to their pronunciation, e.g., mnemonic, pneumonia, and pterodactyl, which
come to English from Ancient Greek. Still, the deleted consonant must be considered part of the root
so that the words can be related. Thus, the root for know must be /kno/ in order to relate know and
acknowledge. As before, we must appeal to an abstract level of representation to do so.

In addition to specific phonotactic rules such as those above involving nasals, there are more
generalized phonotactic rules. For example, if an English word begins with three consonants (CCC),
the first one must be [s], the second one must be a voiceless stop ([p], [t], or [k]), and the third one
must be a liquid ([l] or [r]).
(23) a. spl- splash [splæš] spr- spread [sprd]
b. stl- * str- stream [strim]
c. skl- sclaff [sklæf] skr- scream [skrim]
Notice that, so far, there are no English words that begin with [stl]. Given the other possibilities, that
gap should probably be considered an accidental gap.
2.4 MORPHOLOGY
As one might expect, the study of the sound system of a language cannot be carried out in complete
isolation from the other components of grammar. The phonological rules, in particular, very often
involve reference to MORPHEMES, which are the minimal units of meaning in a language.
Consider the following (ignoring details involving aspiration and release):
(24) MORPHEME: a minimum unit of distinctive meaning.
a. [spikr] ‘speaker’; 2 syllables; 2 morphemes:
/spik/ ‘to utter words’
/r/ ‘agent’
b. [kriyetr] ‘creator’; 3 syllables; 2 morphemes.
/kriyet/ ‘to bring into existence’
/r/ ‘agent’
c. [dalr] ‘dollar’; 2 syllables; 1 morpheme.
d. [æntlr] ‘antler’; 2 syllables; 1 morpheme
e. [bgr] ‘bigger’; 2 syllables; 2 morphemes
/bg/ ‘large in size’
/r/ ‘more’
As the examples in (24) show, morphemes are not necessarily equivalent to either words or syllables.
Further, a particular sequence of sounds can be a morpheme in one word and not another: notice [r]
in speaker is a morpheme, whereas the [r] in antler is not. Similarly [pik] is a morpheme in peeks
[piks], but not in speaks [spiks]. Lastly, observe that the morpheme /r/ either indicates an agent or
expresses the comparative (more).
115

116
Morphemes have different phonetic variants in different contexts just as phonemes do. Paralleling
the PHONEME/ALLOPHONE relation, we have the MORPHEME/ALLOMORPH relation.
An ALLOMORPH is a conditioned phonetic variation of a morpheme. Consider the following:
(25) ALLOMORPH: a conditioned PHONETIC variation of a morpheme:
a. Negative Prefix:
[m]: impossible, imbalance
[n]: intangible, indiscrete
[]: incorrigible, ingratitude
b. The indefinite article:
[]: a book, a use, a hotel
[æn]: an owl, an honor
For the grammarian, one point of particular interest in morphological analysis concerns the nature
of the variations which specific morphemes show in their phonetic shape. Occasionally, this
variation is peculiar and reflects the idiosyncratic speech habits of individual speakers. For example,
some English speakers pronounce abdomen with heavy stress on the first syllable, while others
pronounce it with heavy stress on the second syllable. In such cases as this, the rival forms are said
to be in FREE VARIATION.
More commonly, a variation in the pronunciation of a morpheme is predictable and reflects specific
rules in the grammar of a particular language, sometimes called MORPHOPHONEMIC or
MORPHOPHONOLOGICAL RULES. For example, the plural morpheme in English nouns is
variously pronounced [z] as in busses, [s] as in bucks, and [z] as in bugs.
The occurrence of each of these allomorphic variants is predictable in terms of the phonetic
environment in which the plural morpheme occurs. If the morpheme occurs after [s], [z], [š], [ž],
[], or [], it is realized phonetically as [z], e.g., dresses, sizes, ashes. Otherwise, if it occurs after
a voiceless consonant, it is realized phonetically as [s], e.g., tanks, pipes, slots; if it occurs after a
voiced consonant or after a vowel, it is realized phonetically as [z], e.g., bulbs, trees, eggs. Utilizing
the theory of distinctive features, we can formally represent these rules as follows (where PL means
‘plural’):
(26) a. PL [z] / [+SIBILANT] ___
b. PL [s] / [–VOICED] ___
c. PL [z] / [+VOICED] ___
These same variations are also found in the morpheme which indicates the third person singular
present tense of verbs, e.g., [z] as in blesses and pleases, [s] as in attempts and pollutes, and [z] as
in confides and specifies. They also appear in contractions (he’s for he is) and in possessives, e.g.,
[z] as in Bess’s and Josh’s, [s] as in Pat’s and Skip’s, and [z] as in Budd’s and Don’s.

2.5 THE BIOLOGICAL BASIS OF PHONOLOGY
In Chapter One, we discussed two fundamental problems for linguistic theory, the grammatical
characterization problem and the grammatical realization problem. We pointed out that the ultimate
objective of linguistic research is to uncover the specifications of UG, the collection of features and
principles which are potential characteristics of any human language. This goal is achieved by
relating the facts of grammatical characterization to the facts of grammatical realization.
The rules we have considered so far are descriptively adequate (see Chapter One) as far as the
examples cited are concerned; however, they are not general enough. We can attain a higher level
of descriptive adequacy. In this regard, linguistics is not different from any other discipline,
business, or enterprise. Everyone is searching for the most cost-effective solution to problems.
Consider, first, rule (15a) and notice that both members of the pair [t]–[] are voiceless. If our theory
is correct, we expect the other members of this natural class to behave similarly, that is, we expect
the grammar to contain (27).
(27) [d] [] / ___ [y]
In fact, English does have (27), as (28) shows.
(28) a. If you hide me, I’ll hide you.
b. If you hide me, /ay wl hayd yu/
c. If you hide me, [ayl hay u].
Since the sounds involved form a natural class, we can collapse (15a) and (27) to (29), thereby
attaining a higher level of descriptive adequacy.
(29) [–SIBILANT, +ANTERIOR] [+SIBILANT, –ANTERIOR] / ___[y]
This rule appears to be correct; however, we have gone too far. The feature constellation
[–SIBILANT, +ANTERIOR] includes many consonants other than [t] and [d]; specifically, it
includes [p], [b], [m], [n], [], [ð], [l], [r], and [y]. None of these sounds becomes an affricate before
[y]; therefore, we amend (29) to (30):
(30)
117

118
Notice that (30) involves a change of two distinctive features; the other features remain unchanged.
The result is that an alveolar stop ([t] or [d]) followed by the anterior glide ([y]) becomes an affricate
([] or [], respectively). It is important to realize that one letter or sound does not turn into another.
The letter t and the sound [t] are merely abbreviations for constellations of distinctive features
derived from characteristics of the human vocal apparatus. It is these features that change in the
operation of phonological rules. In (30), two of the features in the feature cluster change value; the
others remain unchanged.
Notice also that our concern for formalism, that is, precision, now pays off, because only now can
we ask the really important questions: Why does the grammar of a human language contain a rule
like (30)? Do children born in an English speaking environment have to learn this rule specifically,
as a peculiar fact about English, in the way that they must learn that the past tense of go is went?
The answer to these questions involves a dynamic that exists between the structure of the human
vocal apparatus and the general phonology of English. This dynamic is influenced by a number of
factors.
First, notice that rapid articulation of the sound sequence [tyu] sounds very much like [u], and the
articulation of [dyu] sounds very much like [u]. The reason for this is due to the position the tongue
must assume during the articulation of the sequences [tyu] and [dyu]. Thus, it would be quite natural
for speakers to merge [t]/[d] and [yu] across a word boundary, substituting [u] for [tyu] and [u]
for [dyu] in rapid articulation.
Second, this substitution would be particularly likely under two conditions: (i) if the phonology of
the language includes the sounds [] and [] as phonemes, and (ii) if the language does not have a
phonological distinction between [ty] and [], or [dy] and []. These two conditions both exist in
English. The sounds [] and [] are common in all dialects (chew, church, chat, chair, cheer; Jew,
judge, jet, jam, jeer, etc.). Furthermore, most dialects of English do not have phonetic sequences
of [t] or [d] followed by [y] unless there is a pause in pronunciation. The usual pronunciation of
tune, for example, is [tun], not [tyun].
For these reasons, in rapid speech, it is natural for an English speaker, knowing the overall
phonology of English, to apply (30). What this means is that, in the overall phonology of any given
language, some phonological rules may follow naturally given other factors. The suggestion,
therefore, is that an English speaking child does not have to learn (30) specifically as a peculiar,
isolated fact about English.
This analysis clearly indicates that some phonological processes are at least partially predictable.
In other words, children learning a language are not faced with the task of acquiring a set of rules
that are isolated, idiosyncratic, and arbitrary. Apparently, language acquisition is possible precisely
because a significant portion of the grammar of all natural languages is a direct reflection of the
nature of the apparatus used to produce them, that is, of the articulatory and auditory organs and the
brain.

This issue is important enough to require careful scrutiny. Therefore, let us consider a more
extended example and return to the rules for forming the plural given in (26). As we discussed in
the Introduction (page 12 ff.), there is an alternation in the allomorph which signals past tense in
English that is similar to the rules in (26). Consider the following (where PST means ‘past tense’):
(31) a. PST [d] / {[t], [d]} ___
b. PST [t] / [–VOICED] ___
c. PST [d] / [+VOICED] ___
Rule (31a) correctly specifies past tense forms like relented ([rilntd]) and applauded
([plawdd]); (31b), disliked ([dslaykt]) and laughed ([læft]); (31c), nagged ([nægd]), riled
([rayld]), and relayed ([riled]). Notice that all past tenses of regular verbs in English are spelled
either –ed or –d; but, this is not relevant to their pronunciation. Compare riled ([rayld]) and relayed
([riled]). As we have seen, the rules of language are based on speech, not writing.
The important question now is, Can the characterizations in (26) for the plural and (31) for the past
tense be related to universal grammar? The answer is, Yes. The following sections specify that
relationship.
2.5.1 FIRST SET OF OBSERVATIONS
Notice that (26) and (31) are very similar. Both sets of rules begin with a rule that deals with the
allomorphic variant that occurs when the morpheme and the stem to which it is suffixed are
HOMORGANIC, that is, share the same or very similar set of distinctive features.
(26a) deals with the suffixation of a [+SIBILANT, +CORONAL] morpheme to a word which ends
in a sound that is [+SIBILANT, +CORONAL]. (31a) deals with the suffixation of an alveolar stop
([–CONTINUANT, –NASAL, +ANTERIOR, +CORONAL]) to a word which ends in an alveolar
stop. In both sets of rules, a separate syllable is created by inserting the neutral vowel [] in order
to distinguish the morpheme from the stem to which it is being attached. Furthermore, if this were
not done, some sequences violating English (if not, universal) phonotactics would be the result, e.g.,
*[æss] and *[plawdd].
In the remaining rules in both sets, a very common phonological process occurs. This process, called
ASSIMILATION, involves making adjacent sound segments similar in distinctive feature
composition. In (26) and (31) the assimilation involves the feature [±VOICED]. Assimilation in
natural languages is so common that it is considered the UNMARKED or expected case. Lack of
assimilation in suffixation is the MARKED or atypical case.
Given these remarks, we specify the morphemes in question as (32), and we replace (26) and (31)
with the more general principles in (33).
119

120
(32) a. The PST morpheme in English is the suffix [d].
b. The PL morpheme in English is the suffix [z].
(33) a. If a suffix and the final segment to which it is attached are homorganic, insert the
neutral vowel [] between the stem and the suffix.
b. In all other cases, assimilate the suffix and the final segment in voicing.
In order to express principles like (33) formally, we must have a way to represent a variable feature
value. This is done using Greek letters in the spot in front of a feature where a “+” or “–” usually
occurs. For example, we express voicing assimilation (33b) as follows:
(34) [+CONSONANT] [VOICED] / [VOICED] + ___
This rule says that a consonant beginning a suffix gets a value of voicing (either “+” or “–”) that is
the same as the final segment before the morpheme boundary. Notice that “+” has two meanings:
before a feature as in [+VOICED], it indicates a positive value for the feature; by itself in a rule like
(34), it indicates a morpheme boundary.
Sometime, several features assimilate at the same time. In this case, more than one Greek letter is
necessary to separate the assimilating features from each other. For example, consider again the
negative prefix given in (25a), repeated here:
(35) a. [m]: impossible, imbalance
b. [n]: intangible, indiscrete
c. []: incorrigible, ingratitude
In this set of data, there is an assimilation of the features ANTERIOR and CORONAL to those of
segment beginning the word. This means that the value (“+” or “–”) of the features ANTERIOR and
CORONAL of the consonant beginning the word must be the same as the prefix. We can express
this as follows:
(36) [+NASAL] [ANTERIOR, CORONAL] / ___+ [ANTERIOR, CORONAL]
In this rule, we must keep the assimilating values for ANTERIOR separate from those of
CORONAL; hence, two Greek letters are needed. The value of “” and “” must be the same in
both places where each occurs in the rule. In short, (36) abbreviates the following four rules:
(37) [+NASAL] [+ANTERIOR, –CORONAL] / ___+ [+ANTERIOR, –CORONAL]
(38) [+NASAL] [–ANTERIOR, +CORONAL] / ___+ [–ANTERIOR, +CORONAL]
(39) [+NASAL] [+ANTERIOR, +CORONAL] / ___+ [+ANTERIOR, +CORONAL]
(40) [+NASAL] [–ANTERIOR, –CORONAL] / ___+ [–ANTERIOR, –CORONAL]

Clearly, a rule like (36) will be needed to express the homorganicity in (33a). For example, if we
wish to express the insertion of [] before the past tense morpheme [d], then (41) is needed.
(41)
Similarly, (42) is needed for the insertion of [] before the plural morpheme [z]:
(42)
Using the –notation, we can collapse these to (43), the formal expression of (33a) .
(43)
2.5.2 SECOND SET OF OBSERVATIONS
We have now correctly identified the morphemes for the plural and for the past tense (cf. (32)),
repeated here as (44).
(44) a. The PST morpheme in English is the suffix [d].
b. The PL morpheme in English is the suffix [z].
We have also formally expressed the distribution of the allomorph in (33a)/(43) and (33b)/(34),
repeated here together as (45).
121

122
(45) a. If a suffix and the final segment to which it is attached are homorganic, insert the
neutral vowel [] between the stem and the suffix.
b. In all other cases, assimilate the suffix and the final segment in voicing.
[+CONSONANT] [VOICED] / [VOICED] + ___
How does this characterization in (44) and (45) relate to what speakers know? It is clear that native
speakers of English know what the two morphemes are and how they are distributed. English
speakers do not have to hear the plural of every noun and the past tense of every verb in order to
produce the correct forms. In fact, if speakers are given nonsense words and asked to make them
plural or past tense, they produce forms which obey (44) and (45). Moreover, speakers make up
words in the correct number and tense all the time. For example, during a TV commentary about
a figure skating competition, Peggy Fleming said, “She two–footed that landing.” Members of the
television audience did not wince at the new verb two–foot; everyone routinely understood it.
Further, the correct allomorph was uttered and understood, even though the expression must have
been a neologism to most who heard it. Examples like this occur very frequently.
Young children also produce forms that they never could have heard: past tenses like [god] ‘goed’
and [rod] ‘throwed’; plurals like [mænz] ‘mans’ and [fts] ‘foots.’ These overgeneralizations give
clear evidence that the substance of (44) and (45) is unconsciously known to native speakers of
English from a very early age.
Do native speakers (unconsciously) know (44) and (45) in the form given? The answer is not clear.
While the substance of (44) and (45) can be expressed in many different ways, the characterization
given is the most generalized. Science usually tries to account for natural phenomena with the most
generalized descriptions possible. But there is no way to determine whether or not people do the
same when they learn their native language. In fact, it may be the case that the rules of language are
redundantly specified. Future psycholinguistic and neurolinguistic investigations may help to settle
this issue.
At this point, linguistics is in the same position as other disciplines, businesses, and enterprises. We
have produced the best possible product our ingenuity and science have allowed, namely, (44) and
(45). It remains to be seen how well our product fares upon further scrutiny.

2.5.3 THIRD SET OF OBSERVATIONS
Summarizing, we can say that (44) and (45) reach a very high level of descriptive adequacy. They
adequately account for all the known cases, and do so in a highly generalized way. Further, the
substance of (44) and (45) is unconsciously known to every native speaker.
Consider next the question whether children have to learn both (44) and (45). It seems clear that
children are not born with unconscious knowledge of (44), because (44) contains facts peculiar to
English. There is no reason in principle why, say, the meanings of the morphemes couldn’t be
reversed in some other language. The rule (45a) is also peculiar to English; however, its effect is to
differentiate morphemes, so we would expect other languages to have rules of similar design.
The status of (45b) is another matter. The rule (45b) is a good candidate for UG status, that is, it is
possible that (45b) is part of every human being’s genetic makeup. This would mean that it follows
directly and naturally from the nature of the human language apparatus, and that a particular form
in a particular language violating (45b) would be MARKED in the sense described above. Human
languages generally have a voicing assimilation rule, apparently because it is easier for a human
vocal apparatus to articulate adjacent sounds with the same voicing than with opposite voicing. This
is particularly true at the end of a word. As a result, we say that children do not have to learn (45b);
it is part of their genetic endowment; it is the UNMARKED case.
While this discussion is far from conclusive, it does show that all the facts in the characterization
of a language are not equal. Some facts may be highly marked, and, therefore, take children a long
time to master. Other facts may reflect inner characteristics of man, and, therefore, be part of the
innate knowledge that children bring to native language acquisition. Because children have a human
language apparatus, some features and principles are more likely to occur in a human language than
others.
2.6 THE PHONOLOGICAL AND LEXICAL COMPONENTS OF A GRAMMAR
The PHONOLOGICAL COMPONENT of a grammar consists of the list of phonological rules.
These are rules like (34), (36), and (43), which relate underlying forms to superficial forms. Also
included in the phonological component are phonotactic rules like (20) and phonological
REDUNDANCY RULES like (46).
(46) a. [+HIGH] [+HIGH, –LOW]
b. [+LOW] [+LOW, –HIGH]
(46a) accounts for the fact that a segment that is [+HIGH] cannot also be [–LOW], and (46b) account
for the reverse. These are redundantly specified features, that is, features that are predictable given
other features.
123

124
Vowels have many redundant features. Since only vowels are [+TENSE], we can specify the
redundancies as follows:
(47) [+TENSE] [+TENSE, +SONORANT,–CONSONANTAL, +VOCALIC,
+CONTINUANT, –NASAL, –SIBILANT, –ANTERIOR,
–CORONAL, +VOICED]
Since all the features except TENSE in (47) are redundant, they are generally left out of charts of
distinctive features for vowels (cf. Figure Three on page 96).
The LEXICAL COMPONENT (LEXICON) of a grammar is the dictionary of morphemes and
words, as well as lexical redundancy rules and rules of word formation. (38) is an example of lexical
redundancy.
(48) [+HUMAN] [+HUMAN, +ANIMATE, +CONCRETE]
The rule (48) accounts for the fact that all HUMAN nouns are also ANIMATE and CONCRETE.
The lexicon also contains morphemes like (49).
(49) Morphemes:
a. /trst/ trust, trusts, trusted, mistrust, mistrustful, entrust, untrustworthy, etc.
b. /red/ raid, raids, raided, raiding, raider, raiders
From these components, we can produce underlying representations like (50).
(50) Underlying phonological representations:
a. /trst+s/
b. /trst+d/
c. /ms+trst/
d. /ms+trst+fl/
Application of various phonological rules to (50) produces the following phonetic representations,
respectively:
(51) Phonetic representations (ignoring details involving aspiration and release):
a. [trsts]
b. [trstd]
c. [mstrst]
d. [mstrstfl]

To these forms, various stress rules must apply assigning the appropriate intonation contours to
words. We will return to a discussion of the lexicon in subsequent chapters.
2.7 ENGLISH SPELLING
It is no secret that English words are not spelled phonetically. Words spelled the same can have
different pronunciations like read [rid]/[rd], wind [wnd]/[waynd], wound [wund]/[wawnd], bow
[bo]/[baw], close [klos]/[kloz], number [nmbr]/[nmr], invalid [´nvld]/[nvæ´ ld], refuse
[rfúz]/[r´fus], etc. In fact, lists of spelling peculiarities like the following are commonplace:
(52) a. rafter / laughter / slaughter
b. slay / sleigh / sleight / slight
c. ear / earn / urn / earth / hearth
d. though / thought / through / tough
e. for / fore / four / fourteen / forty
Often, the same sound has many different spellings:
(53) a. heard / word / bird / curd / nerd
b. rays / raise / raze / weighs / phase / obeys
c. weird / feared / veered / adhered
d. birth / berth / worth / earth
Likewise, similar spellings can signify widely different sounds:
(54) a. beard [bird] / heard [hrd]
b. comb [kom] / tomb [tum] / bomb [bam]
c. meat [mit] / great [gret] / threat [rt] / heart [hart]
d. cough [kf] / bough [baw] / dough [do] / rough [rf] / through [ru]
One frequently hears desperate cries for reform of the English spelling system. While there are many
peculiarities, it is clear that a system of phonetic spelling would undermine the considerable
regularity between English spelling and English morphology. For example, consider the various
spellings of the reduced vowel schwa ([]) in the second syllable of fallacy, remedy, subsidy, history,
luxury. In these, as in other cases, the spelling derives from the spelling of the root word as
comparisons reveal; notice the spelling when the vowel is stressed:
(55) a. Spelled a: fallacy/fallacious; comparable/compare; legacy/allegation
b. Spelled e: remedy/remedial; manager/managerial; competent/compete
c. Spelled i: subsidy/subside; immigrate/migrate; sublimate/sublime
d. Spelled o: history/historian; consolation/console; janitor/janitorial
e. Spelled u: luxury/luxurious; injury/injurious; mercury/mercurial
125

126
Similarly, silent letters frequently show up in related words:
(56) a. Silent n: hymn/hymnal; column/columnar, columnist; damn/damnation
b. Silent g: resign/resignation; sign/signal, designate; phlegm/phlegmatic
c. Others: muscle/muscular; bomb/bombard; soften/soft; know/acknowledge
There are also many regularities. For example, consider the vowel alterations in (47) where the
silent e indicates that the preceding vowel must be tense.
(57) Lax/Tense Vowel Alternations
a. [i]/[]: serene/serenity; redeem/redemption; sleep/slept; meter/ metric
b. [e]/[æ]: profane/profanity; sane/sanity; bathe/bath; shade/shadow
c. [ay]/[]: divine/divinity; ignite/ignition; wide/width; type/typical
d. [u]/[]: lose/lost; crucible/cross; food/fodder
e. [o]/[a]: verbose/verbosity; know/knowledge; nose/nostril; phone/phonic
f. [aw]/[]: profound/profundity; abound/abundant; south/southern
These vowel alternations are rule governed as are the consonant alternations in (58) and (59).
(58) Stop/Fricative Alternations
a. [t/[s]: democrat/democracy; president/presidency; pirate/piracy
b. [k/[s]: electric/electricity; public/publicity; medical/medicine
c. [s]/[š]: express/expression; digress/digression; race/racial; space/spatial
d. [z]/[ž]: revise/revision; infuse/infusion; excise/excision; confuse/confusion
e. [t/[š]: relate/relation; ignite/ignition; native/nation; president/presidential
f. [k/š]: logic/logician; magic/magician; practical/practitioner
g. [d]/[ž]: persuade/persuasion; corrode/corrosion; decide/decision
h. [d]/[s] persuade/persuasive/; corrode/corrosive; divide/divisive
I. [z]/[s] abuse (verb)/abusive; diffuse/diffusive; advise/advice; choose/choice
(59) Stop/Affricate Alternations
a. [d]/[]: grade/gradual; residue/residual; fraud/fraudulent; proceed/procedure
b. [t/[]: right/righteous; Christ/Christian; quest/question; moist/moisture
c. [g]/[]: allegation/allege; regal/regicide; rigor/rigid; gregarious/egregious
Basically, changing English spelling to a system that is phonetic would obscure many of the regular
allomorphic variations in the language. A much better approach would be to teach students
something about English phonology and morphology pointing out that English spelling reveals
underlying regularities. Discussion of the rules governing these alternations requires considerable
technical mastery and is beyond the capabilities of elementary school children trying to cope with
English spelling. However, it would certainly be both possible and helpful to point out the
regularities as the words entered children’s vocabulary studies.

2.8 SUMMARY OF RULE FORMALISM
SYMBOL MEANING
1. [x] x is a PHONETIC symbol or feature
2. /x/ x is a PHONEMIC symbol or feature
3. Ø null
4. is
5. / in the environment of
6. ___ before, after, or between
7. / ___ x before x
8. / x ___ after x
/ x ___ y between x and y
9. # word boundary
/ # ___ at beginning of word
/ ___ # at end of word
10. + morpheme boundary
/ + ___ after a morpheme boundary
/ ___ + before a morpheme boundary
11. C CONSONANT
12. V VOWEL
13. (x) x is an OPTIONAL element
14. {x, y, z} x, y, and z are ALTERNATIVE elements
15. [x, y, z] x, y, and z are a cluster of features
2.9 NOTES ON SYLLABLES
A word has as many syllables as it has vowels. Each vowel can be preceded and followed by any
number of consonants. The superscript on the symbol “C” indicates the maximum number of
consonants (n=any number greater than 1); the subscript indicates the minimum number (0=none).
1. All one syllable words: # C V C #
2. All two syllable words: # C V C V C #
3. All three syllable words: # C V C V C V C #
n n
n
0
n
0
n
n
0
n
0
n
0
n
0 0 0 0
4. Vowel in the last syllable (ULTIMA): / __ C #
5. Vowel in the next to last syllable (PENULT): / __ C V C #
6. Vowel in the third to last syllable (ANTEPENULT): / __ C V C V C #
n
0
n n
n
0
n
0
n
0 0 0
127

128
VOWELS:
SUMMARY OF ENGLISH SOUNDS AND SPELLING
[i] [iy] [i] beat, beet, Caesar, people, we, machine, receive, believe
[] bit, into, been, abyss, fountain, button, language
[u] [uw] [u] boot, who, duty, through, too, to, two, new, nuisance, blue
[] put, wood, would, butcher, woman
[e] [ey] [e] bait, fate, great, eight, rain, reign, rein, they, hay, gauge
[] bet, dead, said, guest, many, friend, says, jeopardy
[] butted, roses, alone, sofa, the
[o] [ow] [o] boat, so, sew, own, though, plateau, toe, soul, brooch
[] bought, caught, chalk, saw, awful, author, sore, soar, four
[æ] bat, laugh, plaid
[] but, does, tough, flood, ton, country
[a] [] pot, palm, far, sergeant, ox
DIPHTHONGS:
[y] [] boy, soil
[ay] [a] sigh, nice, guy, lie
[aw] [a] cow, house
GLIDES:
[w] wish, queen, weather, wear (for some, also when, whether, where)
[] when, whether, where
[y] [j] yes, beyond, use, feud
[h] high, whole
LIQUIDS:
[l] live, along
[] drill, willing
[r] [] rich, car, bury, furry
[] [D] water, later, latter, ladder, writer, rider, party

NASALS:
SUMMARY OF ENGLISH SOUNDS AND SPELLING
[m] mess, hymn, hum, swimming
[n] neat, kin, winning, pneumonia, knife
[ñ] [ny] [nj] onion, union
[] sing, think, tongue
STOPS:
h
[p ] pain, prune, plaque, appear, shrimps
[p] spit, spun, spring, expose
[b] broom, crabs, rubber
h
[t ] tip, tough, pterodactyl, intend, attack
[t] stick, stun, string, extend
[d] dig, bad, killed, heard, mending
h
[k ] keep, cat, chord, quite, pickle, accommodate
[k] skit, scum, school, screw, excuse
[g] give, bag, egg, rogue
FRICATIVES:
[f] first, stuff, telephone, rough
[v] value, sliver, of
[] thin, breath, teeth, ether, both, thing
[ð] then, breathe, teethe, either
[s] sin, this, cross, science, ceiling, race, psychology
[z] zoo, drizzle, was, xylophone, scissors
[š] [] show, chute, flash, issue, conscious, sure, mansion, racial
[ž] [] measure, azure, prestige, illusion
AFFRICATES:
[] [t] chin, kitchen, question, righteous, much
[] [d] jig, judgment, ridge, gem, graduation, region
129

EXERCISES FOR CHAPTER TWO
1. Transcribe the following English words (Answers in Appendix A):
rich ridge ring wring wrist
bush butch budge box buzz
gem Jim sham chum gum
ice eyes east eased oozed
race raise rays rise rose
lose loose louse chose choose
cease seize seas says size
chef chief chic sheik shack
worth earth teethe bathe bath
them this thumb thing then
debt vex Scotch broth knot
maize quiche squish queue shrew
though cough rough bough through
heard word bird curd nerd

weird feared veered cared cord
meat great threat heart cart
chair jeer share shear cheer
wheel whale weigh who how
witch which watch wretch wedge
coat code root rude read
laud loud laid led lewd
knifed sighed prized healed shipped
view dew do poor pour
latch loaf roof tune pure
car oar her hear whore
George vague vast don’t they’re
oath both moth mouth myth
tit Tut tat toot tot
who’d wood would wooed hood
131

132
2. Transcribe the following English words (Answers in Appendix A):
music physics psychic sonic helix
passion cashier cushion caption cashmere
thistle thimble tumble dangle dinghy
champagne cartoon Khartoum careen gangrene
immune impugn impinge syringe obscene
racial racer razor sugar quota
pleasure pledger leisure azure seizure
scissors vision fusion ocean mission
beauty booty purely poorly pearly
quitter water rider writer raider
joining whining singing finger languor
cavil navel ravel gable squabble
Mary marry merry quarry query
numbskull zigzag homegrown deathblow shipshape
outboard marshlands whiplash pastime pitchfork
cutthroat whetstone hangman onslaught crabgrass

3. Transcribe the following words including placement of primary stress (´). You can ignore
aspiration and release. (v) = verb; (n) = noun.
agree debris ennui Pawnee marquee
angry money sunny gypsy litchi
survey (v) Bombay bouquet inveigh parfait
survey (n) bomb bay Tuesday foray subway
review canoe shampoo construe eschew
preview igloo curfew yahoo Zulu
below outgrow plateau Bordeaux although
bellow meadow hobo cocoa depot
outcry (v) apply goodbye sky high untie
outcry (n) ally cacti fish fry pigsty
soda tchotchke bwana chutzpah geisha
4. Transcribe the following words including placement of primary stress (´). You can ignore
aspiration and release. (v) = verb; (n) = noun.
broccoli apogee fricassee symphony pedigree
shillelagh Chianti bologna confetti Hatari
guarantee bourgeoisie chimpanzee Tennessee jamboree
133

134
runaway Saturday ricochet protégée workaday
matinée San José Chevrolet Santa Fé overplay
rendevous residue peekaboo honeydew avenue
Timbuktu kangaroo misconstrue overthrew hitherto
mistletoe Mexico embryo cameo Scorpio
fiasco concerto soprano falsetto crescendo
overflow (v) apropos status quo cheerio to-and-fro
amplify rockaby occupy prophesy underlie
subpoena agenda charisma pariah pagoda
5. Transcribe the following words including placement of primary stress (´). You can ignore
aspiration and release. (v) = verb; (n) = noun.
apostrophe Menominee Penelope fraternity telegraphy
Ypsilanti kamikaze Cincinnati insincerely poison ivy
pistachio Arapaho imbroglio adagio portfolio
armadillo Filipino virtuoso Sacramento lucky fellow
Arabia azalea Bulgaria et cetera America
hullabaloo Kalamazoo Tippecanoe

6. In each of the following, supply English words that contain the sound given in as many
different spellings as you can. Underline the letters representing the sound in the answer.
[f] Answer: physics, fish, enough
a. [i] Answer:
b. [r] Answer:
c. [k] Answer:
d. [š] Answer:
e. [y] Answer:
f. [e] Answer:
g. [o] Answer:
h. [] Answer:
I. [] Answer:
j. [] Answer:
7. Supply the symbol corresponding to each of the following phonetic descriptions.
voiced velar stop consonant Answer: [g]
a. lax mid back rounded vowel Answer:
b. voiced alveolar liquid Answer:
c. voiceless palatal affricate consonant Answer:
d. tense low back unrounded vowel Answer:
e. voiceless alveolar fricative consonant Answer:
f. voiced palatal liquid Answer:
g. voiceless bilabial stop consonant Answer:
135

136
h. tense high front unrounded vowel Answer:
i. voiced velar nasal consonant Answer:
j. voiced labiodental fricative consonant Answer:
8. Write a phonetic description for each of the following symbols.
[g] Answer: voiced velar stop consonant
[o] Answer: tense mid back rounded vowel
a. [h] Answer:
b. [v] Answer:
c. [e] Answer:
d. [o] Answer:
e. [ž] Answer:
f. [] Answer:
g. [æ] Answer:
h. [ñ] Answer:
I. [] Answer:
j. [] Answer:
9. In each of the following, all but one of the sounds listed exhaustively comprise a natural class
of sounds for English. Identify the sound that does not belong in the natural class AND name
the feature or features that define the class.
[k], [d], [g], [b] Choice: [k]
Class: [+VOICED, +STOP]
a. [a], [], [], [æ] Choice:
Class:

. [k], [], [h], [g] Choice:
Class:
c. [z], [], [ð], [ž], [v] Choice:
Class:
d. [t], [], [z], [ð], [s] Choice:
Class:
e. [u], [], [o], [e], [i] Choice:
Class:
f. [w], [h], [y], [] Choice:
Class:
g. [u], [o], [a], [], [] Choice:
Class:
h. [m], [w], [v], [b], [p] Choice:
Class:
I. [u], [i], [e], [], [] Choice:
Class:
j. [], [m], [n], [h], [ñ] Choice:
Class:
10. State the following rules using formal devices (Answers in Appendix A).
Example: A nasal consonant is deleted before another nasal consonant at the beginning of
a word.
Answer: [+CONSONANT, +NASAL] Ø / #___ [+CONSONANT, +NASAL]
a. All intervocalic voiceless stops become voiced.
137

138
b. Nontense vowels in the final syllable of a word are stressed when they are followed by
two or more consonants.
c. Delete all final consonants when the following word begins with a consonant.
d. Reduce all unstressed vowels to schwa.
e. Palatalize all alveolar stops before [y].
f. Devoice all final fricatives.
g. Aspirate all initial voiceless stops.
h. Nasalize all vowels before a nasal consonant.
i. A lax vowel in the antepenultimate syllable of a word is stressed.
j. Palatalize all alveolar stops before [y] when they are preceded by a vowel or when they
are initial.

11. In Classical Latin, nouns have different endings (CASES) depending on their use in a
sentence. For example, a noun functioning as the SUBJECT of a sentence has a
NOMINATIVE CASE ending. One functioning as a POSSESSIVE phrase has a GENITIVE
CASE ending. Consider the following, where a macron over a vowel means the vowel is long:
NOMINATIVE GENITIVE MEANING
arks arkis enclosure
duks dukis leader
daps dapis feast
rks rgis king
sps spis snake
falanks falangis phalanx
urps urbis city
pleps plebis people
a. In terms of these data, what are the morphemes for the NOMINATIVE and the
GENITIVE?
b. What is the Latin morpheme for each of the following?
(1) enclosure
(2) leader
(3) feast
(4) king
(5) snake
(6) phalanx
(7) city
(8) people
c. State any phonological rule(s) applicable in the above Latin data.
139

140
12. Consider the following additional Classical Latin data:
NOMINATIVE GENITIVE MEANING
lis litis strife
fraws frawdis deceit
frns frntis brow
frns frndis leaf
dens dentis tooth
sors sortis lot
a. What are the morphemes for the NOMINATIVE and the GENITIVE?
b. Give the Latin morphemes for each of the following:
(1) strife
(2) deceit
(3) brow
(4) leaf
(5) tooth
(6) lot
c. State any phonological rule(s) in addition to those from the previous problem.

APPENDIX A: ANSWERS TO TRANSCRIPTION EXERCISES
1. rich ridge ring wring wrist
[r] [r] [r] [r] [rst]
bush butch budge box buzz
[bš] [b] [b] [baks] [bz]
gem Jim sham chum gum
[m] [m] [šæm] [m] [gm]
ice eyes east eased oozed
[ays] [ayz] [ist] [izd] [uzd]
race raise rays rise rose
[res] [rez] [rez] [rayz] [roz]
lose loose louse chose choose
[luz] [lus] [laws] [oz] [uz]
cease seize seas says size
[sis] [siz] [siz] [sz] [sayz]
chef chief chic sheik shack
[šf] [if] [šik] [šik] [šæk]
worth earth teethe bathe bath
[wr] [r]
h
[t ið] [beð] [bæ]
them this thumb thing then
[ðm] [ðs] [m] [i] [ðn]
debt vex Scotch broth knot
[dt] [vks] [ska] [br] [nat]
maize quiche squish queue shrew
[mez] h [k iš] [skwš] h [k yu] [šru]
though cough rough bough through
[ðo]
h
[k f] [rf] [baw] [ru]
heard word bird curd nerd
[hrd] [wrd] [brd] h
[k rd] [nrd]
141

142
weird feared veered cared cord
[wird] [fird] [vird] h [k ærd] h [k rd]
meat great threat heart cart
[mit] [gret] [rt] [hart] h [k art]
chair jeer share shear cheer
[ær] [ir] [šær] [šir] [ir]
wheel whale weigh who how
[wil] [wæl] [we] [hu] [haw]
witch which watch wretch wedge
[w] [w] [wa] [r] [w]
coat code root rude read
h
[k ot]
h
[k od] [rut] [rud] [rid]
laud loud laid led lewd
[ld] [lawd] [led] [ld] [lud]
knifed sighed prized healed shipped
[nayft] [sayd]
h
[p rayzd] [hild] [špt]
view dew do poor pour
[vyu] [du] [du]
h
[p ur]
h
[p r]
latch loaf roof tune pure
[læ] [lof] [ruf]
h
[t un]
h
[p yur]
car oar her hear whore
h
[k ar] [or] [hr] [hir] [hor]
George vague vast don't they're
[or] [veg] [væst] [dont] [ðer]
oath both moth mouth myth
[o] [bo] [m] [maw] [m]
tit Tut tat toot tot
h [t t] h [t t] h [t æt] h [t ut] h [t at]
who'd wood would wooed hood
[hud] [wd] [wd] [wud] [hd]

2. music physics psychic sonic helix
[myuzk] [fzks] h [sayk k] [sank] [hilks]
passion cashier cushion caption cashmere
h [p æšn] h [k æšir] [khšn] h [k æpšn] h [k æžmir]
thistle thimble tumble dangle dinghy
[sl] [mbl] h [t mbl] [dægl] [digi]
champagne cartoon Khartoum careen gangrene
h [šæmp en] h h [k art un] h h [k art um] h [k rin] [gægrin]
immune impugn impinge syringe obscene
[myun] h [mp yun] h [mp n] [srn] [absin]
racial racer razor sugar quota
[rešl] [resr] [rezr] [šgr]
h
[k wo]
pleasure pledger leisure azure seizure
h
[p lžr]
h
[p lr] [lžr]/[ližr] [æžr] [sižr]
scissors vision fusion ocean mission
[szrz] [vžn] [fyužn] [ošn] [mšn]
beauty booty purely poorly pearly
[byui] [bui]
h
[p yurli]
h
[p rli]
h
[p rli]
quitter water rider writer raider
h
[k wr] [wr] [rayr] [rayr] [rer]
joining whining singing finger languor
[yni] [wayni] [sii] [figr] [lægr]
cavil navel ravel gable squabble
h
[k ævl] [nevl] [rævl] [gebl] [skwabl]
Mary marry merry quarry query
[mæri] [mæri] [mri]
h
[k wari]
h
[k wiri]
numbskull zigzag homegrown deathblow shipshape
[nmskl] [zgzæg] [homgron] [dblow] [špšep]
outboard marshlands whiplash pastime pitchfork
[awtbrd] [maršlændz] [wplæš]
h h
[p æst aym]
h
[p frk]
cutthroat whetstone hangman onslaught crabgrass
[ktrot] [wtston] [hæmæn] [anslt] [kræbgræs]
143

144
3. agree debris ennui Pawnee marquee
[grí] [dbrí] [anwí] [pní] [markí]
angry money sunny gypsy litchi
[æ´ gri] [m´ni] [s´ni] [ípsi] [líi]
survey (v) Bombay bouquet inveigh parfait
[srvé] [bambé] [boké] [nvé] [parfé]
survey (n) bomb bay Tuesday foray subway
[s´rve] [bámbe] [túzde] [f´re] [s´bwe]
review canoe shampoo construe eschew
[rvyú] [knú] [šæmpú] [knstú] [šú]
preview igloo curfew yahoo Zulu
[prívyu] [´glu] [k´rfyu] [yáhu] [zúlu]
below outgrow plateau Bordeaux although
[bló] [awtgró] [plætó] [brdó] [lðó]
bellow meadow hobo cocoa depot
[b´lo] [m´do] [hóbo] [kóko] [dípo]
outcry (v) apply goodbye sky high untie
[awtkráy] [pláy] [gdbáy] [skayháy] [ntáy]
outcry (n) ally cacti fish fry pigsty
[áwtkray] [æ´ lay] [kæ´ ktay] [f´šfray] [p´gstay]
soda tchotchke bwana chutzpah geisha
[sód] [ák] [bwán] [h´tsp] [géš]/[gíš]
4. broccoli apogee fricassee symphony pedigree
[brákli] [æ´ pi] [fr´ksi] [s´mfni] [pédgri]
shillelagh chianti bologna confetti Hatari
[šléli] [kiyánti] [blóni] [knf´i] [htári]
guarantee bourgeoisie chimpanzee Tennessee jamboree
[gærntí] [buržwazí] [mpænzí] [tnsí] [æmbrí]
runaway Saturday ricochet protégée workaday
[r´nwe] [sæ´ rde] [ríkše] [próte] [w´rkde]
matinée San José Chevrolet Santa Fé overplay
[mætné] [sænhozé] [švrlé] [sæn(t)fé] [ovrplé]

endevous residue peekaboo honeydew avenue
[rándevu] [r´zdu] [píkbu] [h´nidu] [æ´ vnu]
Timbuktu kangaroo misconstrue overthrew hitherto
[tmbktú] [kægrú] [msknstrú] [ovrrú] [hðrtú]
mistletoe Mexico embryo cameo Scorpio
[m´slto] [m´ksko] [´mbriyo] [kæ´ miyo] [sk´rpiyo]
fiæsko concerto soprano falsetto crescendo
[fiyæ´ sko] [kn´o] [spráno] [fls´o] [krš´ndo]
overflow (v) apropos status quo cheerio to-and-fro
[ovrfló] [æprpó] [stæskwó] [iriyó] [tunfró]
amplify rockaby occupy prophesy underlie
[æ´ mplfay] [rákbay] [ákypay] [práfsay] [´ndrlay]
subpoena agenda charisma pariah pagoda
[spín] [´nd] [kr´zm] [práy] [pgó]
5. apostrophe Menomini Penelope fraternity telegraphy
[pástrfi] [mnánmi] [pn´lpi] [frt´rni] [tl´grfi]
Ypsilanti kamikaze Cincinnati insincerely poison ivy
[pslæ´ nti] [kamkázi] [snsnæ´ i] [nsnsírli] [pyznáyvi]
pistachio Arapahoe embroglio adagio portfolio
[pstæ´ šiyo] [ræ´ pho] [mbróliyo] [dážiyo] [prtfóliyo]
armadillo Filipino virtuoso Sacramento lucky fellow
[armd´lo] [flpíno] [vruwózo] [sækrm´nto] [lkif´lo]
Arabia azalea Bulgaria et cetera America
[rébiy] [zéliy] [blgæ´ riy] [ts´r] [m´rk]
hullabaloo Kalamazoo Tippecanoe
[hlblú] [kælmzú] [tpknú]
145

146
APPENDIX B: PHONOLOGICAL RULES
1. a. All intervocalic voiceless stops become voiced.
b. Voiceless stops become voiced when intervocalic.
c.
d.
e.
2. a. Nontense vowels in the final syllable of a word are stressed when they are followed by
two or more consonants.
b. Nontense vowels become stressed when followed by two or more consonants in the final
syllable of a word.
c. [–TENSE] [+STRESS] / _____________ # C2 n
3. a. Delete all final consonants when the following word begins with a consonant.
b. Consonants become zero when final and when the following word begins with a
consonant.
c. C Ø / _____________ C #
4. a. Reduce all unstressed vowels to schwa.
b. Unstressed vowels become schwa.
c. [–STRESS] []

5. a. Palatalize all alveolar stops before [y].
b. Alveolar stops become palatal before [y].
c.
d.
e.
6. a. Devoice all final fricatives.
b. Fricatives become voiceless when final (before #).
c.
d.
e.
7. a. Aspirate all initial voiceless stops.
b. Voiceless stops become aspirate when initial (after #).
c.
147

148
d.
e.
8. a. Nasalize all vowels before a nasal consonant.
b. Vowels become nasal before nasal consonants.
c. V [+NASAL] / _______ [C, +NASAL]
9. a. A lax vowel in the antepenultimate syllable of a word is stressed.
b. Nontense vowels become stressed in antepenultimate syllables.
n n n
c. [–TENSE] [+STRESS] / _______ C 0 V C 0 V C 0 #
10. a. Palatalize all alveolar stops before [y] when they are preceded by a vowel or when they
are initial.
b. Alveolar stops become palatal before [y] when they are preceded by a vowel or by #.
c.
d.
e.

1. Data.
APPENDIX C: SLASH–DASH NOTATION
ROOT MEANING EXAMPLES
/skrb/ write scribble /skrb+l/ [skrbl]
script /skrb+t/ [skrpt]
/fræg/ break fragment /fræg+mnt/ [frægmnt]
refract /ri+fræg+t/ [rifrækt]
2. First Set of Rules.
a. [bt] [pt]
b. [gt] [kt]
3. Revision One.
4. Revision Two.
5. Revision Three.
[b] [p] / _________ [t]
[g] [k] / _________ [t]
149

150
APPENDIX D: PHONOLOGY PROBLEM
Data in phonetic transcription from the language Hip:
SINGULAR PLURAL MEANING
tip tiben girl
pit piden boy
kip kiben child
pik pigen police
tab taben teacher
bid biden dogs
gip giben car
dag dagen cat
tap _____ house
___ giden lions
1. What are the stop phonemes in Hip? Justify your answer with minimal pairs.
2. What is the phonemic form for children?
3. What is the phonetic form for houses?
4. What is the phonemic form for lions?
5. State the rules which accounts for the phonetic differences between the singular and the plural
forms where they occur?
6. What is the morpheme for PLURAL?
7. What is the morpheme for SINGULAR?

APPENDIX E: MORPHOLOGY PROBLEM
Consider the following phonetic transcriptions and translations of Blatin:
pelam I wash nores you work katim I jump
pelat he washes noret he works katis you jump
pelabas you washed norebam I worked katibam I jumped
pelabat he washed norebas you worked katibat he jumped
pelabim I will wash norebis you will work katibim I will jump
pelabis you will wash norebit he will work katibit he will jump
Identify the following morphemes:
first person (I) ______________________
second person (you) ______________________
third person (he) ______________________
present tense ______________________
past tense ______________________
future tense ______________________
What do the following mean:
pelas ______________________
norebim ______________________
How does one say the following in Blatin:
He will wash. ______________________
I work. ______________________
He jumps. _____________________
151

152
APPENDIX F: ENGLISH VOWEL SHIFT
1. e [–HIGH, –LOW, –BACK, +TENSE] [beð] [sen] [ksplen]
(bathe) (sane) (explain)
æ [–HIGH, +LOW, –BACK, –TENSE] [bæ] [sæni] [ksplæntori]
(bath) (sanity) (explanatory)
2. o [–HIGH, –LOW, +BACK, +TENSE] [noz] [fon] [tlskop]
(nose) (phone) (telescope)
a [–HIGH, +LOW, +BACK, +TENSE] [nastrl] [fank] [tlskapk]
(nostril) (phonic) (telescopic)
3. i [+HIGH, –LOW, –BACK, +TENSE] [slip] [mitr] [ridim]
(sleep) (meter) (redeem)
[–HIGH, –LOW, –BACK, –TENSE] [slpt] [mtrk] [ridmpšn]
(slept) (metric) (redemption)
4. u [+HIGH, –LOW, +BACK, +TENSE] [luz] [gus] [fud]
(lose) (goose) (food)
[–HIGH, –LOW, +BACK, –TENSE] [lst] [gsli] [fdr]
(lost) (gosling) (fodder)
**************
5. ay [–HIGH, +LOW, +BACK, +TENSE] [wayd] [layn] [gnayt]
(wide) (line) (ignite)
[+HIGH, –LOW, –BACK, –TENSE] [w] [lnir] [gšn]
(width) (linear) (ignition)
6. aw [–HIGH, +LOW, +BACK, +TENSE] [saw] [bawnd] [profawnd]
(south) (abound) (profound)
[–HIGH, +LOW, +BACK, –TENSE] [sðrn] [bndnt] [profndi]
(southern) (abundant) (profundity)

(1) Root /æg/ ‘do’:
APPENDIX G: ENGLISH MORPHOLOGY
/PHONOLOGICAL FORM/ [PHONETIC FORM] SPELLING
a. /æg + nt/ > [ent] agent
1. Vowel Shift: æ > e in certain contexts
sanity/sane; profanity/profane; explanatory/explain; gratitude/grateful
2. Velar Softening: g > j in certain contexts
rigor/rigid; regal/regicide; allegation/allege; analogous/analogize
b. /æg + nt + i/ > [ensi] agency
president/presidency; democrat/democracy; pirate/piracy; regent/regency
c. /æg + t/ > [ækt] act
regal/direct; fragment/refract; segment/bisect; pigment/depict
d. /æg + t + v/ > [æktv] active
e. /æg + t + v + ti/ > [æktvi] activity
f. /æg + t + v + ti + z/ > [æktviz] activities
g. /n + æg + t + v/ > [næktv] inactive
h. /æg + t + n/ > [ækšn] action
relate/relation; devote/devotion; omit/omission; complete/completion
I. /trænz + æg + t + n + æl/ > [trænzækšnl] transactional
j /ri + æg + t + r + z/ > [riæktrz] reactors
(2) Root /rg/ ‘rule’:
a. /rg + nt/ > [rint] regent
1. Vowel Shift: > i in certain contexts
serenity/serene; severity/severe; delicacy/delicious; metric/meter
2. Velar Softening: g > j in certain contexts
rigor/rigid; regal/regicide; allegation/allege; analogous/analogize
b. /rg + nt + i/ > [rinsi] regency
c. /di + rg + t/ > [drkt] direct
d. /di + rg + t + n + z/ > [drkšnz] directions
e. /n + rg + ul + ær + ti + z/ > [rgyulæriz] irregularities
regular/regulate; tabular/tabulate; circular/circulate; particular/particulate
153

154
(3) Root /fræg/ ‘break’:
a. /fræg + mnt/ > [frægmnt] fragment
b. /fræg + l/ > [fræl] fragile
c. /ri + fræg + t/ > [rifrækt] refract
(4) Root /lg/ ‘law’:
a. /lg + æl/ > [ligl] legal
b. /lg + s + let + n/ > [lslešn] legislation
c. /n + lg + tm + et/ > [lmt] illegitimate
(5) Root /lg/ ‘choose’:
a. /i+ lg + ebl/ > [lbl] eligible
b. /i+ lg + t + n/ > [ilkšn] election
c. /lg + nd + æri/ > [lndæri] legendary
(6) Root /sg/ ‘cut’:
a. /sg + mnt/ > [sgmnt] segment
b. /sg + t + n/ > [skšn] section
c. /dis + sg + t/ > [dayskt] dissect
(7) Root /pg/ ‘paint’:
a. /pg + mnt/ > [pgmnt] pigment
b. /di + pg + t/ > [dipkt] depict
c. /pg + t + or + i+ æl/ > [pktoril] pictorial
(8) Root /mæg/ ‘great, more’:
a. /mæg + sm + æl/ > [mæksml] maximal
b. /mæg + s + ter + i+ æl/ > [mæsteril] magisterial
c. /mæg + or/ > [mer] major
(9) Root /g/ ‘increase’:
a. /g + mnt/ > [gmnt] augment
b. /g + t + n/ > [kšn] auction
c. /n + g + ur + et/ > [ngyuret] inaugurate
(10) Root /pæg/ ‘fasten, agree’:
a. /pæg + nt/ > [pænt] pageant
b. /n + pæg + t/ > [mpækt] impact
c. /pro + pæg + et/ > [prapget] propagate

CHAPTER THREE: SYNTAX
3.1 LINGUISTIC ARGUMENTATION REVISITED
Linguistics, like all sciences, is concerned with explaining natural phenomena, that is, things which
occur naturally in the world and which people can perceive with their senses. The particular
phenomenon investigated in linguistics is language. Linguists look at all human languages that
naturally occur in the world and attempt to understand and describe their properties. One property
that linguists have discovered is that all natural languages consist of an infinite number of sentences
which are constructed in specific ways following specific rules. For example, all the sentences in
this paragraph are English sentences constructed correctly in terms of the rules of English grammar.
The area of linguistics that is specifically concerned with sentence structure is called SYNTAX.
It is possible to look at the structure of sentences and provide a variety of descriptions to account for
the way those sentences are formed. These various descriptions are actually hypotheses about how
sentences are constructed, and like all scientists, linguists must justify their hypotheses with
independent, objective evidence. Let us consider some examples of what such evidence looks like.
3.1.1 LINGUISTIC ARGUMENTATION: SENTENCE TYPES
In most English classes, teachers say that there are three types of complete sentences in English.
There are statements like She is careful, questions like Is she careful, and commands like Be careful,
as well as other examples like the following:
(1) a. Statements: She is careful. He will try to win. They don’t drive too fast.
b. Questions: Is she careful? Will he try to win? Don’t they drive too fast?
c. Commands: Be careful. Try to win. Don’t drive too fast.
English teachers also point out that all English sentences must have a subject, which is usually the
person or thing that performs the action in the verb. For example, the subjects for the statements and
questions in (1) are the italicized phrases in (2).
(2) a. Statements: She is careful. He will try to win. They don’t drive too fast.
b. Questions: Is she careful? Will he try to win? Don’t they drive too fast?
When we consider the commands in (1c), it appears that they have no subjects. Generally, English
teachers say that the subject of all commands is an understood you. In other words, (1c) is
understood as follows:
(3) Commands: (You) be careful. (You) try to win. (You) don’t drive too fast.
Actually, the statement, “the subject of all commands is an understood you,” is a hypothesis, and like
all hypotheses, it must be justified with evidence. When asked to justify this particular hypothesis

156
about commands, people often assert that commands are directed at the person (or persons) one is
talking to and that such a person is referred to with the pronoun you. The pronoun for the speaker,
called the FIRST PERSON, is I; the pronoun for the person spoken to, the addressee, called the
SECOND PERSON, is you. Since commands are addressed to the second person, the subject must
be the second person pronoun, you.
While this seems logical, it is not proof because it relies on the judgement and opinion of the person
making the assertion about the data. It is subjective. Since different people can view the same
phenomenon in different ways, what is needed as proof of the above description is independent,
objective evidence – evidence that is not based on the way in which a particular person views the
data.
To see how we can support a hypothesis with objective evidence, consider first the following
sentences paying particular attention to the italicized words which are called REFLEXIVE
PRONOUNS:
(4) a. I see myself.
b. The little girls dressed themselves.
c. Mary voted for herself.
Notice that the reflexive pronouns refer back to the subject, that is, in (4a) myself refers back to I,
in (4b) themselves refers back to the little girls, and in (4c) herself refers back to Mary. In fact, if
the reflexive pronoun does not agree with the subject, then the sentence is ungrammatical as follows:
(5) a. *I see yourself. (lack of agreement in person, first person versus second person)
b. *The girls dressed herself. (lack of agreement in number, singular versus plural)
c. *Mary voted for himself. (lack of agreement in gender, female versus male)
Given the above, we can make the following statements about reflexive pronouns in English.
(6) a. Reflexive pronouns occur after the main verb and refer back to the subject.
b. Reflexive pronouns must agree with the subject in person, number, and gender.
Returning to the issue of commands, notice that the only reflexives that are possible in a command
are the second person reflexive pronouns, yourself and yourselves:
(7) a. Dress yourself/yourselves.
b. *Dress myself/himself/herself/ourselves/themselves.
Since reflexive pronouns must agree with the subject in person (cf. (6b)) and since the only reflexive
pronouns possible in commands are second person, then the subject of a command must be second
person, that is, the addressee(s) or you. This argument constitutes independent, objective evidence
for the assertion, “the subject of all commands is an understood you.” Notice that all native speakers

of English unconsciously know the facts listed in (6) even if they have never heard the term reflexive
pronoun. We know that is correct because native speakers utter sentences like (4) and never utter
ones like (5).
Like other scientists, linguists look for more than one argument to support their hypotheses;
therefore, let us explore the matter further.
Consider next tag questions, in which the subject of the main sentence is repeated in a question
tagged onto a statement, such as the following:
(8) a. He didn’t do it, did he?
b. They will go, won’t they?
c. *She wasn’t there, was he?
d. *I’ll be invited, won’t they?
Notice that tag questions after a command must contain the second person pronoun you; hence, the
subject of a command must be an understood you.
(9) a. Dress yourself, won’t you (please)?
b. Don’t bother me, will you?
c. *Dress yourself, won’t he?
d. *Don’t bother me, will she?
As a third proof, consider idioms like the following and notice that the POSSESSIVE PRONOUNS
(in italics) must agree with the subject of the sentence; if they don’t the sentence is ungrammatical:
(10) a. She gave the children her full attention.
b. The man lost his temper.
c. *She gave the children my full attention.
d. *The man lost your temper.
Notice that idioms like the above can only contain second person possessive pronouns when they
occur in commands; hence, the subject of commands must be second person or you:
(11) a. Give the children your full attention.
b. Don’t lose your temper.
c. *Give the children my full attention.
d. *Don’t lose her temper.
The three arguments above provide justification for the hypothesis, “the subject of all commands is
an understood you.” Such justification is not derived from the opinion of the person making the
assertion, that is, it is not subjective. Instead, it is derived entirely from the objective observation
of facts about English sentences as they occur naturally in everyday speech. The arguments can be
157

158
verified or falsified through additional examination of other data; therefore, they constitute
independent evidence to support the hypothesis.
3.1.2 LINGUISTIC ARGUMENTATION: PARTS OF SPEECH
When trying to describe some aspect of syntactic structure, linguists always rely on independent
criteria to support their analyses. Let us consider here another example with the following sentences:
(12) a. John looks Italian.
b. John speaks Italian.
These two sentences appear to have the same structure, that is, they appear to be made up of the same
types of phrases with the same parts of speech (noun, verb, etc.). In fact, native speakers know that
the word Italian is used differently in the two sentences. We can prove that by examining other data
like the following.
(13) a. John looks very Italian.
b. *John speaks very Italian.
(14) a. *John looks fluent Italian.
b. John speaks fluent Italian.
Native speakers would agree that (13a) is grammatical, but (13b) is not, and that (14a) is
ungrammatical while (14b) is perfectly well-formed. These facts indicate the word Italian is used
differently in the two sentences. Specifically, in (12a) it functions as an adjective, while in (12b) it
functions as a noun. Confirmation of this hypothesis comes from examining other words whose
function is unquestioned (adjective or noun) and conjoining those words with Italian in the two
contexts under investigation (after looks and speaks):
(15) a. John looks tall, handsome, Jewish, and Italian (so he’s right for the part in the
movie).
b. John speaks Hebrew, Sanskrit, Italian, and several other languages.
(16) a. *John looks Hebrew, Sanskrit, Italian, and several other languages.
b. *John speaks tall, handsome, Jewish, and Italian.
A linguist must take great care when examining data. It is easy to be mislead, which is why linguists
search for as many arguments as they can. For example, consider the following:
(17) a. John looks more Italian (than Bill).
b. John speaks more Italian (than Bill).

Given these examples, it appears that Italian functions in the same way in the two sentences since
both sentences allow the comparative more. Further examination of other comparative-type
expressions reveals that, in fact, they are different as we have hypothesized:
(18) a. *John looks too much Italian.
b. John speaks too much Italian.
(19) a. John looks too Italian (to play that part in the movie).
b. *John speaks too Italian.
The additional data indicate that the grammaticality of both examples in (17) is only an apparent
falsification of our hypothesis that Italian functions differently in the two sentences in (12). There
must be some other fact which makes both of the examples in (17) good. In those cases, it is because
the comparative more can freely occur before both adjectives and nouns:
(20) a. John is more handsome, educated, courteous, etc. (than Bill).
b. John has more money, grandchildren, hair, etc. (than Bill).
3.1.3 LINGUISTIC ARGUMENTATION: PHRASES
Consider now one final example of linguistic argumentation using the following sentences:
(21) a. My niece hit the bus stop.
b. My niece heard the bus stop.
c. My niece saw the bus stop.
These sentences appear to involve the same succession of elements with only the verbs being
different (hit, heard, and saw):
(22) a. My niece - hit - the bus stop
b. My niece - heard - the bus stop
c. My niece - saw - the bus stop
Other data, however, indicate that the structure of these sentences is quite different. The difference
hinges on the way the word stop is analyzed. Notice that the word stop can be used either as a noun
(This bus makes too many stops) or a verb (The bus stopped at the corner). When stop functions as
a noun, other words can combine with it to form a COMPOUND, which is a phrase composed of
two or more elements that act like a single unit (bus stop, mail box, hair dryer, time machine, etc.).
When stop functions as a verb, it can be modified by an adverb (suddenly stop, quickly go, happily
sing, easily succeed, etc.). Thus, in (21a), the word stop is a noun and part of the compound bus
stop; in (21b), the word stop is a verb. Further, in (21c), either analysis is appropriate (noun or verb)
159

160
so the example is ambiguous. We can support this analysis of the sentences in (21) with data like
the following:
(23) a. *My niece hit the bus suddenly stop.
b. My niece heard the bus suddenly stop.
c. My niece saw the bus suddenly stop.
(24) a. My niece hit both the mail box and the bus stop.
b. *My niece heard both the mail box and the bus stop.
c. My niece saw both the mail box and the bus stop.
What the above sentences indicate is that, after the verb hit, the words bus and stop can form a
compound. Sentence (24a) is grammatical because the word and connects two compound nouns
(mail box and bus stop). On the other hand, sentence (23a) is ungrammatical because the word
suddenly separates the two elements of what is supposed to be a compound (cf. *mail probably box,
*hair thoroughly dryer, etc.).
After the verb heard in (23b), the words bus and stop do not form a compound and the word stop
is itself a verb. As a result, it can be modified by the adverb suddenly, and (23b) is grammatical.
On other hand, (24b) is ungrammatical under the ordinary interpretation of the verb hear and the
objects mail box and bus stop. Judgements of grammaticality are based on ordinary language in
ordinary circumstances, i.e., ones in which mail boxes and bus stops do not talk or make noise by
themselves though that is certainly imaginable. Significantly, (24b) only makes sense under such
special circumstances.
The example in (23c) is unambiguous allowing only the reading in which stop is a verb. As
expected, (24c) is also unambiguous having only the opposite reading that involves conjoining two
compound nouns (mail box and bus stop). Other data that support our analysis include the following:
(25) a. The bus stop was hit by my niece.
c. *The bus stop was heard by my niece. (Bus stops do not ordinarily make noise.)
c. The bus stop was seen by my niece. (unambiguous; stop must be a noun)
(26) a. *It was the bus that my niece hit stop. (Compound elements cannot be separated.)
b. It was the bus that my niece heard stop.
c. It was the bus that my niece saw stop. (unambiguous; stop must be a verb)
(27) a. *My niece hit the busses stop. (The first element in a compound cannot be plural.)
b. My niece heard the busses stop.
c. My niece saw the busses stop. (unambiguous; stop must be a verb)
All these examples add independent and objective support for our analysis of the phrase structure
of the sentences in (21).

3.1.4 LINGUISTIC ARGUMENTATION: STRATEGIES
The three examples of linguistic argumentation given here should be viewed as a model for how to
proceed in syntactic analysis. With these examples, we can formulate four strategies that can be used
not only for linguistic analysis, but also for the analysis of any problem. Here is a summary of these
four strategies.
(28) Strategy One: Proceed in a stepwise fashion.
It is often useful to scan data for patterns and regularities. If nothing appears to fall into place,
try to find some commonality in a few examples in the data and proceed from that starting
point to gradually incorporate other examples. Separate examples into mutually exclusive
classes defined by the same types of criteria, so that each of your examples can only fall into
one class. If some examples or some data don’t fit into the criteria, put them aside. It is better
to leave some examples out of your analysis than to fit them in by relaxing your criteria.
(29) Strategy Two: Search for generalizations.
An analysis says very little if it does not lead to generalizations; therefore, your description of
the data should result in a statement of some general principles. As above, it is better to leave
some elements out of an analysis if their inclusion requires giving up a generalization or
violating a well-established principle. Perhaps they are misunderstood cases or exceptions.
An analysis that describes most of the data with interesting, contrastive generalizations is
better than one that includes all the examples but says little of interest. A proposal that covers
most concerns with well-supported data is better than one which tries to include everything
by adding extra conditions and special cases.
(30) Strategy Three: Make your hypotheses and your proposals data-driven.
Your hypotheses and your proposals should arise from an examination of the data. Always
begin with the examples and show how they drive your descriptive apparatus, terminology,
and representations. Do not propose a hypothesis unless there are data to support it. Once you
make a proposal make sure your representations (symbols, discussion, etc.) always follow it.
The more you show how that data drive your proposals, the more robust your proposals will
be, so much so that your proposals become inevitable.
(31) Strategy Four: When presenting your hypothesis or making your proposal, consider your
audience.
You should choose your examples carefully, be consistent, and remove extraneous material
from your description. You should only mention what is essential to make your point, and
don’t introduce unnecessary and potentially distracting complexities. Your proposal should
be presented in such a way that it can be easily followed and understood. Presentation matters.
161

162
3.2 ORDERING CONSTRAINTS
With the above discussion about the nature of linguistic argumentation in mind, let us now turn to
the focus of this chapter, describing the structure of English sentences and developing a theory of
syntax. We begin by noting that sentences are not random strings of words. Rather, the words of
every sentence are arranged into units called PHRASES. Further, there is both a left to right
constraint on the order of phrases, called LINEAR ORDER, and a top to bottom constraint on the
order of phrases, called HIERARCHICAL ORDER. Learning English consists in part of
mastering these ordering constraints. Failure to observe the constraints on linear order can lead to
ungrammaticality at any time during language acquisition. In adult language, (32b)is ungrammatical;
in child language, (32d) is ungrammatical for the meaning given.
(32) a. That is daddy’s chair.
b. *That is daddy chair.
c. Daddy chair. (= ‘That is daddy’s chair.’)
d. *Chair daddy. (= ‘That is daddy’s chair.’)
In addition, failure to observe linear order can lead to meaning changes in both adult language
(33a/b) and child language (33c/d).
(33) a. John found Bill.
b. Bill found John.
c. Mommy find. ( = ‘Mommy come and find me; I’m hiding.’)
d. Find mommy. ( = ‘I will find mommy.’)
Changes in hierarchical order often result in ambiguity, that is, multiple meanings. For example, the
phrase the Spanish history books can be interpreted like the Spanish math books or like the art
history books, as follows:
(34) a. John has the Spanish history books. (= the history books in Spanish)
John has the Spanish math books.

. John has the Spanish history books. (= the books on Spanish history)
John has the art history books.
As we noted above, SYNTAX is the area of linguistics that is specifically concerned with sentence
structure, including the linear and hierarchical ordering of phrases. Formally, a PHRASE is a word
or group of words that functions as a unit, which means that the language treats it as a whole.
Phrases are identified by one or more of three syntactic processes mentioned in the Chapter One and
reviewed here.
3.2.1 TESTING HYPOTHESES: REFERENCE, OMISSION, AND PLACEMENT
In this section, we will review the three basic syntactic processes used to identify individuals phrases
in sentences: reference, omission, and placement.
REFERENCE is a binding relationship between phrases. Some phrases get their meaning only
from identification with other phrases. For example, in John hurt himself, the reflexive pronoun
himself gets its meaning from its identification with John. We say himself refers to or is bound to
John, and the relationship between the two is an instance of reference. Reference occurs in simple
sentences like (35), in complex sentences like (36), and between sentences like (37) and (38).
(35) a. The painter cut himself. (himself = the painters)
b. The painters painted each other. (each other = the painters)
(36) a. The portrait of himself that Rembrandt liked the most is the one he painted last.
(himself = Rembrandt; he = Rembrandt)
b. After the children finished dinner, they brushed their teeth.
(they = the children; their = the children)
(37) a. Impressionist painters seem to use more blue than expressionist painters.
b. At least, I think they do. (they =impressionist painters; do = use more blue.)
(38) a. The children should not be playing with matches.
b. Why did you give them them?
(The first them refers to the children; the second them refers to matches.)
163

164
OMISSION is the deletion of a phrase. Repeated occurrences of phrases can be left out. In such
cases, the omitted phrase is understood to be identical to the given phrase. Consider the following
examples where the deleted units are in italics:
(39) a. Alex will jump on the trampoline and his friends will jump on the trampoline too.
b. Alex will jump on the trampoline and his friends will too.
(40) a. Olivia has many toys and Liliana has many toys.
b. Olivia has as many toys as Liliana has.
(41) a. Bennett will clean his room if his father asks him to clean his room.
b. Bennett will clean his room if his father asks him to.
PLACEMENT is the occurrence of a phrase in a particular position. Consider the varying positions
in which constituents can be placed in the following examples:
(42) a. Fletcher often smiles at his father.
b. Fletcher smiles at his father often.
(43) a. Fiona drank up all her milk.
b. Fiona drank all her milk up.
Note that, in the above, the word often can occur on either side of the verb phrase smiles at this
father, which indicates that smiles at his father is a unit. Also, the preposition up can occur on either
side of the direct object all her milk, which indicates that all her milk is a unit.
These three processes can be used to determine the units in any English sentence; however, they do
not apply equally easily in all cases. Students should not expect the testing process to run smoothly
all the time. To see the difficulties, we will use the following sentence as our target sentence:
(44) The company handed out the dividends to the stockholders last week.
Using REFERENCE, we can identify the major units in a rather straightforward way. Basic units
can be replaced easily with question words (who, what, when, where, etc.) and pronouns (it, they,
them, etc.). Consider the following:
(45) a. Who handed out the dividends to the stockholders last week?
b. The company.
c. It handed out the dividends to the stockholders last week.
(46) a. What did the company handed out to the stockholders last week?
b. The dividends.
c. The company handed them out to the stockholders last week.

(47) a. Who did the company handed out the dividends to last week?
b. The stockholders.
c. The company handed out the dividends to them last week.
(48) a. When did the company hand out the dividends to the stockholders?
b. Last week.
c. The company handed out the dividends to the stockholders then.
(49) a. What did the company do last week?
b. It handed out the dividends to the stockholders.
c. The company did it.
OMISSION, on the other hand, is much more tightly constrained given the nature of English syntax.
The units most easily omitted are verbal units in examples like the following:
(50) a. The company handed out the dividends to the stockholders last week because the
board of directors told it to hand out the dividends to the stockholders last week.
b. The company handed out the dividends to the stockholders last week because the
board of directors told it to.
(51) a. This company handed out the dividends to the stockholders last week, and that
company handed out the dividends to the stockholders this week.
b. This company handed out the dividends to the stockholders last week, and that
company this week.
In the sections below, we will see further uses of omission in other units, especially noun phrases.
Turning to PLACEMENT, we find very few elements than can be permuted easily among all
sentential units. One such element is the word only:
(52) a. Only the company handed out the dividends to the stockholders last week.
b. The company handed out only the dividends to the stockholders last week.
c. The company handed out the dividends only to the stockholders last week.
d. The company handed out the dividends to the stockholders only last week.
Notice that only is placed at the beginning of each individual unit in the above sentences.
In some cases, different units can be placed at the beginning of the sentence for emphasis, but with
varying degrees of acceptance by speakers:
(53) a. ?The dividends, the company handed out to the stockholders last week.
b. To the stockholders, the company handed out the dividends last week.
c. Last week, the company handed out the dividends to the stockholders.
165

166
3.2.2 THREE OTHER DIAGNOSTICS TO DETERMINE STRUCTURE
In this section, we will look at three other diagnostic tests that linguists use to verify phrase structure
again using as our target sentence the example The company handed out the dividends to the
stockholders last week.
(54) THE CLEFT SENTENCE TEST: English has a type of sentence called the cleft sentence
with the form “It was X that Y.” The item that fills the X slot is considered a unit.
a. It was the company that handed out the dividends to the stockholders last week.
b. It was the dividends that the company handed out to the stockholders last week.
c. It was to the stockholders that the company handed out the dividends last week.
d. It was the stockholders that the company handed out the dividends to last week.
e. It was last week that the company handed out the dividends to the stockholders.
f. *It was out the dividends that the company handed to the stockholders last week.
g. *It was stockholders last that the company handed out the dividends to week.
Applying the cleft sentence test to the target sentence produces some sentences that are grammatical,
which indicates that the clefted elements (in italics above) are units. Attempts to make other
sequences of elements units, such as out the dividends (54f) and stockholders last (54g), produces
an ungrammatical sentence, which indicates that those sequences are not units.
(55) THE PSEUDO-CLEFT SENTENCE TEST: English has another type of sentence called
the pseudo–cleft sentence with various forms beginning with a WH-phrase and then
following with a pattern very similar to a cleft sentence.
a. What the company did was hand out the dividends to the stockholders last week.
b. What the company handed out to the stockholders last week were the dividends.
c. Who the company handed out the dividends to last week were the stockholders.
d. When the company handed out the dividends to the stockholders was last week.
In the examples in (55), the putative units are the ones in italics that follow the forms of the verb be
(was and were). The form of the pseudo-cleft test is “WH–phrase X be Y,” where the sequence in
the Y slot is identified as a unit.
(56) THE PARAPHRASE TEST: Sentences can often be paraphrased in various ways that
help to identify their structure.
a. The company handed the dividends out to the stockholders last week.
b. The stockholders were handed out the dividends by the company last week.
c. The dividends were handed out to the stockholders by the company last week.
d. Last week, the company handed out the dividends to the stockholders.
e. The company handed out the dividends last week to the stockholders

f. *The company handed out the dividends to the last week stockholders.
g. *The company handed the dividends to out the stockholders last week.
h. *The stockholders last week, the company handed out the dividends to.
As always, one must exercise great caution using these tests. It is easy to be mislead. Therefore, it
is best to try as many different tests as possible to determine phrase structure. The paraphrase test,
in particular, must be applied in such a way that the words remain essentially unchanged altering
only things like the number of a noun or the tense of a verb. Thus, in (56a), the preposition out has
merely been repositioned; in (56b) and (56c), the verb has been put in the passive voice; and, in
(56d) and (56e), the phrase last week has been repositioned. In (56f), the phrase last week has broken
up the unit the stockholders; in (56g), the preposition out has broken up the unit to the stockholders;
thus, both are ungrammatical. Similarly, (56h), which attempts to make a single unit out of the
stockholders last week, is ungrammatical.
In using any of the above tests, one should vary as little as possible and perform only one
manipulation at a time. Scientists always alter only variable at a time, so they know which variable
is the determining factor.
The diagnostic tests given in this section and the last one show that all sentences are organized into
discrete units or phrases. As we will see, the sentences of every human language contain phrases.
Furthermore, as we saw in the Introduction and Chapter One, phrase structure is a feature of the
earliest sentences used by children regardless of the language they are learning. The exact character
of phrases varies from language to language, but there is no known language which lacks phrase
structure. Therefore, phrase structure cannot be an arbitrary convention; it must be related in some
way to the nature of human beings, that is, it must be rooted in human biology.
3.2.3 LANGUAGE AND MEMORY
The biological origins of phrase structure become clear when one examines the nature of human
memory. It is a well–known fact that human beings have very limited SHORT TERM MEMORY
(STM), which is the immediate memory for new information, such as a telephone number just
received from the operator. STM contrasts with our virtually limitless LONG TERM MEMORY
(LTM), the storage and recall of information that is not new, such as one’s mother’s maiden name.
Generally, STM is thought to be limited to about seven (±2) bits of information. STM can be
facilitated by organizing information into units called CHUNKS. For example, the string of
numbers in (57a) is much more readily recalled when given as (57b).
(57) a. 1–4–9–2–2–0–0–1–1–7–7–6–1–9–9–7
b. 1492 – 2001 – 1776 – 1997
167

168
Given the reality of the chunk in psychological studies and the phrase in linguistic studies, it seems
plausible to derive (58).
(58) The sentences of a natural language consist of phrases because human STM requires
organizing a string of words into chunks to facilitate processing.
In short, a phrase is a chunk of a sentence. Syntax is the study of phrases. A linguist specifies phrase
structure by examining the units that participate in , omission, and placement. This methodology
is theory–independent; in fact, this is the method by which linguists develop theories of grammar.
We can see this if we examine the development of grammatical theory in this century, specifically,
the development from traditional grammar to structural grammar to generative grammar.
3.3 TRADITIONAL AND STRUCTURAL GRAMMAR
The oldest grammatical model in western civilization is TRADITIONAL GRAMMAR, a
grammatical model that has evolved from the language studies of the Greeks and Romans. Such
grammars contain definitions like (59).
(59) A SENTENCE is a structured expression consisting minimally of a SUBJECT and a
PREDICATE. The subject specifies the topic of the sentence, and the predicate specifies
what is asserted of the topic. Predicates always contain a VERB, which is a word that
serves to express an action, occurrence or state of being; sometimes predicates also contain
an OBJECT, which specifies the person or thing directly affected by the action defined in
the verb. Since they must at least contain a verb, predicates are also referred to as VERB
PHRASES. Subjects and objects are phrases containing a NOUN, which is the name of
a person, place, or thing, and a DETERMINER (DET), which is a word like the, this, or
that.
According to these definitions, a sentence like (60a) might be graphically represented by the TREE
DIAGRAM (60b).
(60) a. The passengers obey the stewardess.
b.

While (60b) appears to be a fairly adequate representation of the phrases in (60a), there are problems.
The first concerns an auxiliary or helping verb like must when it appears in a similar sentence like
The passengers must obey the stewardess. (59) does not tell us what to do with it, in particular, it
does not tell us whether an auxiliary is part of the SUBJECT node or the SENTENCE node or the
VERB PHRASE node. This is important because we would like to be able to devise an explicit set
of rules that describe the phrase structure of all English sentences. Since the left to right order of the
elements in a sentence must be considered, we have three possibilities for connecting an auxiliary
(AUX) with the rest of the sentences. We can represent the choices as follows, where the dotted
lines indicate the possible attachments of AUX:
(61) a. The passengers must obey the stewardess.
b.
If AUX is part of the SUBJECT, then the tree diagram for an English sentence is (62a), and the
PHRASE–STRUCTURE RULES (PS–RULES) that describe English sentences are (62b), where
the arrow can be interpreted as ‘consists of.’
(62) a.
b. (1) SENTENCE SUBJECT + VERB PHRASE
(2) SUBJECT DET + NOUN + AUX
(3) VERB PHRASE VERB + OBJECT
(4) OBJECT DET + NOUN
169

170
If AUX is part of the VERB PHRASE, then the tree diagram for an English sentence is (63a), and
the PS–RULES that describe English sentences are (63b).
(63) a.
b. (1) SENTENCE SUBJECT + VERB PHRASE
(2) SUBJECT DET + NOUN
(3) VERB PHRASE AUX + VERB + OBJECT
(4) OBJECT DET + NOUN
Lastly, if AUX is part of the SENTENCE, then the tree diagram for an English sentence is (64a), and
the PS–RULES that describe English sentences are (64b).
(64) a.
b. (1) SENTENCE SUBJECT + AUX + VERB PHRASE
(2) SUBJECT DET + NOUN
(3) VERB PHRASE VERB + OBJECT
(4) OBJECT DET + NOUN
Traditional grammarians were not concerned much with the objective justification of the statements
they made about grammar. The choice between analyses like the above was often, quite simply, a
matter of opinion. All modern theories of grammar take a very different point of view. All modern
theories of linguistics (those of the twentieth century) attempt to ground their analyses in terms of
structural principles which can be justified, that is, proven. The further into the twentieth century

one gets in the development of grammatical theory, the more stringent the requirement for proof
becomes.
To determine which of the above sets of rules is correct, we can use the processes of reference,
omission, and placement. Consider first (62), the hypothesis that AUX is part of SUBJECT. A
well–known fact about English is that subjects and objects can be replaced with pronouns like they,
them, she, her, etc. The use of pronouns is very common in language: by replacing a phrase with a
pronoun we do not have to keep repeating the whole phrase. Instead of saying The stewardess told
the passengers that the passengers had to obey the stewardess, we say The stewardess told the
passengers that they had to obey her. Rules of reference tell us what they and her refer to.
Furthermore, specific pronouns must be used in specific positions: subjects must be replaced with
pronouns like I, he, she, we, they, etc., and objects must be replaced with pronouns like me, him, her,
us, them, etc. Thus, alongside of The passengers must obey the stewardess, we have sentences like
(65), but none like (66).
(65) a. They must obey her.
b. She must obey them.
(66) a. *Them must obey she.
b. *Her must obey they.
Following (62), we produce a fully grammatical sentence when we attempt to substitute the
SUBJECT node, which contains the AUX, and the OBJECT node with the appropriate pronouns:
(67) a. [The passengers must] obey [the stewardess].
b. [They] obey [her].
However, if the nouns in SUBJECT and OBJECT position of (62) are reversed as in (68a), then
substitution of SUBJECT and OBJECT with pronouns produces the ungrammatical (68b).
(68) a. [The stewardess must] obey [the passengers].
b. *[She] obey [them]. (cf. She obeys them).
The ungrammaticality of (68b) tells us that the analysis in (62) is wrong. If the rules were correct,
then we would not get an ungrammatical sentence when we applied them in various circumstances.
We therefore must reject (62). Intuitively, (62) seems incorrect for another reason. Subjects and
objects appear to be the same kinds of phrase; yet, (62) claims that AUX is part of SUBJECT.
Clearly, AUX can never be part of OBJECT: *They obey the stewardess must. If subjects and
objects are the same kinds of phrase, we ought to be able to get AUX in either. As a result, we must
reject (62) which means that our alternatives are (63) or (64). To help us choose between (63) and
(64), let us consider the following data:
171

172
(69) a. Generally the passengers must obey the stewardess.
b. *The generally passengers must obey the stewardess.
c. The passengers generally must obey the stewardess.
d. The passengers must generally obey the stewardess.
e. *The passengers must obey generally the stewardess.
f. *The passengers must obey the generally stewardess.
g. The passengers must obey the stewardess generally.
(70) a. *Strictly the passengers must obey the stewardess.
b. *The strictly passengers must obey the stewardess.
c. *The passengers strictly must obey the stewardess.
d. The passengers must strictly obey the stewardess.
e. *The passengers must obey strictly the stewardess.
f. *The passengers must obey the strictly stewardess.
g. The passengers must obey the stewardess strictly.
These data indicate that there are two broad classes of adverbs in English: SENTENCE ADVERBS
like generally that modify the whole sentence, and MANNER ADVERBS like strictly that modify
the verb phrase. Only the diagram in (64) and the rules in (64b) capture this classification, that is,
we arrive at a very efficient characterization of the two classes of adverbs with the following logic:
(71) a. If the rules in (64b) specify sentence structure, then we can say that sentence adverbs
occur around the three major sentential elements, namely, the four positions indicated
by the pound sign as follows:
SENTENCE # SUBJECT # AUXILIARY # VERB PHRASE #
b. If the rules in (64b) specify verb phrase structure, then we can say that manner
adverbs occur either at the beginning or the end of the verb phrase, namely, the two
positions indicated by the pound sign as follows:
VERB PHRASE # VERB + OBJECT #
The diagram and rules in (63) do not lead to such an efficient characterization. In short, the
PLACEMENT of adverbs proves what the units in sentences are. Furthermore, these rules predict
that (72a) is grammatical, and that (72b) is not.
(72) a. The passengers must generally strictly obey the stewardess.
b. *The passengers must strictly generally obey the stewardess.
There are sentences that corroborate the choice of (64). Notice that the VERB PHRASE node in
(64a) is substantiated by its possible OMISSION in a sentence like The passengers must obey the
stewardess if the pilot orders them to (obey the stewardess). If the VERB PHRASE included AUX,

as in (63), then repetition of the VERB PHRASE would produce *The passengers must obey the
stewardess if the pilot orders them to must obey the stewardess. Again, following the rules logically
we see that they are confirmed or rejected by the sentences that result, specifically whether the rules
lead to other sentences that are grammatical or ungrammatical.
As final confirmation of (64), consider the use of do it in a sentence like The pilot told the
passengers that they must obey the stewardess and actually had no trouble convincing them to do
it (do it = obey the stewardess). Again, if the AUX were part of the VERB PHRASE, then do it
would have to mean must obey the stewardess not obey the stewardess. Indeed, there is no
pronominal form like do it in English that stands for a VERB PHRASE that contains an
AUXILIARY. If one existed, we would have evidence supporting (63).
As a result of these considerations, we prove that English sentences contain three basic units;
therefore, the analysis in (64) is justified. Stated this way, a grammatical description is not a matter
of opinion. It is a matter of fact. Furthermore, we attain a very high level of adequacy by basing our
description on the concept of phrase, since phrase structure is the result of chunking, and chunking
is the result of biological constraints on human memory. We have a direct connection between the
nature of man and the nature of grammar.
As we saw in the Chapter One, there are three levels of adequacy that a linguist can attain in
grammatical analysis. The first level, called OBSERVATIONAL ADEQUACY, is attained when
the facts are noted. The second level, called DESCRIPTIVE ADEQUACY, is attained when the
facts are described with generalized principles or rules, e.g., (64). The third and highest level, called
EXPLANATORY ADEQUACY, is attained when those principles are related to the nature of the
language apparatus of man. As a linguist moves from observational to explanatory adequacy, an
ever higher level of generality is reached. At the highest level, linguists begin to relate individual
languages like English to human language in general and, therefore, to offer a real explanation for
linguistic phenomena. In attempting to attain explanatory adequacy, a linguist is also able to separate
out what is arbitrary and language specific, from what is predictable in terms of the nature of the
human language apparatus (the brain and the vocal and auditory organs).
Over a period of many years, traditional grammar gave way to STRUCTURAL GRAMMAR
because the latter could provide more descriptively adequate characterizations of natural language.
Traditional definitions such as those in (59) have been replaced by PS–RULES like those in (64b).
Thus far, we have justified a model of English which contains tree structures like (64a). However,
there are improvements that can be made to (64) that will allow us to attain a higher level of
descriptive adequacy. Notice that any word or group of words that can be a subject of an English
sentence can also be an object (with a meaning change, of course):
(73) a. The handsome prince rescued the beautiful princess.
b. The beautiful princess rescued the handsome prince.
173

174
(74) a. The new tenor from Italy met the members of the orchestra.
b. The members of the orchestra met the new tenor from Italy.
A categories like VP (Verb Phrase) or NP (Noun Phrase) is a STRUCTURAL CATEGORY. It
abbreviates a class of syntactic structures that can occur in certain positions in a syntactic
HIERARCHY. On the other hand, a category like subject or object is a FUNCTIONAL
CATEGORY, which names what function or role a particular structure plays in the sentence. Thus,
in (74a) and (74b), the new tenor from Italy is structurally a noun phrase (NP), consisting of the noun
tenor preceded by the determiner the and the adjective new, and followed by the prepositional phrase
from Italy; functionally, this NP is the subject of (74a) and the object of (74b).
This additional information about English syntax, and more specifically about the sentence in (60a),
can be expressed if the tree diagram in (60b) is replaced with the one in (75), an entirely structural
representation (S = Sentence; NP = Noun Phrase; VP = Verb Phrase; N = Noun; V = Verb):
(75)
3.4 TENSE AND THE INTERNAL STRUCTURE OF SENTENCES
The representation in (75) describes the structure of the sentence The passengers must obey the
stewardess. Naturally, our rules must do more than describe the structure of just one sentence since
there is an infinite number of sentences in every language. Also, some VP do not contain direct
objects and some NP do not contain determiners:
(76) a. The passengers can deplane.
b. Passengers must obey stewardesses.
These two facts can easily be accounted for by stipulating that direct object NPs and determiners are
optional. Borrowing the parenthesis notation we used in phonology to indicate optional elements,
we revise the PS-Rules as follows (cf. (64b)):

(77) a. S NP + AUX + VP
b. VP V + (NP)
c. NP (DET) + N
Other facts about English cannot be so easily accommodated. Consider the following sentences
which appear to contain no AUX:
(78) a. The passenger obeys the stewardess.
b. The passenger obeyed the stewardess.
We might say that such sentences indicate that the AUX should be optional in (77a); however, this
fails to account for the fact that the sentences in (78) have a tense. Example (78a) is in the present
tense (PRS); (78b) is in the past tense (PST). Furthermore, there can be only one tense in each
sentence:
(79) a. The passenger does obey the stewardess.
b. The passenger did obey the stewardess.
c. *The passenger does obeys the stewardess.
d. *The passenger did obeyed the stewardess.
e. *The passenger does obeyed the stewardess.
f. *The passenger did obeys the stewardess.
Summarizing, it is clear that all native speakers of English know the following facts about tense and
the internal structure of sentences:
(80) a. Every sentence (S) must contain a tense.
b. There is only one tense in each S.
c. If there is no auxiliary verb, then tense is marked on the main verb (cf. (78)).
d. If there is an auxiliary verb, then tense is marked on the auxiliary verb (cf. does in
(79a) and did in (79b)).
Of course, the knowledge that native speakers have of these facts is unconscious. That much is clear
because they observe the facts when they produce new sentences, that is, they never say things like
(79c) - (79f). The facts are part of the grammar of English that native speakers have internalized.
The facts exist whether or not there are people around who want to write the facts down as we are
trying to do. When we write down the grammatical facts, all we are doing is bringing those facts to
consciousness. We are not making them up.
175

176
Additional facts about tense that native speakers know are the following:
(81) a. Two tense markers are signaled by a change in the shape of the verb: present tense
(PRS) and past tense (PST).
[1] The passenger obeys the stewardess.
[2] The passenger obeyed the stewardess.
[3] The passenger does obey the stewardess.
[4] The passenger did obey the stewardess.
b. One tense marker is signaled by the bare verb stem alone: the imperative tense
(IMP), which is used to express commands.
[1] Obey the stewardess.
[2] Do (not) obey the stewardess.
c. Two other tense markers are signaled by preceding the main verb with an auxiliary
verb called a MODAL: future tense (FUT) and conditional tense (CND).
[1] The passenger will/shall obey the stewardess.
[2] The passenger could/should obey the stewardess.
Since a tense is obligatory and a modal is optional, we should replace (77) with the following where
TP stands for Tense Phrase:
(82) a. S NP + TP + VP
b. VP V + (NP)
c. NP (DET) + N
d. TP (MOD) + T
e. T {PRS, PST, IMP, FUT, CND}
In PS–Rules like the above, the PARENTHESES indicate OPTIONAL ELEMENTS and the
BRACES (CURLY BRACKETS) indicate ALTERNATIVE ELEMENTS. These notations, of
course, are the same ones we saw above in Chapter Two in P-Rules (Phonological Rules).
The symbol T stands for TENSE, which is used here as a special abbreviation for the various TIME
and MODE/MOOD distinctions in English. The five tenses in English include the three TIME

distinctions, PRESENT (PRS), PAST (PST), and FUTURE (FUT), and the two MODE/MOOD
distinctions, CONDITIONAL (CND), and IMPERATIVE (IMP). We have examples like the
following:
(83) a. She lives in a dorm. (lives is present tense, PRS)
b. She lived in a dorm. (lived is past tense, PST)
c. She will live in a dorm . (will is future tense, FUT)
d. She should live in a dorm. (should is conditional mode, CND)
e. Live in a dorm. (live is imperative mode, IMP)
(84) a. She goes to college. (goes is present tense, PRS)
b. She went to college. (went is past tense, PST)
c. She will go to college. (will is future tense, FUT)
d. She should go to college. (should is conditional mode, CND)
e. Go to college. (go is imperative mode, IMP)
Notice that some of these distinctions are not realized as individual words, but as inflectional
endings, that is, changes in the shape of the verb (lives, lived, etc.). Further, in English, inflections
aren’t always an identifiable element at the end of a word: they can appear as modifications of the
root or as nothing at all. For example, went is the past tense of go; will is the modal will in future
time; should is the modal shall in the conditional mode.
In English, a MODAL (MOD) is a word that colors the meaning of the main verb and expresses
notions like possibility (can, may), necessity (must, should), potentiality (might, could), likelihood
(may), obligation (must, should), and so on. The most common modals are can, could, shall, should,
will, would, may, might, must, do, does, did, etc. These words are actually composite forms, e.g.,
does is PRS of do, did is PST of do, etc.
Here is a summary and some other examples of regular and irregular verbs in various tenses:
(85) PRS Singular PRS Plural PST Singular PST Plural FUT
a. obeys obey obeyed obeyed will obey
b. lives live lived lived will live
c. goes go went went will go
d. becomes become became became will become
e. writes write wrote wrote will write
f. dives dive dove/dived dove/dived will dive
g. is are was were will be
h. does do did did will do
i. has have had had will have
177

178
3.5 DETERMINERS AND THE INTERNAL STRUCTURE OF NOUN PHRASES
Continuing our investigation of the phrase structure of English, let us now consider the internal
structure of noun phrases like the following (the NPs are in italics):
(86) a. Passengers obey stewardesses.
b. Those passengers obey these stewardesses.
c. All the passengers obey all the stewardesses.
d. All the nervous passengers obey the stewardesses.
(87) a. *Those all passengers obey the stewardesses.
b. *Nervous the passengers obey the stewardesses.
Again, it is clear that native speakers unconsciously know a variety of facts about the internal
structure of NP because they will utter sentences like (86), but not ones like (87). Among other
facts, native speakers know that the left to right order of the elements in an NP must be as follows:
(88) NP (QNT) + (DET) + (ADJ) + N
In this rule, QNT stands for quantifier (a word like all, both, or each ), DET stands for determiner
(a word like the, this/these, or a/an), and ADJ stands for adjective (a word like nervous, tall, or
happy). Native speakers know which category (QNT, DET or ADJ) each word belongs to because
they always put them in the proper order which, again, is why they utter sentences like (86) but not
ones like (87). They also know exactly how the categories differ from each other, for example, that
only adjectives can be compared:
(89) a. tall, taller tallest
b. nervous, more nervous, most nervous
c. all, *aller, *allest
d. these, *theser, *thesest
In addition to the above, native speakers also know that there are two broad categories of nouns in
English: PROPER NOUNS, which refer to a specific entity (Heather, Daisy, Mary, Michigan,
Oakland University, etc.), and COMMON NOUNS, which are general in their reference (heather,
daisy, girl, state, education, university, etc.).
The major distinction between the two categories of nouns is that proper nouns cannot be specified
by a determiner (*this Michigan, *that Oakland University, etc.), whereas common nouns can (this
state, that university, etc.). Further, an overt determiner is obligatory for singular COUNT NOUNS,
which are nouns that can be counted (one daisy, two daisies, many daisies, etc.), but optional for
plural count nouns and for noncount nouns or MASS NOUNS, which cannot be counted (*one
heather, *two heathers, *many heathers, etc.). Here is a summary of the relevant facts (notice the
capital letters on the proper nouns):

(90) Noun Class Overt Determiner No Overt Determiner
a. Proper: *I like the Heather and the Daisy. I like Heather and Daisy.
b. Singular Count: I like the daisy. *I like daisy.
c. Plural Count: I like the daisies. I like daisies.
d. Noncount (Mass): I like the heather. I like heather.
As we can see above, proper nouns are usually capitalized in writing; however, that “clue” is missing
in speech. Still, the entire classification of nouns given here is unconsciously known by all native
speakers, which is why they only use determiners when they should in speech. Remember that young
children do not have access to things like the rules for capitalization; they speak the language long
before they learn to read and write. The unconscious knowledge they acquire is our focus. Our
grammar must account for that knowledge to be descriptively adequate.
3.5.1 ABSTRACT ELEMENTS: Ø
As the above description indicates, the use of determiners in noun phrases is complex and subtle.
People learning English as a second language (ESL) often make mistakes in using determiners,
putting them in where they shouldn’t occur and leaving them out where they should occur. This is
especially true of people whose native language has no obligatory determiners, such as Japanese.
Adult native speakers of English never make such mistakes. They don’t say things like *I saw child
or *I live in the Michigan. Importantly, the reason that native speakers do not say such things is not
because they have never heard anyone say such things. The number of sentence is infinite so a
speaker could not have heard everything. The reason that native speakers use determiners correctly
is because they have unconsciously internalized the facts about determiners described above. They
unconsciously know the rules. They have not memorized all the good sentences.
Students new to linguistics are often overwhelmed by the multitude of facts that linguists discuss
about grammar. But students need to remember that they unconsciously already know those facts.
For example, they know not to say things like *I saw child or *I live in the Michigan. Students do
have to learn the terminology that linguists use, such as “proper noun,” but they do not have to learn
the facts; they only need to learn how to recall the facts. Thus, a major focus of this book is on
learning to ask the right questions to become consciously aware of the facts one already knows
unconsciously. This is not as easy as it sounds. If it were simple, linguists would have uncovered
all the facts of grammar long ago.
As we have observed, the basic distinction in the above classification of nouns is that proper nouns
cannot be specified by a determiner and common nouns can:
(91) a. *I like the Heather and the Daisy.
b. I like the heather and the daisy.
179

180
The grammar we are developing must specify the distinction between proper and common nouns in
some explicit way. Most grammars of English account for that distinction by postulating a
phonologically null determiner, represented by the symbol Ø. Grammarians argue that this
determiner is present when it appears that a common noun does not have a determiner. The full
analysis is as follows:
(92) a. I like the heather. [ay layk ð hðr]
b. I like Ø heather. [ay layk hðr] - phonetically equivalent to (92f)
c. *I like heather. (ungrammatical because there is no determiner before heather)
d. *I like the Heather. (ungrammatical because the occurs before Heather)
e. *I like Ø Heather. (ungrammatical because of the presence of Ø before Heather)
f. I like Heather. [ay layk hðr] - phonetically equivalent to (92b)
The example in (92c) is ungrammatical because heather, a common noun, requires Ø when there is
no overt determiner (cf. (92b)). The example in (92e) is ungrammatical because Heather, a proper
noun, cannot be specified by a determiner, including the phonologically empty determiner Ø (cf.
(92f)). The bottom line here is that our grammar must account for the fact that native speakers know
the difference between thousands of proper nouns like Heather, Daisy, Bill, John, Grace, Jack,
Dawn, etc., and thousands of common nouns like heather, daisy, bill, john, grace, jack, dawn, etc.
Note that (92b) and (92f) are spoken and heard in exactly the same way. Postulating Ø enables us
to distinguish the two types of nouns formally.
Summarizing, the above analysis says this: proper nouns never have a determiner; common nouns
always have a determiner; if a common noun lacks an overt determiner, then Ø is present.
The above analysis extends to possessives (Heather’s, women’s, her, etc.) and to quantifiers (all,
both, etc.). In English, an NP can contain either a possessive or a determiner, but not both – another
fact that native speakers unconsciously know. Further, an NP can only contain a quantifier if it also
contains a determiner. Consider the following, where POS represents the possessive suffix
(apostrophe) and notice that POS is in the same position as determiners, including Ø:
(93) a. all the children
b. all Ø children
c. *all POS children (*all’s children)
d. *Heather the children
e. *Heather Ø children
f. Heather POS children (Heather’s children)
g. *her the children
h. *her Ø children
i. her POS children (her children)

A possessive pronoun like her is actually a composite of she and the possessive morpheme (POS),
just as went is a composite of go and the past tense morpheme (PST) and did is a composite of do
and the past tense morpheme (PST) as we saw above.
Again, all the above facts about noun phrases must be known by native speakers, at least
unconsciously, because they know which nouns can be specified by determiners and which can’t.
It is worth repeating here that the facts are facts in the grammar of English. Students often think that
grammatical facts and rules are things that teachers make up. That is incorrect. Languages have
grammar whether or not there are teachers to talk about the grammar. There is no human language
that does not have a grammar because languages require grammar. Grammar is, as we noted in the
beginning of this chapter, a biological necessity.
What we are trying to do here is to account for the grammatical facts native speakers know in the
most generalized way possible, an objective of all science. All of the revisions we have made in our
formal description of English are DATA-DRIVEN: we observe the data, discover the facts in
English grammar, and then adjust our rules in such a way that our written, descriptive grammar
accounts for the data and the facts (see Chapter One, Section 1.2). Linguists postulate revisions in
their grammatical descriptions and rules only when there are data to support those revisions.
Now notice that possessives themselves are noun phrases; in the following examples people is
modified by the adjective wealthy and donations is modified by the adjective generous:
(94) a. the people’s generous donations
b. those wealthy people’s generous donations
c. those very wealthy people’s very generous donations
d. their very generous donations
Furthermore, the possessive marker (the apostrophe) is a separate element affixed to the end of the
entire possessive noun phrase:
(95) a. the people’s donations [the people]’s donations
b. the people in that room’s donations [the people in that room]’s donations
c. *the people’s in that room donations
Lastly, as we have seen, determiners and possessives cannot specify the same noun:
(96) a. those donations
b. the people’s donations
c. their donations
(97) a. *those the people’s donations *the people’s those donations
b. *those their donations *their those donations
181

182
In other words, determiners and the possessive occupy the same position in noun phrases. A
quantifier like all can only be followed by a determiner (all the books, *all’s books), and an ordinary
noun like John or child can only be followed by the possessive marker (John’s book, that child’s
books, *John the book, *that child the books (cf. (93)). The facts require us to modify (88), repeated
here as (98), to (99).
(98) NP (QNT) + (DET) + (ADJ) + N
(99) NP ( {QNT, NP} ) + (DET) + (ADJ) + N
A cluster like ( {QNT, NP} ) in above rule, means that either a QNT or an NP can optionally occur
in the first position of an NP.
Continuing to explore the internal structure of noun phrases, we note that noun phrases can contain
whole sentences:
(100) a. John heard a rumor.
b. John heard (that) Mary kissed Bill.
(101) a. A rumor upset John.
b. That Mary kissed Bill upset John.
The direct object of (100a) is a rumor, and the direct object of (100b) in the sentence Mary kissed
Bill, which must be a sentence because it contains a tensed verb (kissed).
To accommodate all the varieties of NP that occur in the above examples, we require the following
PS-Rule for NP:
(102) NP
The rule (102) says that an NP has one of two forms: either a noun, which can be preceded by some
optional elements, or a sentence. Thus, (102) is an abbreviation for all the rules in (103).
(103) a. NP QNT + DET + ADJ + N
b. NP QNT + DET + N
c. NP QNT + ADJ + N
d. NP QNT + N
e. NP DET + ADJ + N
f. NP DET + N
g. NP ADJ + N
h. NP N
i. NP S

Again, observe that the left to right order in PS–RULES is crucial: it expresses the fact that the NP
in (104) are grammatical while those in (105) are not:
(104) a. the new tenor from Italy
b. the tenor
c. tenors from Italy
(105) a. *the from Italy new tenor
b. *tenor the
c. *from Italy tenors
3.5.2 PHRASE STRUCTURE RULES: FIRST PROPOSAL
Expanding our corpus, we can allow for more complex phrases such as those in (106) if we make
the additions to our rules specified in the right-hand column (note that (106d) is ambiguous).
(106) a. He will not go. Add a negative (NEG) to the TP rule.
b. Mary quickly finished the assignment. Add an adverb (ADV) to the VP rule.
c. They elected Bill president. Add a second NP to the VP rule.
d. The thief hit the lady with the hat. Add a PP (Prepositional Phrase) to both
VP rule and NP rule.
These new examples gives us the set of PS–Rules (107).
(107) a. S NP + TP + VP
b. TP (MOD) + T + (NEG)
c. VP (ADV) + V + (NP) + ( {NP, PP} )
d. PP P + (NP)
e. NP
f. T {PRS, PST, FUT, CND, IMP}
These rules will generate the appropriate structure for the sentence in (106). In particular, they
account for the ambiguity of (106d): if the PP with the hat is what the thief used to hit the lady, then
the PP is part of the VP (107c); if the PP with the hat describes what the lady is wearing, then the
PP is part of the direct object NP (107e). As before, the arrow in these rules is an instruction to
rewrite the symbol to the left as the symbol or symbols to the right. Parentheses indicate optional
183

184
constituents; braces indicate alternative constituents. The only function of the plus symbol in
PS–RULES is to mark a clear separation of constituents; it is, therefore, optional. Note again that
the left to right order of constituents is an entirely empirical matter; once a particular order is
determined from an examination of possible phrase structures, it cannot be changed unless there is
other evidence indicating that the change is justified. The rules in (107) are not a complete list of
all the PS–Rules for English; for example, notice that conjunctions like and are not included. These
rules will be added to and modified as we proceed.
The name given to grammars formulated in terms of phrase structure rules and tree diagrams is
PHRASE STRUCTURE GRAMMARS; the theory behind them, as we have indicated, is called
STRUCTURAL GRAMMAR. In developing a set of phrase structure rules like (107), a linguist
will attempt to reach the highest level of generality possible, which means proposing rules that
describe the greatest number of cases (out of the infinite number possible) in the simplest way. For
example, postulating tense as occurring after MOD and before V accounts most simply for did play
(do PST play) and played (PST play). Considering only a sentence like they played, one might
conclude that an analysis which places tense to the left of V is unnecessarily abstract. But remember,
a grammar attempts to account for all possible sentences. In English, the tense marker (PRS, PST,
etc.) is affixed to the closet verb, whatever that is (modal or main verb).
In addition to identifying all of the major constituents of a sentence and indicating how these
constituents are related to each other, phrase structure grammars offer a formal way of formulating
such functional notions as subject and object. Notice that in a tree diagram like (75) some of the
nodes in the tree immediately dominate others, that is, are directly above them. For example, the S
node immediately dominates an NP node, a TP node, and a VP node; the VP node immediately
dominates a V node and an NP node; and so on. Using this relationship of immediate dominance,
one can define the subject of a sentence as the NP node which is immediately dominated by S, and
the object as the NP node which is immediately dominated by VP. Similarly, the predicate can be
defined as the VP immediately dominated by S, and the main verb can be defined as the V
immediately dominated by VP. Linguists also use “family relationships” to describe positions in
phrase structure, for example, NP, TP and VP are sisters as well as daughters of S.
Such formal definitions are preferable to definitions like those in (59) because they are explicit, i.e.,
they refer to specific entities standing in unique relationships. A definition like the direct object
specifies the person or thing directly affected by the action defined in the verb, on the other hand,
is not explicit and, therefore, raises many problems. For example, consider the direct objects in the
following sentences:
(108) a. The man will burn the money.
b. The man will earn the money.
Under phrase structure analysis, the sentences in (108) would be assigned the constituent structure
given in (109).

(109)
This tree characterizes the object NP of both (108a) and (108b) in the same way. A comparison of
the following sentences indicates that this characterization is correct.
(110) a. How much money did the man burn?
b. How much money did the man earn?
(111) a. The money was burned quickly.
b. The money was earned quickly.
If there were something different about the occurrence of the NP the money in (108a) and (108b),
one would not expect them to behave in the same way. But, the sentences in (110) and (111), and
many others, show that they are, in fact, the same.
Now, according to the definitions in (59), the NP the money in (108a) is the direct object because
it specifies the thing directly affected by the action defined in the verb. This seems defensible. With
regard to the NP the money in (108b), however, a problem arises, because it is difficult to conceive
of the money as being directly affected by the action of earning. This problem becomes even greater
for sentences like the following:
(112) Sue earned a bad reputation.
(113) John suffered a blow on the head.
Generally speaking, this problem always arises when one attempts to describe language on the basis
of definitions like those in (59). Another example occurs with the definition usually given to nouns,
i.e., a noun is the name of a person, place, or thing. This definition combined with (59) leads to
statements like A noun can be the name of a thing which is the topic of an expression. Such
statements reveal little about English sentence structure. On the other hand, a description of a
category like noun, formulated in terms of phrase structure grammar, will explicitly separate those
items which are nouns from those which are not. For this reason, most linguists have abandoned
definitions like those in (59) in favor of more rigorous and formal ways of characterizing sentence
structure, such as those afforded by phrase structure grammar. Of course, in aiming for such
explicitness and generality, linguists are implementing the methodology of all scientists.
185

186
3.6 THE ENDOCENTRICITY CONSTRAINT
PS-Rules like (107) and the structural diagrams they generate like (109) also allow us to begin to
make some important generalizations about phrase structure. For example, notice that, in (109), the
major phrases (NP, TP and VP) each dominate a category of the same type (N, T, and V,
respectively). The technical term for this is ENDOCENTRICITY, which means that every
constituent X must be dominated by a phrase of the same type XP, whatever X is. Further, every
phrase XP must dominate an X. What that means, for example, is that noun phrases must contain
nouns and verb phrases must contain verbs. We can formulate this Endocentricity Constraint as
follows:
(114) ENDOCENTRICITY CONSTRAINT.
a. Each XP must dominate one and only one X; each X must be dominated by one and
only one XP. The constituent X is called the HEAD of the XP. The phrase XP is
called the MAXIMAL PROJECTION of X.
b. Each head X must dominate one and only one word or abstract category (PRS, PST,
FUT, POS, Ø, etc.).
The Endocentricity Constraint stipulates the following:
(115) a. XP X
b. *XP Y
c. *XP X + X
There are three constituents in (109) which do not obey the Endocentricity Constraint, DET, MOD
and S. Implementing the Endocentricity Constraint for DET and MOD, we modify (109) as follows
(we will consider S later in this chapter):
(116)
In the above representation, with the exception of S, every X is dominated by an XP, and every XP
dominates X. Further, each head X dominates only one word or abstract category.

It is important to note, at this point, that good scientific inquiry always seeks to uncover
generalizations like the Endocentricity Constraint even when there are apparent exceptions such as
S. Also, given the examples we have considered so far, it appears as though DP and MP are
HEAD–ONLY constructions, that is, phrases that only can contain a head.
Returning to our original example, we now have the following representation for The passengers do
obey the stewardess:
(117)
Implementing the Endocentricity Constraint, we can now generalize the internal structure of phrases
in our PS rules as follows:
(118) a. S NP + TP + VP
b. TP (MP) + T + (NEGP)
c. VP (ADVP) + V + (NP) + ( {NP, PP} )
d. PP P + (NP)
e. NP
f. T {PRS, PST, FUT, CND, IMP}
g. XP X
Notice that the structure and rules above are more generalized than those we have had in the past,
because the internal structure of phrases is now more generalized. In our investigation, there has
been a gradual progression toward greater generality from (60b) to (75) to the representation and
rules above. In fact, as we proceed in this chapter, we will uncover evidence supporting the principle
that all phrases have the same internal structure.
187

188
3.7 THE LEXICON
In order to generate a tree structure like (117) from rules like (118), our grammar must also contain
a lexicon or dictionary of all the words in the language. We must have a way to ensure that words
are inserted into the appropriate slots in structure, that is, that verbs are inserted under V, nouns
under N, determiners under D, and so on. Recalling our use of features from phonology, let us
propose a lexicon with SYNTACTIC FEATURE distinctions such as the following:
(119) a. truth [+N] N=Noun
b. explain [+V] V=Verb
c. the [+D] D=Determiner
d. from [+P] P=Preposition
Some words can clearly function as more that one category:
(120) a. love [+N; +V]
b. patient [+N; +ADJ]
c. swell [+N; +V; +ADJ]
d. down [+N; +V; +ADJ; +P]
Given feature specifications like (119) and (120), we ensure that words are inserted into tree
structures with the following principle of lexical insertion:
(121) For each X, freely insert a lexical item (word) with the feature specification [+X, ...].
Given (121), only words with the feature [+X, ...] will be inserted into tree structures under the
category X.
In addition to ensuring that words are inserted under the appropriate category, we must also ensure
that they are inserted into the appropriate environments (before and/or after the appropriate
structures). For example, we must ensure that verbs which require a direct object (Transitive Verbs)
are inserted into structures where an NP follows the verb, such as (117). Similarly, verbs which
cannot have direct objects (Intransitive Verbs) should not be inserted into structures before an NP.
Linguists call such restrictions CO-OCCURRENCE RESTRICTIONS. They stipulate the
structural environments in which lexical items can occur. Adopting the formalism we used in
phonology for environments, we have examples like the following (see Rule Formalism on page
127):
(122) a. obey [+V, +/___NP] Transitive Verb
b. smile [+V, –/___NP] Intransitive Verb
c. break [+V, ±/___NP] Transitive or Intransitive Verb
d. give [+V, +/___NP + PP] Transitive Verb + Indirect Object

These entries stipulate that a verb like obey can only be inserted into trees having a transitive
structure, that is, a VP that dominates a V followed by an NP such as (117).
Using syntactic features like the above, we can make improvements in the rules in (118) to attain an
even higher level of generality and descriptive adequacy. Consider the following examples of adverb
phrases (ADVP) and adjective phrases (ADJP):
(123) ADVP
a. very beautifully
b. as helpfully as Mary
c. so much more skillfully than Mary
(124) ADJP
a. very beautiful
b. as helpful as Mary
c. so much more skillful than Mary
Clearly, adverbs and adjectives are essentially the same type of syntactic category such that one can
easily be turned into the other. The major difference, of course, is that ADVP modify verbs and
ADJP modify nouns, facts which are directly expressed in the PS-Rules (118). Notice that ADVP
is to the left of V and dominated by VP, while ADJP is to the left of N and dominated by NP. Given
these relationships, it is clear that adverbs and adjectives are members of some larger category,
which linguists call ADJUNCT or A. Implementing this observation, we can replace both ADVP
and ADJP with AP, for ADJUNCT PHRASE. An adjunct is a modifier. The position of the AP
in structure will determine whether it is functioning adverbially or adjectivally: in a VP, it will be
adverbial; in an NP, it will be adjectival. Using features, let us say that adverbs are adjuncts that are
nonadjectival and adjectives are adjuncts that are adjectival:
(125) a. [+A, –ADJ] beautifully, helpfully, skillfully, happily, quickly, etc.
b. [+A, +ADJ] beautiful, helpful, skillful, happy, quick, etc.
c. [+A, ±ADJ] hard, fast, daily, etc.
In a similar way, we do not need to specify both QP and NP in (118e) because quantifiers are a type
of noun. In fact, sometimes the same word can function as either a quantifier noun or a nonquantifier
noun. Consider the following (in the “a” examples, the italicized words function as quantifier nouns;
in the”b” examples, as nonquantifier nouns):
(126) a. Half the men were eliminated. (cf. All/both the men were eliminated)
b. We cut the team in half. (cf. We cut the team into two groups.)
(127) a. We made three times the profit. (cf. We made all the profit.)
b. Times are hard. (cf. Decisions are hard.)
189

190
(128) a. We made double the profit. (cf. We made twice the profit.)
b. She is my double. (cf. She is my stand-in.)
Quantifier nouns (all, both, etc.) and determiners are actually in complementary distribution with
nonquantifier nouns (John, child, etc.) and the possessive suffix. Determiners can only be preceded
by quantifier nouns and POS can only be preceded by nonquantifier nouns. Consider again examples
like the following:
(129) NP DP HEAD
a. all the books
b. all Ø books
c. *all POS books (*all’s books)
d. *John the books
e. *John Ø books
f. John POS books (John’s books)
Since a quantifier is a type of noun, we can replace ({QP, NP}) with just (NP) in the rule (118e);
and, given the above discussion, we can also replace ADJP with AP in that rule. The new rule is as
follows:
(130) NP
Using features, we can distinguish individual common nouns and quantifiers in the following way:
(131) a. [+N, –Q] book, girl, truth, group, decision, profit, etc.
b. [+N, +Q] all, both, each, twice, etc.
c. [+N, ±Q] half, time, double, etc.
All the data taken together show that determiners, quantifiers, and possessives have the same
function, namely, they specify nouns (this child, all those books, Mary’s decision, etc.). That
indicates that they are members of a larger category, which linguists call SPECIFIERS.
We can also use syntactic features to distinguish subcategories of nouns and pronouns as follows:
(132) a. [+N, +COMMON] heather, jack, state, etc.
b. [+N, –COMMON, –PRO] Heather, Jack, Michigan, etc.
c. [+N, –COMMON, +PRO, +REFLEXIVE] herself, himself, itself, etc.
d. [+N, –COMMON, +PRO, –REFLEXIVE] she, he, it, etc.
Proper nouns ([+N, –COMMON, –PRO]) share many features with pronouns ([+N, –COMMON,
+PRO]). Both subcategories of nouns cannot occur after quantifiers, determiners and adjectives or
before PP:

(133) a. *That happy Richard laughed.
b. *A tall she married a short him.
We can account for these co-occurrence restrictions borrowing the formalism for redundancy rules
in Chapter Two (cf. (46) on page 123):
(134) a. [–COMMON] [–\NP__ ]
b. [–COMMON] [–\DP__ ]
c. [–COMMON] [–\AP__ ]
d. [–COMMON] [–\__ PP]
These redundancy rules state that all noncommon nouns (proper nouns and pronouns) cannot occur
after quantifiers, determiners, and adjectives or before PP. Given the Endocentricity Constraint, we
can collapse the four redundancy rules above into one rule as follows:
(135) [–COMMON] [–\ XP __, –\ __XP]
This redundancy rule states that all noncommon nouns cannot occur after or before any other type
of phrase within an NP.
3.8 A NOTE ON SCIENTIFIC INQUIRY
In the preceding sections, we have discussed some attempts to develop an adequate theory of
grammar beginning with traditional grammar and culminating in a version of structural grammar that
reaches a high level of generality with the Endocentricity Constraint. Reviewing this, we can see
how linguistics has been able to deepen and broaden its characterizations of human language in an
effort to attain ever higher levels of adequacy. The search for descriptively and explanatorily
adequate models of grammar continues today.
It is often dismaying to come to a discipline for the first time and find that it does not provide a
succinct, unified, and fully justified theory. Linguistics is such a discipline; however, although there
is no grammatical theory which is championed by every living linguist, the issues today are quite
clearly drawn, and so are the objectives of research. The discovery of the grammatical facts
themselves is much more significant than the details of particular representations of those facts.
No matter what grammatical model is proposed, it must be able to provide a characterization of all
sentences in such a way that it accounts for the native speaker’s knowledge of them. This involves,
among other things, accounting for ungrammatical sentences in a principled way, that is, in a way
which predicts that they will be ungrammatical. The facts to be accounted for are the same
regardless of the theory invented to characterize them. The objective of the discipline is to provide
a theory which will be consistent with the ever–expanding corpus of data. In this respect, linguistics
is no different from any other science. Consider, for example, the study of celestial objects.
191

192
Astronomy is one of the oldest sciences. Even before 2000 B.C., crops were sowed according to the
movements of the sun and moon. Modern theories of the universe which displace the earth from the
center are due to Copernicus (1473–1543). There have been many modifications over the centuries
even to the point that separate areas of inquiry or methodology have been identified like radio
astronomy, astrophysics, and spectroscopy. The discovery of pulsars, quasars, black holes, and other
cosmic phenomena has engendered modifications in existing theories of the universe. As our
knowledge of the universe has expanded, the field of astronomy has expanded, and a number of
distinct theories for a number of distinct issues have emerged, e.g., theories on the origin of the
universe, the big bang theory, string theory, brane theory, and so on. In the process, the explanatory
power of the combined theories has increased. The discovery of some new cosmic phenomena never
nullifies everything that precedes, that is, astronomy does not begin again from scratch. Rather,
existing theories are appropriately modified or, if the phenomena are not understood, left entirely
unmodified despite apparent conflicts or inconsistencies. It is always possible that putative
exceptions are not exceptions at all, only misunderstood cases.
People have been studying language almost as long as they have been speculating on the heavens.
Terms like verb which form part of the theories described in this book were coined by ancient
grammarians. However, it is only in the last fifty years or so that linguists have been concerned with
many of the issues discussed in this book, in particular, those issues formulated here as the
grammatical characterization problem and the grammatical realization problem. In short, students
of linguistics must understand that new models of characterization are inevitable as new data are
discovered. During this evolution, the issues have become more clearly defined, and the
characterizations have become more explanatorily adequate. More than anything else, this process
and progress of scientific inquiry must be understood.
In addition, students must understand that grammars are theories about language. In writing
grammars, linguists attempt to describe observed phenomena, such as the different ways in which
they observe native speakers using proper nouns and common nouns. Sometimes, the description
will make use of abstract elements such as Ø postulated above. In this regard, linguists are no
different from other scientists. For example, physicists postulate the existence of dark matter, that
is, “invisible” matter, to account for abnormalities in the observed rotation of galaxies. No one has
ever seen dark matter anywhere in the universe, just as no one has ever heard Ø. Nevertheless,
scientists assume that these theoretical constructs exist so that they can account for the phenomena
they do observe.
In our discussion, various approaches to grammatical characterization are adopted, discussed, and
either discarded or maintained as our inquiry proceeds. To some, this may seem to be a waste of
time. Why not ignore the historical perspective and just discuss the right analysis? The answer is
quite simple: there is no right analysis. There is only the analysis which covers the known facts at
the present time. Furthermore, science sometimes takes one or even a few steps backwards,
returning to an older theory once discarded and resurrecting it with modifications.

Since our understanding of language is increasing all the time and analyses, consequently, can be
expected to be falsified by newly discovered data, the most important thing to be learned is how to
do grammatical analysis to begin with. This prepares students for the changes that are inevitable in
so young an area of inquiry as formal linguistics. Therefore, our focus must be on the methods and
objectives of linguistic inquiry. At every stage in this inquiry, we will formulate hypotheses and
examine their consequences for the theory of language. This process of hypothesis development is
common not just to linguistics, but to scientific method in general.
Everything said here about scientific inquiry applies to the analysis of problems that confront society,
businesses, industry, government, and all other institutions and enterprises. As we noted in the
Introduction, the problems facing today’s workers are complex and subtle, as difficult to solve as the
nature of the human mind, language, and the universe. What we have done throughout this book is
devise steps for breaking down problems into manageable parts, for making proposals that are robust
and workable, for testing hypotheses, and for finding solutions that satisfy the most factors and
constituents in the best way possible. Linguistic analysis has been our means to develop expertise
in problem solving. At an abstract level, formulating a theory of language, which is our specific
problem, is like all problems.
3.9 X–BAR SYNTAX
Returning to our main discussion, let us continue to explore the nature of an adequate theory of
syntax. Thus far, our discussion has pivoted around the concept of SYNTACTIC UNIT. To find
these units, we considered the three syntactic processes of REFERENCE, OMISSION, and
PLACEMENT. We have the following definitions:
(136) a. PHRASE: a word or group of words that functions as syntactic unit in
REFERENCE, OMISSION, and/or PLACEMENT.
b. HEAD: the minimal element of a phrase.
A phrase whose head is a noun is a noun phrase; one whose head is a verb, a verb phrase; one whose
head is a preposition, a prepositional phrase; and so on. The phrases we have examined in the
sections above have been relatively simple. Consider now the following increasingly complex
examples of noun phrases:
(137) a. those paintings
b. all those paintings
c. all those acquired paintings of landscapes
d. all those recently acquired paintings of landscapes of the Louvre’s
e. all those ten recently acquired Impressionist paintings of landscapes of the Louvre’s
on display
f. all those ten recently acquired Impressionist paintings of landscapes of the Louvre’s
on display, which are so beautiful
193

194
If we attempt to diagram the NP’s in (137) in accordance with the mechanisms we have so far
developed for English phrase structure, we encounter a number of immediate problems that require
a solution. The most difficult problem concerns the fact that both traditional and early structural
grammar provide only the most meager structural framework for basic syntactic units. Quite clearly,
we cannot suspend all the structures in (137f), for example, from the same NP node. Such an
analysis ignores the fact that NP’s have internal hierarchical structure as well as internal linear
structure. To see this, consider the simple NP in (138) provided by the rules developed so far.
(138)
A diagram like (138) leaves a number of important facts unaccounted for. First, note that we have
examples like (139).
(139) a. John saw those Spanish paintings, and Bill saw these Italian ones.
b. John saw those Spanish paintings, and Bill saw these ____.
The word ones in (139a) substitutes for the noun paintings indicating that paintings is an internal
unit within its own phrase. That fact is accounted for by (138): paintings is under one node by itself,
the N node. However, the same diagram cannot account for the gap in (139b). In this sentence, the
gap stands for Spanish paintings which is not identified as a unit in (138). For Spanish paintings
to be a unit the two elements, the AP (Spanish) and the N (paintings), must be under one node,
separate from the DP (those) node. This means that a better diagrammatic representation of (138)
is (140), where the "?" identifies the missing, as yet unlabeled, node.
(140)
This deficit in the internal structure of phrases has not gone unnoticed in modern grammatical
analysis, and, in the 1970’s, a phrase structure framework was developed that ultimately came to be
known as X–BAR SYNTAX. This theory replaces diagrams like (140) with ones like (141), in
which progressively larger syntactic units of the same phrasal type are indicated by progressively
higher primes; for example, N´ designates a larger unit than N. In the original version of this theory,

primes were not used; rather a bar was placed over a symbol, e.g., N¯. Since bars are difficult to
produce on typewriters and word processors, they are generally replaced with primes as follows:
(141)
Notice that, in (141), Spanish paintings is contained in a separate node labeled N´. This node
represents a separate, intermediate noun phrase level between N and NP. The representation claims
that phrases, in particular, noun phrases, must have more internal hierarchical structure than the
amount usually accorded them. One of the many examples that demands this is (139b) which
contains a gap that stands for the unit Spanish paintings, represented by the N´ in (141). The
representation in (141) replaces the "?" of (140) with a new intermediate noun phrase level and,
accordingly, directly accounts for the gap in (139b). Basically, X–Bar syntax recognizes more
syntactic levels than the two traditionally recognized, that is, more than just XP over X.
Before continuing, a somewhat trivial matter should be mentioned – one which seems to help
students read complicated phrase structure diagrams. In X–Prime (or X-Bar) notation, it is useful
to line up all the head elements (XP, X´, X) in a straight vertical sequence making the location of the
head of each separate phrase straightforward. Thus, the representation of (141) is (142).
(142)
Note that the X´ phrase is explicitly mentioned in a tree only if it branches; thus, N´ is mentioned but
neither D´ nor A´ is. Also, for simplicity, we can represent tree diagrams like the above linearly as
LABELED BRACKETS:
(143) [ NP [ DP [ D those ] ] [ N´ [ AP [ A Spanish ] ] [ N paintings ] ] ]
195

196
Note that each bracket on the left is paired with one on the right beginning with the smallest units
and proceeding to increasingly larger units. The following diagram illustrates which brackets are
paired (first the X phrase, then the X´ phrase if there is one, and lastly the XP phrase):
(144) ¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯
¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯
¯¯¯¯¯¯¯¯¯¯ ¯¯¯¯¯¯¯¯¯¯¯¯¯
¯¯¯¯¯¯ ¯¯¯¯¯¯¯¯ ¯¯¯¯¯¯¯¯¯
[ [ [ those ] ] [ [ [ Spanish ] ] [ paintings ] ] ]
NP DP D N´ AP A N
Returning to the main discussion, consider now a phrase like (145), which can have either of the two
meanings indicated.
(145) those Spanish teachers
a. those teachers who are Spanish
b. those teachers of Spanish
In addition, we have the following examples:
(146) a. those Spanish Braille teachers
b. *those Braille Spanish teachers
(147) a. John saw those Spanish math teachers, and Bill saw these ________.
b. John saw those Spanish math teachers, and Bill saw these Italian ones.
These phrases require that compound nouns like Braille teachers in (146a) and math teachers in
(147) be contained under a separate N node, giving structures like the following:
(148) (149)

The above representations contain the compound nouns Braille teachers and math teachers. In such
compounds, the first element is actually the COMPLEMENT of the head, it completes it’s
meaning: a Braille teacher is a person who teaches Braille; a math teacher is a person who teaches
math. In these representations, both Braille teachers in (148) and math teachers in (149) are
contained under one N´ node. Such representations account for the use of ones in examples like
(147b). Observe also that these representations can account for the ambiguity of a phrase like (145).
If it means (145a), then Spanish is a descriptive adjective and branches off the higher N´ if it means
(145b), then Spanish is part of a compound and branches off the lower N´. Further, this analysis
predicts that both should be able to occur together, and that is correct: those Spanish Spanish
teachers is possible; where the first Spanish (an adjective phrase or AP) refers to origin, and the
second Spanish (a noun phrase or NP) refers to language.
These data indicate that the order of elements within an NP is fixed: SPECIFIERS must dominate
ADJUNCTS and ADJUNCTS must dominate COMPLEMENTS. We have no examples like
(150).
(150) a. *(John met) those math Spanish teachers.
b. *(John met) Spanish those teachers.
c. *(John met) math Spanish those teachers.
In the following two diagrams, only (151) is grammatical because the adjunct (AP) Spanish
(indicating national origin) dominates the complement (NP) Spanish (indicating the language). In
(152), the two constituents are reversed so the phrase is not grammatical (cf. *those math Spanish
teachers).
(151) (152) *
197

198
The data that we have observed and the facts that we have discovered indicate that the Endocentricity
Constraint (114) must be expanded as follows:
(153) ENDOCENTRICITY CONSTRAINT.
a. Each XP must dominate one and only one X; each X must be dominated by one and
only one XP. The constituent X is called the HEAD of the XP.
b. Each head X must dominate one and only one word or abstract category (PRS, PST,
FUT, POS, Ø, etc.).
c. Between XP and X, multiple recursions of X´ are possible.
d. Specifiers must dominate adjuncts; adjuncts must dominate complements.
3.10 GENERALIZING PHRASE STRUCTURE: S AND NP
While the revisions that we have implemented reach a much higher level of descriptive adequacy
than the rules of traditional grammar, they still leave many generalizations unexpressed. For
example, in every language, there are abundant instances of cross-categorial principles, that is,
principles which spread over many different categories like sentences (S) and noun phrases (NP).
Consider the following:
(154) S: The people generously donate to that charity.
NP: The people’s generous donations to that charity...
(155) S: The barbarians suddenly destroyed the city.
NP: The barbarians sudden destruction of the city...
(156) S: John steadfastly refuses to do anything illegal.
NP: John’s steadfast refusal to do anything illegal...
(157) S The Mafia illegally shipped drugs into the country during June.
NP: The Mafia’s illegal shipments of drugs into the country during June...
Notice that in examples like (154) and for any given verb/noun pair like donate and donation, it is
possible to substitute the same kinds of subjects (the people, Americans, etc.), the same kinds of
complements (to that charity, to Detroit, etc.) and the same kinds of adjuncts (generous, generously,
etc.). Some of these generalizations can be expressed by the principles we have formulated. For
example, consider the following representations for the sentence and noun phrase in (154):

(158) S: The people generously donate to that charity.
(159) NP: The people’s generous donations to that charity...
199

200
Given these structures, we can make the necessary generalizations about complements and adjuncts
because they occupy exactly parallel positions. Both adjuncts generous and generously are
dominated by an X´ over another X´; the complements to that charity follow the head X, which is
dominated by an X´. Generalizations regarding the subjects, however, cannot be expressed, because
the above diagrams do not completely match each other; in particular, there seems to be an extra VP
node in (158). Furthermore, the symbol S does not obey the Endocentricity Constraint. We can
solve these problems if we replace (158) with the following structure:
(160) S: The people generously donate to that charity.
The structure in (160) exactly parallels (159) so all relationships can be generalized between
sentences and noun phrases. Just as DP (POSSESSIVE) specifies the head noun, TP (TENSE)
specifies the head verb; the subjects of both structures is the phrase the people. Again, both adjuncts
generous and generously are dominated by an X´ over another X´; the complements to that charity
follow the head X, which is dominated by an X´.
Essentially, (160) represents a sentence as a large verb phrase (VP), which is appropriate since the
main constituent of a sentence is the verb.
3.11 ADJUNCTS (MODIFIERS) AND COMPLEMENTS
There is one remaining problem with the above structures and that concerns the distinction between
adjuncts (modifiers) and complements. Notice first that the distinction between adjuncts and
complements is real. Adjuncts are always higher in the tree structure than complements, and the
same phrase often can have either function. We have examples like the following:

(161) Prehead NP position.
a. the cotton sack
[1] ‘the sack made of cotton’ or ‘the sack for cotton’
[2] the cotton cotton sack
[3] the cotton potato sack
[4] *the potato cotton sack
b. a grand piano
[1] ‘a really nice piano’ or ‘a type of piano’
[2] a grand grand piano
[3] a terrific upright piano
[4] *an upright terrific piano
(162) Posthead NP position.
a. the choice of the men
[1] ‘the choosing of the men’ or ‘the choice by the men’
[2] the choice of the men of the men
[3] the choice of the men by the men
[4] *the choice by the men of the men
b. the rumor that they made up
[1] ‘the rumor about their making up’ or ‘the rumor which they made up’
[2] the rumor that they made up that they made up
[3] the rumor that they made up which they made up
[4] *the rumor which they made up that they made up
(163) Posthead VP position.
a. He decided on the boat.
[1] ‘He chose the boat’ or ‘He was on the boat when he decided.’
[2] He decided on the boat on the boat.
[3] He decided on his vote on the boat.
[4] *He decided on the boat on his vote.
b. He reads the most.
[1] ‘He reads the most things’ or ‘He reads the most often.’
[2] He reads the most the most.
[3] He reads the most books the most often.
[4] *He reads the most often the most books.
201

202
The problem with the structures (158) and (160) concerns the ambiguities in the above examples;
specifically, there is no way to tell whether a phrase is functioning as an adjunct or as a complement
in representations like the following:
(164) a grand piano
(165) He decided on the boat.
The constituents grand and on the boat in the above trees may either be adjuncts or complements
given our statement of the Endocentricity Constraint (153). In both cases, the ambiguities are
structural, not lexical, that is, the ambiguities pertain to syntactic relationships, not the meaning of
words. In a sentence like I found the bug, the ambiguity derives from the multiple meanings of the
word bug (insect, computer problem, etc.). Other examples of LEXICAL AMBIGUITIES are as
follows:

(166) a. The bandit was shot by the bank. (bank = financial institution, edge of river, etc.)
b. The students are revolting. (revolting = disgusting, uprising, etc.)
Other examples of STRUCTURAL AMBIGUITIES occur in the following examples, where the
constituents can be grouped in more than one way:
(167) a. The farmer looked over the fence.
[1] The farmer - looked over - the fence.
[2] The farmer - looked - over the fence.
b. The man who left suddenly returned.
[1] The man - who left suddenly - returned.
[2] The man - who left - suddenly returned.
The structural ambiguity in (164) and (165) cannot be resolved by different hierarchical groupings
because the problem does not involve the left-to-right hierarchical order of the constituents. To
solve this problem, we might propose to make a distinction between an X´ over another X´, on the
one hand, and an X´ over an X, on the other, as follows:
(168) a. XP ...X´... Specifier Rule
b. X´ ...X´... Adjunct Rule
c. X´ ...X... Complement Rule
This proposal means that He decided on the boat would have the following two representations,
where on the boat is an adjunct in the first and a complement in the second:
(169) a. He decided on the boat. (on the boat is a adjunct; it is where he decided)
203

204
b. He decided on the boat. (on the boat is a complement; he chose the boat)
But this proposal will not work because complements can be recursive just as adjuncts can. Note
first that we can have multiple adjuncts, that is, recursive X´ levels:
(170) a. That pretty, little, yellow house is for sale.
b. He drives a sleek, green, speedy Maserati.
Similarly, complements and compound nouns can be stacked recursively:
(171) a. That kitchen appliance factory guarantee is very generous
(cf. The factory guarantees the kitchen appliance.)
b. The terrorist cell police raid stunned everyone.
(cf. The police raided the terrorist cell.)
c. The sexual harassment policy court decision was of national interest.
(cf. The court decided on the policy on sexual harassment.)
d. She filled out the customer satisfaction store survey.
(cf. The store surveys customer satisfaction.)
e. They dedicated a cancer patient radiation treatment facility.
(cf. A facility treats cancer patients with radiation.)
According to (168b) and the proposals we have made, a phrase like kitchen appliance factory
guarantee would have the following representation, which would make kitchen appliance an adjunct,
rather than a complement (direct object) of guarantee:

(172)
To solve this problem, we will make one final revision to phrase structure levels. Let us assume a
three level version of the Endocentricity Constraint such that each level is distinctly identified with
numbers and specifiers are always daughters of X3, adjuncts are always daughters of X2, compounds
and complements are always daughters of X1, and the head is always X0. Just as the original
Endocentricity Constraint stipulated that each phrase minimally consists of an XP over an X, the new
revision stipulates that each phrases minimally consists of an X3 over an X0. This new phrasal
architecture yields diagrams like the following (note again that intermediate levels are only
mentioned in a diagram if they branch):
(173) a. b.
205

206
With this revision, we can diagram the noun phrase the kitchen appliance factory guarantee and its
related sentence (The factory guaranteed the kitchen appliance) as follows:
(174) a.
b.
Note that in both of the above trees, kitchen appliance is a complement (X1 level constituent). Since
each level is uniquely specified with a different number, recursion is possible at all levels without
ambiguity.

Adding a quantifier and adjuncts (both adjectives and an adverb) to the above phrase, we have
examples like all those unexpected, very generous kitchen appliance factory guarantees:
(175)
207

208
With all the above revisions, it is now possible to accurately represent ambiguous phrases in which
constituents can be interpreted either as an adjunct or as a complement:
(176) a. [ N3 [ N2 [radio station] [ N1 night manager]]] = a night manager at a radio station
b. [ N3 [ N1 [radio station] [ N1 night manager]]] = a night manager of a radio station
A radio station manager can be either a night manager at a radio station (different
managers at the same station), in which case radio station is an adjunct (N2), or a
night manager of a radio station (someone who manages the whole radio station), in
which case radio station is a complement (N1).
(177) a. peanut oil preservatives = oil preservatives made of peanuts
[ N3 [ N2 peanut [ N1 oil preservatives]]]
b. peanut oil preservatives = oil preservatives for peanuts
[ N3 [ N1 peanut [ N1 oil preservatives]]]
c. peanut oil preservatives = preservatives made of peanut oil
[ N3 [ N2 [peanut oil] [ N0 preservatives]]]
d. peanut oil preservatives = preservatives for peanut oil
[ N3 [ N1 [peanut oil] [ N0 preservatives]]]
(178) a. American embassy sniper attack = attack of/on the American embassy by snipers
[ N3 [ N1 American embassy [ N1 sniper [ attack]]]]
b. American embassy sniper attack = attack at/from the American embassy by snipers
[ N3 [ N2 American embassy [ N1 sniper [attack]]]]
c. American embassy sniper attack = an attack by American embassy snipers
[ N3 [ N1 American embassy sniper [attack]]]
d. American embassy sniper attack = an American attack by embassy snipers
e. etc.
[ N3 [ N2 American [ N1 embassy sniper [attack]]]]

Returning to our original examples, we have the following representations in X3 notation (cf. (159)
and (160)):
(179) N3: the people’s generous donations to that charity...
(180) V3: The people generously donate to that charity.
209

210
Summarizing, we have the following generalizations:
(181) a. Elements that branch from X3 are SPECIFIERS
in N3 Determiners, possessive marker (POS), quantifiers
in V3 Subjects, tenses
b. Elements that branch from X2 are ADJUNCTS
in N2 Numerals (three, third, etc.), adjectives (silky dress)
Compound elements expressing source or material (silk dress)
in V2 Expressions of time and place, adverbs, predicate nominatives
c. Elements that branch from X1 are COMPLEMENTS and COMPOUNDS
in N1 Complements, direct objects
Compound elements expressing complements (evening dress)
in V1 Complements, direct objects
d. Elements that branch from X0 are parts of words
in N0 Affixes that form nouns (-er in singer, teacher, etc.)
in V0 Prepositions that form two part verbs (put on, throw out, etc.)
Lastly, we reformulate the Endocentricity Constraint as follows:
(182) a. Every phrase has three syntactic levels above the X0 level, which is the head and
contains a word or abstract category (PRS, PST, POS, Ø, and so on).
b. Every word or abstract category must be dominated by an X0 which, in turn, must be
dominated by an X3. Intermediate levels (X2 and/or X1) are explicitly mentioned
in diagrams and labeled brackets only if the phrase contains X2 and/or X1 level
elements other than the head. Implicitly, intermediate levels are always there.
c. There may be multiple recursions of every level; however, each higher level of a
phrase must be greater than or equal to the level below, that is, all the specifiers must
be higher than all the adjuncts which, in turn, must be higher than all the
complements.
X3 above X3 above X2 above X2 above X2 above X1 above X1 above X0
[ all [ those [ pretty [ little [ yellow [ ranch [ vacation [ homes ]]]]]]]]
N3 N3 N2 N2 N2 N1 N1 N0
*X3 above X2 above X3 above X2 above X1 above X2 above X1 above X0
*[ those [ pretty [ all [ little [ ranch [ yellow [ vacation [ homes ]]]]]]]]
N3 N2 N3 N2 N1 N2 N1 N0

3.12 RESIDENTIAL GRAMMAR (RG)
The major unresolved issue in the previous discussion is that our theory still does not provide a way
to express the relationships between syntactic categories like verbs (V3) and nouns (N3). Let us,
therefore, reexamine the matter. Traditionally, syntactic categories have been viewed as atomic, that
is, not composed of other elements. The data from the last few sections suggest that that view is
incorrect. As we have seen, syntactic categories share characteristics which we have accounted for
with features in much the way that sounds share characteristics in phonology.
Feature theory is a standard part of phonological descriptions. For example, consider again the
following features:
(183) a. [+VOICED] produced with vocal cord vibration
b. [–VOICED] not produced with vocal cord vibration
(184) a. [+NASAL] involving exhalation through the nose
b. [–NASAL] not involving exhalation through the nose
(185) a. [+ANTERIOR] articulated at the front of the mouth
b. [–ANTERIOR] not articulated at the front of the mouth
Given these features, we can say that the sounds for m as is mass, b as in bass, p as in pass, and g
as in gas, are as follows:
(186) a. [m] [+VOICED, +NASAL, +ANTERIOR]
b. [b] [+VOICED, –NASAL, +ANTERIOR]
c. [p] [–VOICED, –NASAL, +ANTERIOR]
d. [g] [+VOICED, –NASAL, –ANTERIOR]
Thus, a sound like [m] is nothing more than an abbreviation for a specific set of phonological
features. Let us propose a similar feature analysis for syntax.
We begin by noting the difference between verbs and all other categories, in particular, that verbs
are the only category that carries tense (broke), the only category that shows mode (may break) and
aspect (has broken) distinctions, and the only category that can occur in present and past participial
forms (breaking and broken). Thus, we make the following division:
(187)
211

212
The feature [+VBL] specifies those categories which are inherently marked for distinctions in tense,
aspect, voice, and/or mode; [–VBL] specifies categories that are not so marked. [–VBL] categories
also fall neatly into two major classes, those that behave like nouns, and those that do not.
A distinctive attribute of nouns is that they are inherently marked for distinctions in number, person,
gender, and/or case. Let us specify the former as [+NML]; the latter, [–NML]. The same distinction
occurs in some verb forms. Verb forms that are inherently marked for nominal distinctions [+NML]
include verbal nouns like the gerund basing in Basing the argument on so many assumptions is
unwise; verb forms that are not inherently marked for nominal distinctions include tensed verbs like
bases in He bases his argument on many assumptions, in which the singular number on bases is a
result of agreement and is not an inherent feature of the tensed verb (cf. They base their argument
on many assumptions). To capture these distinctions, we modify (187) to (188) and refer to
nonnouns as CHARACTERIZERS since they characterize a head.
(188)
[+NML] categories differ from characterizers in a number of ways. First, [+NML] categories have
a number of morphological specifications that characterizers lack. They occur before specific
suffixes like –ment, –ness, –ing, etc. (basement, baseness, basing). They show number distinctions:
base versus bases. They occur in the possessive: the base’s shape and the shape of the base.
Second, [+NML] categories fulfill a number of specific syntactic functions, which means they
occupy specific positions in syntactic hierarchies. For example, they can be SUBJECT or
OBJECT. Third, [+NML] categories are always distributed before [–NML] categories on every X
level in both prehead and posthead positions. Notice that in every diagram in this monograph, when
two categories are suspended from the same level, [+NML] categories (NOUNS) always precede
[–NML] categories (CHARACTERIZERS and VERBS).
We now have four major syntactic categories in language, which RG specifies as follows:
(189) a. Verbs (V): [+VBL, –NML] go, goes, went, gone, going
b. Nouns (N): [–VBL, +NML] woman, women, she, all
c. Characterizers (C): [–VBL, –NML] happy, happily, probably, on, after
d. Verbal Nouns (VN): [+VBL, +NML] swimming, eating, reading

Given these major categorial distinctions, we note that many parts of speech are distinguished from
each other mainly by what kind of phrase they can occur in and where they can occur within that
phrase, that is, they have a specific residence. Formally, we define a RESIDENCE as a specific
position in syntactic structure which has a specific semantic or syntactic function. The theory based
on the concept of residence is called RESIDENTIAL GRAMMAR (RG).
The morphosyntactic feature oppositions [±VBL, ±NML] are not sufficient to distinguish the major
syntactic categories that occur in the world’s languages. For example, within the category verb, there
are subclasses which include finite verbs, nonfinite verbs (infinitives, participles, etc.), auxiliaries,
and modals. Within the category characterizer, there are prepositions, subordinators (subordinating
conjunctions), adjectives, adverbs, determiners, and so on. To specify these subclasses, we will use
two features [±OPEN PHRASE] ([±OPH]) and [±OPEN CLASS] ([±OCL]) defined as follows:
(190) a. [+OPH]: occurring as the head of a phrase that freely contains other elements
b. [–OPH]: occurring as the head of a phrase that does not freely contain other
elements
(191) a. [+OCL]: unlimited in number
b. [–OCL]: limited in number
The syntactic feature specification [±OPH] separates categories whose internal phrase structure is
open in the sense that it can freely contain specifiers (words like the, this, and that), quantifiers
(words like all, some, and much), or adjuncts (adjectives and adverbs), from other categories whose
internal phrase structure is highly restricted. The [±OPH] distinction is most clearly seen in the
variety of structures possible for common nouns on the one hand, which can be specified, quantified,
and modified, and pronouns and proper nouns on the other hand, which cannot:
(192) a. That very old woman left early./*That very old she left early.
b. All the tall women walked in./*All the tall they walked in.
c. A woman that is tall will get the job./*Mary that is tall will get the job.
The feature [+OCL] specifies those categories which form an open class to which an unlimited
number of items may be added. Since classical antiquity, most grammarians have observed the sharp
distinction between the very large number of nouns, verbs, adjectives, and adverbs that occur in
languages as opposed to the rather small fixed number of prepositions, conjunctions, determiners,
pronouns, etc. The opposition [±OCL] is intended to capture this dichotomy. For the most part, in
English, open class words are marked by distinct inflectional suffixes: nouns show variations in
number (boy, boys); verbs show variations in tense (plays, played); and adjectives and adverbs show
variations in comparison (fast, faster, fastest; long, longer, longest). Examples of [+OCL] and
[–OCL] categories are the following:
(193) a. [+OCL]: nouns, verbs, adjectives, adverbs
b. [–OCL]: auxiliaries, modals, pronouns, determiners, prepositions, etc.
213

214
Within a particular kind of phrase, certain categories are restricted to X3 level, others to X2 level,
and still others to X1 level. For example, all categories denoting specification (DETERMINERS)
or quantification (QUANTIFIERS) are X3 level residents. All descriptive modifiers of the head are
X2 Level residents. All elements of compounds (history in history teacher) and complements
(history in teacher of history and teach history) are X1 Level residents. Therefore, with reference
to the X hierarchy, we recognize the following level features:
(194) a. [+X3L] can freely occur immediately dominated by X3 (N3, V3, or C3)
b. [–X3L] cannot freely occur immediately dominated by X3 (N3, V3, or C3)
(195) a. [+X2L] can freely occur immediately dominated by X2 (N2, V2, or C2)
b. [–X2L] cannot freely occur immediately dominated by X2 (N2, V2, or C2)
(196) a. [+X1L] can freely occur immediately dominated by X1 (N1, V1, or C1)
b. [–X1L] cannot freely occur immediately dominated by X1 (N1, V1, or C1)
We note further, that some categories are restricted to phrases with specific heads, that is, can be
suspended from some X level only when X is a specific phrase type. DETERMINERS can only
occur in noun phrase, suspended from N3; ADJECTIVES can occur in either noun phrases or verb
phrases, suspended from N2 or V2; MANNER ADVERBS can occur only verb phrases suspended
from V2. Therefore, we recognize the following environment features:
(197) a. [+ENH] can freely occur immediately dominated by N3, N2, or N1
b. [–ENH] cannot freely occur immediately dominated by N3, N2, or N1
(198) a. [+EVH] can freely occur immediately dominated by V3, V2, or V1
b. [–EVH] cannot freely occur immediately dominated by V3, V2, or V1
(199) a. [+ECH] can freely occur immediately dominated by C3, C2, or C1
b. [–ECH] cannot freely occur immediately dominated by C3, C2, or C1
Lastly, we recognize that some categories occur in prehead position (DETERMINERS) and others
in posthead position (PREPOSITIONS). We, therefore, add these feature oppositions:
(200) a. [+PRH] can freely occur before the head of a phrase
b. [–PRH] cannot freely occur before the head of a phrase
(201) a. [+PSH] can freely occur after the head of a phrase
b. [–PSH] cannot freely occur after the head of a phrase
Ultimately, the concept of residence is linked to the syntactic features of RG. These are proposed
as substantive syntactic universals, meaning that the languages of the world must construct their
syntactic inventories out of these features, just as they must construct their phonological inventories

out of substantive phonological features. Despite this constraint, it is important to note that the
theory does not demand that all languages choose the same features in the same combinations. Quite
the contrary occurs. Different languages select different sets of features out of the pool of possible
features, and even the same language, at different points in its history, utilizes features in different
ways. For example, at an earlier stage of English, word order constraints were less stringent than
they are today; therefore, the features [±PREHEAD] and [±POSTHEAD] did not have the
significance they have in modern English. The task of children acquiring a language is to discover
which system they have been exposed to by chance.
Note that, with features, we can directly account for the relationship between examples in which
constituents occupy different positions, such as the available books and the books available as in
(202).
(202) a. b.
Recall that such diagrams omit unbranching intermediate X levels, by convention. Thus, N2 is
present because it branches; but N1, not relevant here, is omitted. If the phrases were the available
history books and the history books available, then both N2 (dominating available) and N1
(dominating history) would be present in the diagrams.
In addition, individual levels must be recursive to distinguish units: [ N3 the [ N2 intelligent [ N2 young
[ N men]]]] contrasts with [ N3 the [ N2 young [ N2 intelligent [ N men]]]]. In both of these examples, there
are two successive, separate N2 levels distinguishing the meanings. In the first case, [ N3 the [ N2
intelligent [ N2 young [ N men]]]], one is talking about the young men who are intelligent. But in the
second, [ N3 the [ N2 young [ N2 intelligent [ N men]]]], one is taking about the intelligent men who are
young.
This method of recursion is also necessary to account for the use of ones in sentences like John met
the young men from Cleveland, and Bill met the old ones, where ones refers to the subunit men from
Cleveland. Such examples contrast with others like John met the young men from Cleveland, and
Bill met the ones from New York, where ones refers to the subunit young men.
215

216
Since all categories on the X2 Level are modifiers of the head, it no longer matters whether the
adjective precedes the noun or follows it. We simply state idiosyncratic facts about individual words
in the lexicon next to each lexical entry. This means that the word available has a dictionary
specification like [–VBL, –NML, +X2L, +PRH, +PSH] meaning that it is an X2 Level characterizer
(Adjective) that can occur in either prehead or posthead position.
We can extend this analysis to show the relationship between a courageous man and a man of
courage as in (203) and a Braille teacher and a teacher of Braille as in (204).
(203) a. b.
(204) a. b.
The structure of courage in (203b) is called a genitive of quality; notice that, as an X2 Level element,
it serves as a modifier of the head, exactly like the adjective courageous in (203a). The structures
in (204) show the relationship between compounds and complements. In English, this is most
clearly seen in nouns, although it exists for other categories like verbs (bartend versus tend bar) and
adjectives (life supporting versus supporting life).

The of in (203b) and (204b) is not a full preposition introducing a prepositional phrase. Actually,
it is little more than a phonological filler necessary in English to separate the head from posthead
constituents. In English, noun phrases can contain very long sequences of compound elements as
we have seen. Further, the last noun in a sequence of nouns is the head of the phrase and only the
head can carry number distinctions. Consider the following examples which show that only the head
in a sequence of nouns can show number distinctions and it is the head of the subject NP that the
verb agrees with:
(205) a. The kitchen appliance factory guarantee is/*are generous.
b. The kitchen appliance factory guarantees *is/are generous.
c. *The kitchen appliance factories guarantee is/are generous.
d. *The kitchen appliances factory guarantee is/are generous.
e. *The kitchens appliance factory guarantee is/are generous.
f. A ten thousand word paper is/*are required.
g. *A ten thousand words paper is/are required.
Also, as the following data show, the head (in italics) of an NP must be separated from posthead
constituents by a characterizer (double underlined):
(206) a. The votes against gun control *is/are greater than expected.
b. John’s reliance on his parents is/*are pathetic.
c. The bird in that tree is/*are a yellow bellied sapsucker.
d. The language teachers in the auditorium *is/are having a meeting.
If there is no overt characterizer separating the head noun from posthead elements, then of must be
inserted directly after the head. We have contrasts like the following:
(207) a. His love stories *is/are charming.
b. His love of stories is/*are charming.
c. The language teachers *is/are monitored by the principal.
d. The language of teachers is/*are monitored by the principal.
e. The number systems *is/are complicated.
f. The number of systems is/*are complicated.
g. The student letters *is/are very complimentary.
h. The student of letters is/*are very complimentary.
All of the above intricacies about English noun phrases seem to involve the rule of verb-subject
agreement. Without agreement, many English sentences are uninterpretable. There is a very large
number of words in English that can be either a noun or a verb, e.g., school, show, play, book, table,
chair, man, etc. Consider the following and observe the number on the verb:
217

218
(208) a. The school shows play here. (Subject is shows; the verb play is plural.)
b. The school show plays here. (Subject is show; the verb plays is singular).
c. The school shows plays here. (Subject is school; the verb shows is singular.)
d. The schools show plays here. (Subject is schools; the verb show is plural.)
Without agreement, the sentences are uninterpretable:
(209) a. *The school show play here.
b. *The schools shows plays here.
Since verbs agree with the head of the subject phrase, it is important that speakers be able to identify
the head. The three constraints mentioned above facilitate that identification: (i) only the head can
show number distinctions; (ii) the head must be separated from posthead constituents by a
characterizer; and (iii), if there is no characterizer after the head, then of must be inserted effectively
marking the head. It is interesting that the word that is inserted to mark the head is of, which is
perhaps the most generic preposition in English, often pronounced [].
Given this analysis, we might have an explanation for why the past tense has become regular in
English for all persons and numbers, but the present tense has not. Compare the following.
(210) a. Past Tense:
played for all persons and numbers: She/They played.
b. Present Tense:
plays for third person singular; play elsewhere: She plays versus They play.
The basic problem is that singular and plural noun inflection is identical to verb inflection:
(211) a. play: either a singular noun or a plural present
b. plays: either a plural noun or a singular present
There are many nouns in English that also function as verbs. On the other hand, there are only a few
past tense verbs that are also nouns, and they are all irregular, e.g., cut and hit. Further, locating the
head of an English noun phrase is not a trivial matter. There are many opportunities for ambiguity.
All these potential ambiguities are eliminated by the rule of subject verb agreement. Quite simply,
without the inflectional differences between singular and plural count nouns and between singular
and plural present verbs, speakers would be unable to locate the head of a noun phrase. Thus, the
present tense of verbs cannot become regular
We can generalize the above description of the English N3 to V3 and account for the very large
number of nominals that are related to verbs, e.g., destruction and destroy, evolution and evolve,
advancement and advance, shipment and ship, etc. As we have seen, this is essential if we are to
generalize notions like subject–of and object–of across categories: we want to be able to relate the

N3 his insistence on that to the V3 he insisted on that, the N3 her unexpected refusal to comply to
the V3 she unexpectedly refused to comply, and so on. Consider the following representations which
illustrate the close relationships between V3 and N3:
(212) The Mafia illegally shipped drugs into the country during June.
(213) the Mafia’s illegal shipments of drugs into the country during June...
219

220
The representations in (212) and (213) make significant improvements in the search for descriptive
adequacy. For example, notice that such representations as the above make possible a generalized
statement of subject and object. The subject N3 (the Mafia) has the same relationship to the V3 head
shipped in (212) as it has to the N3 head shipment in (213). Similarly, the object N3 (drugs) has the
same relationship to both V1 in (212) and N1 in (213); the adverb illegally is in the same relationship
to shipped as the adjective illegal is to shipment. In short, the RG framework goes a long way in
solving the problems in previous discussions.
In (212), PST is the morpheme for the past tense discussed earlier (see page 177). Thus, TENSE
markers in V3 occupy the same residence as DETERMINERS and POSSESSIVES in N3. In (213),
POS stands for the possessive suffix, which is usually either [s] as in book’s (the book’s cover), [z]
as in John’s (John’s son), or [z] as in Bess’s (Bess’s boss). Notice that, in forming a possessive like
Mafia’s, the noun Mafia is under the N3 and POS is under the C3. This is necessary because English
contains expressions like the boy with the stick’s hand is sore, where the possessive phrase consists
of the N3 the boy with the stick followed in a separate node by the possessive suffix. Notice that we
do not say *The boy’s with the stick hand is sore.
With these improvements, we are able to represent a complex example like (137h) as follows:
(214) all those ten recently acquired Impressionist paintings of landscapes of the Louvre’s on
display, which are so beautiful...

Another example of the relationship between V3 and N3 is the following:
(215) The friendly flight attendant patiently described the safety precautions to the nervous
passengers in their native language.
(216) the friendly flight attendant’s patient description of the safety precautions to the nervous
passengers in their native language...
221

222
Observe again that, by convention, nodes do not appear in trees explicitly unless they branch:
(217) this very young student (218) this Renaissance history student
(219) this very curious very young Renaissance history student

In all the representations above, we see that specifiers are higher in the tree structures than modifiers
and modifiers are higher in the structures than compounds and complements. The phrase “higher”
is formally expressed in linguistics as the C–COMMAND (for CATEGORY COMMAND)
relationship, defined as follows:
(220) A syntactic category C–commands a syntactic category if does not dominate , if
does not dominate , and if the first branching syntactic category n dominating also
dominates .
Thus, we say that specifiers C–command modifiers and modifiers C–command compounds and
complements. This three level version of X–bar theory has been thoroughly examined in English
(Binkert 1984) which is a head–center language, as well as in languages like Thai, where the head
is initial, and Newari, where the head is final. Despite the different positions of the head in the
phrase, the C–command requirements regarding specifiers, modifiers and complements are the same
as in English. The data supporting this for N3 are given in (221) and (222).
(221) Thai (data from Deepadung, 1989).
a. nók phíraâp tua lék tua nán
bird kind of bird small that
‘that small pigeon’
b. rôm kradaàt khan yày khan nií
umbrella paper big this
‘this big paper umbrella’
(222) Newari (data from Tuladhar, 1985).
a. Jigu thwa Râmã dekâbyugu chakhâ cheÞ
my this Ram build one house
‘this one house of mine which Ram built’
b. Râmyâ wa murkhamha Bengâli pâsâ
Ram’s that stupid Bengali friend
‘that stupid Bengali friend of Ram’s’
c. mistegu nhugu dhebâ kame yâgu hak
women’s new money earn to do right
‘women’s new right to earn money’
In these examples, we see determiners and possessives C–commanding modifiers, and modifiers
C–commanding complements and all compounds.
223

224
In Japanese, a head final language, we find the same C–command relations realized, although the
language is frequently described as nonconfigurational. The most interesting parallel involves the
postposition no (‘of’), which can signal either a possessive, an agentive modifier or a complement.
Thus, a phrase with the head e (‘picture’), like (223), can mean any of the given translations, similar
to the English noun phrase John’s picture.
(223) John–no e
a. the picture which John has
b. the picture painted by John
c. the picture with John in it
Crucially, a double occurrence of –no, as in (224), can only mean one of the three given translations
(data verified by Yukari Mori and Keiko Noji, both native speakers).
(224) Mary–no John–no e
a. the picture painted by John which Mary has
b. the picture with John in it which Mary has
c. the picture with John in it which Mary painted
d. *the picture painted by Mary which John has
e. *the picture with Mary in it which John has
f. *the picture with Mary in it which John painted
These data are exactly what our C–command relationships predict. If possessives must C–command
modifiers and modifiers must C–command complements, then the only possible combinations are
(225).
(225) a. POSSESSIVE + MODIFIER (cf. 224a)
b. POSSESSIVE + COMPLEMENT (cf. 224b)
c. MODIFIER + COMPLEMENT (cf. 224c)
Furthermore, (226a) can only mean (226b).
(226) a. Mary–no John–no Bill–no e
b. the picture of Bill that John painted that Mary has
Given these examples, we might propose that, in the unmarked case, the C–command relations are
universal: specifiers must C–command modifiers, and modifiers must C–command compounds and
complements. Further, on any given level, [+NML] categories precede [–NML] categories, the order
in noun phrases being especially rigid. In particular, the ordering restrictions seen in examples like
these five large houses are stated as a universal by Greenberg (1963:87):

(227) Greenberg’s Universal 20.
225
“When any or all of the items (demonstrative, numeral, and descriptive adjective) precede
the noun, they are always found in that order. If they follow, the order is either the same
or its exact opposite.”
Hawkins (1983: 117–120) discusses some problems with (227) and provides the following
illustrative examples:
(228) a. Three modifiers on the left/none on the right.
Dem–Num–Adj–N English, Finnish, Hindi, Hungarian, Mandarin, Maung
b. Two modifiers on the left/one on the right.
Dem–Num–N–Adj French, Italian
Dem–Adj–N–Num No examples
Num–Adj–N–Dem No examples
c. One modifier on the left/two on the right.
Dem–N–Adj–Num Kabardian, Warao
Num–N–Adj–Dem Basque, Easter Island, Indonesian, Jacaltec, Maori,
Vietnamese, Welsh
Adj–N–Num–Dem No examples
d. No modifiers on the left/three on the right.
N–Adj–Num–Dem Selepet, Yoruba, Akan
On the basis of the data in (228), Hawkins (1983: 119–120) revises (227) to (229).
(229) “When any or all of the modifiers (demonstrative, numeral, and descriptive adjective)
precede the noun, they (i.e., those that do precede) are always found in that order. For
those that follow, no predictions are made, though the most frequent order is the
mirror–image of the order for preceding modifiers. In no case does the adjective
precede the head when the demonstrative or numeral follow.”
This revision is consistent with the observations made here. It is further supported by recent work
in generative syntax (Chomsky 1995; Haegeman 1994; Carnie 2012). Even though adjuncts, i.e.,
modifiers, are not accorded their own level distinct from complements in some other theories of
X–bar syntax, adjuncts are always on higher recursions of X´ than compounds and complements.

226
The complete RG morphosyntactic feature matrix is summarized in Figure Eight. Elsewhere in this
text, syntactic categories are often referred to using the following abbreviations:
V: VERB ADV: MANNER ADVERB
AUX : AUXILIARY PPN: PREPOSITION
MOD : MODAL SCJ: SUBORDINATING CONJUNCTION; SUBORDINATOR
N: NOUN CCJ: COORDINATING CONJUNCTION; COORDINATOR
PRO: PRONOUN CPL: COMPLEMENTIZER
QNT: QUANTIFIER DET: DETERMINER
ADJ: ADJECTIVE DGR: DEGREE WORD
One of the most significant advantages of the above framework is that it enables us to state the
formal universal (230):
(230) [+NML] categories must precede [–NML] categories in both prehead and posthead position
on every X level.
Given the matrix in Figure Eight, the potential difficulty of constructing a revealing set of phrase
structure rules is solved. All possible phrase structure sequences in English are reducible to one
PHRASE STRUCTURE FILTER (PSF), namely, (231a) which is equivalent to the tree in (231b).
n m
(231) a. [X ([+NML]) ([–NML]) X ([+NML]) ([–NML]) ]
b.
where m n
(231) asserts that each X level must dominate a lower X level with an equal or lesser numerical
value, e. g., X2 over X2, X1, or X0 but not X2 over X3. Further, (231) asserts that each X level may
have at most one [+NML] and one [–NML] category, in that order, to the right and/or to the left of
the head X. Using this feature system, we can identify several natural classes of syntactic categories
by where they occur in phrase structure:
(232) a. N–SPECIFIER: a specifier that is a noun, e.g., a quantifier or a possessive noun
b. C–SPECIFIER: a specifier that is a characterizer, e.g., a determiner, degree
word, or complementizer
c. N–ADJUNCT: an adjunct that is a noun, e.g., a cardinal or ordinal numeral
d. C–ADJUNCT an adjunct that is a characterizer, e.g., an adjective, adverb, or
preposition

FIGURE EIGHT: FEATURES FOR SOME ENGLISH MORPHOSYNTACTIC CATEGORIES
CATEGORIES & EXAMPLES VBL NML OPH OCL PRH PSH ENH EVH ECH X1L X2L X3L
VERB: go, goes, went + – + + – + + + + + + +
AUX: have, has, had + – – – + + – + + – – –
MODAL: can, could, must + – – – + – – – + – – –
NOUN: boy, milk, Mary – + + + + + + + + + + +
PRONOUN: she, he, herself – + – – + + + + + + + +
QUANTIFIER: all, both, half – + – – + – + – – – – +
NUMERAL: three, third – + + – + – + – – – + +
ADJECTIVE: happy, big – – + + – + + + – + + –
MANNER ADVERB: happily – – + + + + – + – – + –
SENTENCE ADVERB: probably – – – + + + – + + – – +
DETERMINER: a, the, that, this – – – – + – + – – – – +
DEGREE WORD: so, too, that – – – – + – + – – – – +
COMPLEMENTIZER: that, if – – – – + – – + – – – +
EMPHATIC: so, too, not – – – – – + – – + – – +
PREPOSITION: in, out, after – – + – – + + + + + + +
SUBORDINATOR: since, after – – + – – + + + + – + +
COORDINATOR: and – – – – – + + + + + + +
VBL: VERBAL inherently marked for distinctions in tense, aspect, voice, and/or mode
NML: NOMINAL inherently marked for distinctions in number, person, gender, and/or case
OPH: OPEN PHRASE occurring as the head of a phrase that freely contains other elements
OCL: OPEN CLASS unlimited in number
PRH: PREHEAD can freely occur before the head of a phrase
PSH: POSTHEAD can freely occur after the head of a phrase
ENH: ENVIRONMENT OF N can freely occur immediately dominated by N3, N2, or N1
EVH: ENVIRONMENT OF V can freely occur immediately dominated by V3, V2, or V1
ECH: ENVIRONMENT OF C can freely occur immediately dominated by C3, C2, or C1
X1L: X1 LEVEL can freely occur immediately dominated by X1 (N1, V1, or C1)
X2L: X2 LEVEL can freely occur immediately dominated by X2 (N2, V2, or C2)
X3L: X3 LEVEL can freely occur immediately dominated by X3 (N3, V3, or C3)
227

228
Given (231), we have solved the categorial problem and can eliminate from our grammar complex
and unrevealing PS–Rules like those in (107). (231) also makes clear the relationship between a tree
diagram such as (231b) and an equivalent, linear representation such as (231a), referred to as
LABELED BRACKETS. For example, the following representations are equivalent; again, note
that labeled brackets always occur in left–right pairs as the arrows indicate:
(233) ¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯
¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯
¯¯¯¯ ¯¯¯¯¯¯ ¯¯¯¯¯¯¯ ¯¯¯¯¯
a. [ V3 [ N3 he] [ C3 PST] [ V1 [ V0 leave] [ N3 her] ] ]
b.
Note that her is the object of leave and is part of V1. As before, PST + leave = left.
Labeled brackets are difficult to read and, therefore, often simplified. For example,
[ N3 he] above is technically [ N3 [ N0 he]]; however, since the internal structure of the
noun phrase is clear, it is given in simplified form. If only the V1 is important to a
discussion, the above might even be simplified to [ he PST [ leave her]].
V3 V1
(234) ¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯
¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯
¯¯¯¯ ¯¯¯¯¯¯ ¯¯¯¯ ¯¯¯¯¯¯¯¯¯¯¯¯¯¯
a. [ V3 [ N3 he] [ C3 PST] [ V2 [ V0 leave] [ C3 happy/happily] ] ]
b.
Note that happy/happily is a modifier of the predicate leave and is part of V2 (the
distinction between X1 and X2 is discussed further on the next page).

As the above diagrams show, the most significant difference between RG and other versions of
X–Bar syntax is the level distinctions, in particular, the status of the X2 level. In all phrases, the X2
level contains descriptive modifiers of the head. These may be either characterizers or nouns;
compare (235a) and (235b).
(235) a. a courageous act b. an act of courage
In V3, the V2 level dominates manner adverbs and, in copulative constructions, predicate modifiers
of the subject:
(236) a. He acts courageous(ly). b. He is courageous.
The parallelism between V2 and N2 can be seen in the following representations:
(237) a. It smells bad. [ V3 [ N3 it] [ C3 PRS] [ V2 [ V0 smell] [ C3 bad] ]]
229
b. It has a bad smell. [ V3 [ N3 it] [ C3 PRS] [ V1 [ V0 have] [ N3 [ C3 a] [ N2 [ C3 bad] [ N0 smell]]] ]]
c. Its smell is bad. [ V3 [ N3 [ N3 it] [ C3 POS] [ N0 smell] ] [ C3 PRS] [ V2 [ C3 bad] ]]

230
The need for three X levels above the head X proceeds from the following assumptions we have
made all along:
(238) a. The rules of language are structure dependent, that is, they refer to structural units or
phrases.
b. The rules refer to whole phrases, not pieces of phrases. In terms of a tree diagram,
this means that rules make reference to nodes.
Given (238), consider the RG structure for this young history student:
(239)
Verification of each level proceeds as follows:
(240) a. The entire phrase (the whole N3) is replaced in the sentence This young history
student said that he would pass. (he = this young history student)
b. The anaphor one can occur with N2 elements: John met this young history student,
and Mary met that one. (one = young history student)
c. Conversely, one cannot occur with N1 elements: *John met a history student, and
Bill met a biology one. (one = student)
The need for the distinction between N2 and N1 is confirmed by a phrase like the Spanish Spanish
teacher. Without two levels below the, there would be no way to disambiguate the senses of
Spanish. The first one means from Spain; the second, of Spanish. Thus, the phrase equals the
teacher of Spanish from Spain (compare (241a) and (241b)).

(241) a. the Spanish Spanish teacher b. the teacher of Spanish from Spain
Residency requirements for this phrasal architecture predict the grammaticality of (242a) and (242b),
and the ungrammaticality of (242c) and (242d).
(242) a. the French math teacher
b. the teacher of math from France
c. *the math French teacher
d. *the teacher from France of math
In addition, these structures account for the ambiguity of phrases like the practical nurse, an
intellectual historian, and the issue of student grants. All of these facts result from the following
residency requirements of English: Descriptive modifiers of the head must hang from X2; elements
of compounds and complements must hang from X1.
Lastly, observe that this system allows the first member of a compound noun to be any one of the
three major syntactic categories: in math teacher the noun head (teacher) is preceded by another
noun (math); in physical therapist the head noun (therapist) is preceded by a characterizer (physical);
and, in go cart, the head noun (cart) is preceded by a verb (go).
231

232
The three level hypothesis for V3 can be confirmed with reference to the following structure:
(243) Sue will get the mail on Monday.
Given (238), verification of the three V nodes proceeds as follows:
(244) a. The existence of the V3 level is indicated by its pronominalization in the following
sentence: Sue will get the mail on Monday; at least, that’s what Bill says. (that =
Sue will get the mail on Monday)
b. The existence of the V2 level is indicated by its omission in the following sentence:
Sue will get the mail on Monday, and Mary will ___ too. (get the mail on Monday
is left out.)
Further, V2 level elements can occur after the pro–form do so as in the following:
Sue will get the mail on Monday, and Mary will do so on Tuesday. (do so = get the
mail)
c. The existence of the V1 level is indicated by its omission in the following sentence:
Sue will get the mail on Monday, but Mary won’t ____ until Tuesday. (get the mail
is left out.)
Further, V1 elements can NOT occur after the pro–form do so as the following
indicates: *Sue will get the mail on Monday, and Mary will do so the pay checks on
Tuesday.

The existence of V1 (direct object and object complement position) as opposed to V2 (predicate
nominative position, referring back to subject) also resolves the ambiguity in (245).
(245) a. John left the house messy.
[1] messy an object modifier:
[ [ John] [ PST] [ [ [ leave] [ the house] [ messy ] ] ] ]
V3 N3 C3 V2 V1 V0 N3 C3
[2] messy a predicate adjective, subject to agreement rules:
[ [ John] [ PST] [ [ [ leave] [ the house] ] [ messy] ] ]
V3 N3 C3 V2 V1 V0 N3 C3
b. John defied Bill to get even.
[1] to get even a complementary infinitive with subject Bill:
[ [ John] [ PST] [ [ [ defy] [ Bill] [ to get even] ] ] ]
V3 N3 C3 V2 V1 V0 N3 C3
[2] to get even a purposive infinitive with subject John:
[ [ John] [ PST] [ [ [ defy] [ Bill] ] [ to get even] ] ]
V3 N3 C3 V2 V1 V0 N3 C3
233

234
c. The boys made good models.
[1] make a transitive verb like build:
[ [ the boys] [ PST] [ [ [ make] [ good models] ] ] ]
V3 N3 C3 V2 V1 V0 N3
[2] make a copulative verb like become:
[ [ the boys] [ PST] [ [ [ make] ] [ good models] ] ]
V3 N3 C3 V2 V1 V0 N3
In addition, the V1/V2 distinction explains the grammaticality of (246), and predicts the
ungrammaticality of (247).
(246) a. John ate the food raw, nude.
b. John left the house unlocked, drunk as a skunk.
(247) a. *John ate the food nude, raw.
b. *John left the house drunk, unlocked.
The V1/V2 distinction also accounts for the ambiguity of sentences like those in (248):
(248) a. They left good friends.
b. They left the country great men.
c. They considered the candidates very concerned about the future of the country.
The most striking feature of these examples is the parallelism between the N1/N2 distinction and the
V1/V2 distinction, clearly seen in the following:
(249) a. a young math teacher
[ [ a] [ [ young] [ [ math] [ teacher ] ] ] ]
N3 C3 N2 C3 N1 N3 N0
b. a teacher of math that young (is rare)
[ [ a] [ [ [ teacher] of [ math] ] [ that young] ] ]
N3 C3 N2 N1 N0 N3 C3
c. He taught math young.
[ [ he] [ PST] [ [ [ teach] [ math] ] [ young] ] ]
V3 N3 C3 V2 V1 V0 N3 C3
(250) a. a superb math teacher
[ [ a] [ [ superb] [ [ math] [ teacher ] ] ] ]
N3 C3 N2 C3 N1 N3 N0
b. a teacher of math that superb (is rare)
[ [ a] [ [ [ teacher] of [ math] ] [ that superb] ] ]
N3 C3 N2 N1 N0 N3 C3
c. He taught math superbly.
[ [ he] [ PST] [ [ [ teach] [ math] ] [ superbly] ] ] ]
V3 N3 C3 V2 V1 V0 N3 C3

3.13 TRANSFORMATIONAL GENERATIVE GRAMMAR
By themselves, phrase structure rules can describe a great number of the individual sentences that
occur in languages; however, if our corpus is expanded to include more types of constructions, it
appears that some relationships between sentences will not be able to receive adequate
characterizations with phrase structure rules alone. For example, consider sets of English sentences
like the following and note that each pair means basically the same thing:
(251) a. The city must burn this book.
b. This book, the city must burn.
(252) a. The senators agreed on one particular compromise.
b. On one particular compromise, the senators agreed.
(253) a. He will never be very strong.
b. Very strong, he will never be.
The construction illustrated in (251b), (252b), and (253b) is TOPICALIZATION, the placement
of a phrase in sentence initial position for emphasis. Such examples rarely occur in isolation, which
is why they sound somewhat strange. However, in context they are not unusual. For example,
during a discussion of the books to be burned, (251b) would sound quite appropriate.
Students often claim that they would never utter topicalized phrases, but careful attention to language
reveals that, in fact, they do. Further, whether one utters topicalized phrases or not is completely
irrelevant. The fact is that such sentences are immediately understood by native speakers in context.
Given this, the sentences become part of the corpus of data that a linguist must analyze. As our
discussion of prescriptive versus descriptive grammar in Chapter One indicated, modern linguistics
is concerned with describing what is possible, not what is considered correct.
These three pairs of sentences begin with phrases that occur before the NP subject phrase. In (251b),
the subject phrase is preceded by another N3; in (252b), by a prepositional phrase; and, in (253b),
an adjective phrase. We could accommodate these data by adding a PS-Rule like the following
where X3 is a phrase with a [–VBL] head (noun, preposition, or adjective):
(254) V3 (X3) + N3 + T3 + V2
But this will not do for several reasons. First, the optional phrases at the beginning of (254) are
understood as the complements of the verbs, that is, phrases that complete the meaning of the verbs.
For instance, this book in (251b) is the object of burn. Our structural definition of object says that
the object is the N3 dominated by the V1. While this definition holds for (251a), it fails for (251b)
because the initial NP of (254) is not dominated by V1. Furthermore, our structural definition of
subject is now compromised. Given (254), there are now two N3 dominated by V3. Worse yet, if
we adopt (254), the whole classification of English verbs comes to grief. Broadly speaking, there
235

236
are two types of verbs in English: TRANSITIVE VERBS like catch, which must take a direct
object; and, INTRANSITIVE VERBS like die, which never do:
(255) a. John caught Bill.
b. *John caught.
(256) a. John died.
b. *John died Bill.
If, to account for (257), we allow transitive verbs to occur in verb phrases without a following N3,
we lose the distinction between transitives and intransitives.
(257) Foul balls like that, no one can catch.
Somehow, we must indicate that the initial phrases in (251b), (252b), (253b), and (257) are
complements of the main verb. Since a phrase structure solution via (254) failed, another method
must be sought.
PS–RULES generate tree structures with phrases in specific positions. Suppose that we now allow
grammars to contain other rules that move elements from such positions to new positions, that is,
transform one structure into another. For example, suppose that, as before, the PS–RULES in (107)
generate a structure like (258).
(258)
Now suppose a TRANSFORMATIONAL RULE moves the object NP to the front of the sentence
leaving behind a TRACE (a kind of footprint) of that object in the form of a co–indexed NP. The
result is (259), where "[e]" indicates the EMPTY NP TRACE designating the original home of the
object.

(259)
This produces the sentence (260a) or, more accurately, (260b).
(260) a. This book, the city must burn.
b. [ this book] the city must burn [ e]
NP i NP i
A crucial point in the representation (260b) is that there really is a gap in the sentence at the point
of the [e]. When speakers hear a sentence that begins with a “displaced” object, they must wait for
the gap in the sentence before they can assign an interpretation to that “displaced” phrase. Consider
(52).
(261) a. This book, you absolutely must try to get your students to convince each other to tell
their friends to read.
b. [ NP this book] i you absolutely must try to get your students to convince each other to
tell their friends to read [ e]
NP i
Given representations like the above, we can say that [ NP e] i is bound to [ NP this book] i via
REFERENCE, not unlike the way pronouns are bound to their referents in sentences like John i saw
himself in the mirror.
i
A grammar that contains transformational rules is called a TRANSFORMATIONAL
(GENERATIVE) GRAMMAR (TG). In such grammars, a sentence like (260a) is associated with
two structures: the structure generated by the phrase structure rules, that is, (258); and the structure
generated by the transformational rules, that is, (259). Technically, we say that phrase structure rules
generate the DEEP STRUCTURE of sentences such as (258), and transformational rules generate
the SURFACE STRUCTURE such as (259). Crucially, functional relations like subject and object
are not altered by transformational rules. However, the rules do switch the order of constituents in
a sentence meaning that the order in the deep structure can be different from the order in the surface
structure. More importantly, our description of the grammars of human languages has now changed.
Originally, we claimed that human languages could be described by PS-Rules alone. Now, we must
say that they require an additional rule, the Transformational Rule or T-Rule.
237

238
3.14 NONTRANSFORMATIONAL GENERATIVE GRAMMAR
With transformational rules, we say that human languages cannot be adequately described without
the following two assumptions:
(262) a. Every sentence has two structural levels, an abstract underlying structure roughly
equivalent to what is understood, and a real superficial structure roughly equivalent
to what is heard.
b. The order of phrases in the abstract underlying structure can be different from the
order of phrases in the real superficial structure.
It seems quite clear that (262a) must be assumed. The syntactic processes of REFERENCE,
OMISSION, and PLACEMENT demand it. For example, there is no way to understand the
superficial structure (263a) without the underlying representation (263b).
(263) a. John will sing if you ask him to.
b. John will [VP sing ] if you ask him to [VP e ]
i i
On the other hand, (262b) is not a necessary assumption. That assumption is linked to
transformational grammar. Rather than postulating a new rule type that moves constituents to
different positions leaving behind a co-indexed empty category, supposed we allow [e] to be directly
generated as follows:
(264) XP [e]
If we allow empty phrases to be generated directly via PS–Rules, then we can give up (262b). All
that remains is to propose principles for finding the referent for [e] something that we need to do
anyway to account for sentences like the following which contain co-indexed constituents:
(265) a. After John i got up, he i had a big breakfast.
b. After he i got up, John i had a big breakfast.
c. First John entered the room, and then the SOB insulted the hostess.
i i
Therefore, we do not need transformational rules at all. What we need is a mechanism for relating
[NP e] to full noun phrases, something that is needed anyway. Again, it is useful to stop at this point
and take stock of what we are trying to accomplish. Native speakers of every language completely
understand every possible sentence in their language. The sentence that you are now reading is a
sentence which you have probably never encountered before in your life, and yet you understand it
immediately as you read it. It must be that, at some unconscious level, you understand precisely the
relationships we are trying to uncover. The number of sentences in every language is infinite;
learning a language does not consist of memorizing a list of the possible sentences. Rather, it must
consist of learning the principles upon which every possible sentence is constructed. These

principles form native speakers’ unconscious knowledge of their language. Throughout this
discussion, we have been trying to ascertain just what this knowledge is, that is, what must native
speakers know to understand all the sentences we have been considering.
3.15 COMMAND RELATIONS IN PHRASE STRUCTURE
We are now in a position to formalize relationships like SUBJECT and OBJECT and to account for
so–called displaced phrases. Consider first (258) recast with the RG innovations as (266).
(266)
In a diagram like (266), it is possible to follow a path from one X category to another. Some paths
prove to be crucial in determining syntactic relationships. One particularly useful relationship in
syntax is the C–COMMAND (for CATEGORY COMMAND) relationship, which we defined above
(see page 223) and repeat here for convenience:
(267) A syntactic category C–commands a syntactic category if does not dominate , if
does not dominate , and if the first branching syntactic category n dominating also
dominates .
Applying (220, 267) to the diagram in (266), we see that the N3 the city C–commands the verb burn
because the node immediately above that N3 is the top V3 node and that top V3 node dominates the
verb burn. Similarly, the N3 this book C–commands burn because the node immediately above that
N3 is the V1 in the diagram and that V1 dominates the verb burn.
As we will see in Chapter Five, this relationship is at the basis of all the rules for finding the
referents of pronouns in English. For now, we will use it to formally define the functional notions
subject–of and object–of:
(268) a. The subject of a node â is the N3 which immediately C–commands and precedes â.
b. The object of a node â is the N3 which immediately C–commands and follows â.
239

240
The phrase “immediately C–commands” means ‘is closest to’ or ‘is separated by the fewest number
of intervening nodes in the tree’ (see page 383 for a formal definition of “immediate C–command”
and pages 381 to 386 for other technical terms).
Given (220, 267) and (268), we determine the subject and object of a verb such as the one diagramed
in (266) as follows:
(269) a. The N3 above city C–commands and precedes the verb burn. The N3 is á; the V0
node burn is â; the node immediately above á, which is the topmost V3 in the
diagram, dominates â. Since á precedes â, á (the city in this example) is the subject
of â (burn).
b. The N3 above book C–commands and follows the verb burn. The N3 is á; the V0
node burn is â; the node immediately above á, which is V1, dominates â. Since á
is to the right of â, á (this book in this example) is the object of â (burn).
The same definitions will work for all the other relevant diagrams in this text. For example, The
Mafia is the subject and drugs the object of shipped in (212) and shipment in (213); and Sue is the
subject and mail the object of get in (243).
As we have seen, RG permits the terminal nodes of (231) to dominate two kinds of empty categories,
that is, categories that are phonologically null: [u], which is an understood element in a sentence like
Let’s eat [u]; and [e], which is a bound ANAPHOR, that is, an item that requires a referent. The
availability of [e] together with (220, 267) will account for the interpretation of so–called displaced
constituents. Consider again (260) recast with the RG innovations as (270).
(270)

In (270), as in (266), city is the subject of burn. Importantly, book is not the subject. Although the
N3 above book C–commands burn, it does not immediately C–command it. The N3 above city is
closer. In (270), the empty N3 ([e]) immediately C–commands and follows burn, so it is the object;
however, it is also an anaphor, so it must have a referent. To find the referent, we use the following
concepts:
(271) IMMEDIATE NEIGHBORHOOD: An immediate neighborhood embraces all phrases
that are immediately C–commanded by the same C–Specifier (DET or POS in N3; CPL or
TNS in V3).
(272) EXTENDED NEIGHBORHOOD: An extended neighborhood embraces all X3
recursions dominating the immediate neighborhood.
(273) THE EMPTY CATEGORY CONDITION (ECC): An empty category and its referent
must have the same syntactic feature matrix, that is, they must be categories of the same
type.
(274) THE EXTENDED NEIGHBORHOOD CONDITION (ENC): A WH-phrase or focused
phrase in the extended neighborhood must be bound to an [e] that it C–commands.
Given these specifications, consider (270). The tense characterizer PRS is a C–Specifier; therefore,
all items within the first V3 above it are in its immediate neighborhood. This includes the anaphor
[ N3 e]. The phrase this book is in a V3 recursion above the immediate neighborhood; hence, it is in
an extended neighborhood. Both this book and the [e] are phrases of the same type. Both the ECC
and the ENC are met; accordingly, this book is the referent for [e] and the object of burn even though
it precedes the subject.
While the definitions and conditions presented in this section have been simplified for expository
purposes, they nonetheless satisfy all the major difficulties that have been presented in previous
sections. There are, quite naturally, some more difficult examples to be considered, but the essential
direction of the solution presented here is correct and extends, with some elaboration, to other very
complicated examples.
The crucial result of our investigation is that it is possible to formalize even complex relationships
in terms of linear and hierarchical structure. We have succeeded in reducing all phrases to variations
on one frame (231). Furthermore, we have provided formal, explicit definitions of relations like
SUBJECT, OBJECT, REFERENCE, and so on. We have done this without transformational rules.
In short, our entire solution is base on the concept phrase that is rooted in human cognitive capacity.
241

242
3.16 EMPTY CATEGORIES
As we have seen, sentences sometimes contain missing elements or gaps which speakers must fill
either from information elsewhere in the sentence or from some general principles. Examples of
sentences with gaps are the following:
(275) a. This book, you really should read ____.
b. He shrugged ____.
Native speakers know intuitively that the gaps in the above sentences are understood as follows:
(276 ) a. This book, you really should read it.
b. He shrugged his shoulders.
There are two major types of empty categories in language: [e], a bound element that has a referent
somewhere inside the same sentence, and [u], an understood element. Consider the following
examples, where [1] is the sentence as it is spoken, where [2] indicates the way it must be
understood, and where the subscripts indicate which two elements are bound (refer to each other):
(277) a. [1] Who will cook the dinner?
[2] [ V3 Who i [ V3 [e] i will cook the dinner] ]
(The [e] is bound to who, the logical subject of cook.)
b. [1] What will he cook today?
[2] What i will he cook [e] i today.
(The [e] is bound to what, the logical object of cook.)
c. [1] He shrugged.
[2] He shrugged [u].
(The [u] indicates some understood phrase.)
Note that [e] is bound to some referent elsewhere in the sentence, but [u] is unbound. In other
words, the meaning of [e] is recovered from information given elsewhere in the sentence, whereas
the meaning of [u] must be supplied by some other general principles. In the case of (277c), shrug
belongs to class of English verbs, body-part verbs, which allow the omission of a specific body-part.
Membership in this class must be learned. For example, the class includes blink, clap, nod, and
wave, but not braid, crane, pluck, or sprain. While it’s possible to say She blinked, it is not possible
to say *She sprained. Speakers must simply learn which verbs allow an understood body-part.
As the above examples indicate, an empty category can only occur in a position that can contain a
fully specified phrase. Each [e] and [u] above can be replaced by a fully specified phrase. [u] occurs
in a position that can contain an NP; [e] occurs in a position that can contain its referent.
Conversely, [e] and [u] cannot occur where fully specified phrases cannot occur:

(278) a. John persuaded Bill that Mary should go.
b. *John persuaded Bill Mary to go.
c. *Who did John persuade Bill [e] to go?
i i
(279) a. He donates money to charities.
b. He donates [u] to charities.
c. *He donates charities money.
d. *He donates charities [u].
The data above indicate clearly that grammars must contain empty categories as well as rules for
where empty categories can or must occur. As a result, we have postulated above both the Empty
Category Condition (ECC) and the Extended Neighborhood Condition (ENC).
Understandably, the matter can be complicated. Consider, for example, the following sentence:
(280) Who did John promise that Bill would visit?
This sentence has two possible meanings, which are disambiguated by the following representations:
(281) a. Who i did John promise [e] i that Bill would visit [u]?
b. Who did John promise [u] that Bill would visit [e] ?
i i
In (281a), the question being asked is who John promised, whereas in (281b) the question being
asked is who Bill would visit. Note that only [e] is bound to who, and that empty categories can
occur in the middle of sentences.
For a sentence like Who did John promise that bill, the grammar must generate the following:
(282) Who i did John promise [e] i that bill?
Note that the grammar does not generate (283) because then there would be no slot for an [e] bound
to who, and the ENC requires a WH-phrase to be bound to an [e]:
(283) *Who did John promise [u] that bill?
Declarative variants of (283) do tolerate a [u] in object position for some speakers:
(284) a. John promised me that bill.
b. John promised [u] that bill. (cf. I never promised you/[u] a rose garden.)
Crucially, [e] and [u] must occur in every WH–question where a verb allows a [u] object; otherwise,
grammars would not produce the correct reading for classic examples like the following:
243

244
(285) a. What i did the teacher read [u] to the children from [e] i?
b. *What i did the teacher read [e] i to the children from [u]?
c. *The teacher read to the children from [u].
An important feature of empty categories is that they block contraction so that grammars of natural
languages must contain the following condition:
(286) THE CONTRACTION BLOCK CONDITION (CBC)
Contraction is not possible over an empty category.
Examples of this condition involving [e] include the following:
(287) a. Who i did you want [e] i to win the race?
b. *Who did you wanna win the race?
(288) a. Could i we [e] i have gone there?
b. *Could we’ve gone there? (cf. We’ve gone there.)
The empty [e] found in passive constructions also blocks contraction:
(289) a. He i was dared [e] i to go.
b. *He was [dæ] go. (cf. Did he [dæ] go?)
(290) a. She i was allowed [e] i to go.
b. *She was [law] go.
The empty category [u] also blocks contraction:
(291) a. Sue will now dress [u] to get ready for the prom. (reflexive object understood)
b. *Sue will now [drs] get ready for the prom.
(292) a. They met [u] to discuss the wedding. (reciprocal object understood)
b. *They [m] discuss the wedding.
(293) a. Sue will nod [u] to indicate her approval. (understood body part)
b. *Sue will [na] indicate her approval.
(294) a. They eat [u] to live. ([u] = some type of food)
b. *They [i] live.

The CBC also extends to idioms such as the following where a very specific N3 is understood:
(295) a. Did they get a chance to go?
b. Did they get [u] to go?
c. Did they get a go? (= ‘Did they get a go ahead?’)
d. Did they [g] go? (= ‘Did they get a go?’ not ‘Did they get to go?’)
Examples like (295) contrast with the following where no empty category occurs:
(296) a. They have got to go. (= ‘They must go.’)
b. They have got a go. (= ‘They have got a go ahead.’)
c. They’ve [ga] go. (= ‘They have got to go’ or ‘They have got a go.’)
Consider now a sentence like (297), which must be related to the sentences in (298).
(297) When was Sally interviewed?
(298) a. Someone interviewed Sally at a particular time.
b. Sally was interviewed.
c. Someone interviewed Sally.
The question (297) is clearly related to the statements in (298). Notice that (297) presupposes, in
fact, that there was someone who actually interviewed Sally at a particular time. As a result, the
grammar should give (297) the representation in (299).
(299) When i was j Sally k [e] j interviewed [e] k [e] i?
a. The empty category with subscript “i” marks the ordinary position of a phrase
expressing time (compare (298a)).
b. The empty category with subscript “j” marks the ordinary position of the auxiliary
in a sentence (compare (298b)).
c. The empty category with subscript “k” marks the ordinary position of a direct object
(compare (298c)).
It is clear from examples such as the ones above that empty categories are necessary to give
sentences the full interpretation of their meaning. All native speakers know that there is a gap in a
sentence like He can come if he wants to, and that the gap can only be filled by come. As a result,
the complexity that empty categories introduce into a grammar is unavoidable if one wishes to impart
meaning to sentences.
245

246
Note that the matter is not simply one of providing a meaningful interpretation to a sentence. Empty
categories are justified in representations on purely linguistic grounds. Observe that the following
sentences are ungrammatical:
(300) a. *He wants to.
b. *He admires.
(300a) is ungrammatical because the infinitive construction (the one introduced by to) requires a verb
phrase after the to; similarly, (300b) is ungrammatical because admire is a transitive verb, meaning
it requires a direct object. The following representations satisfy these constraints:
(301) a. He can come i if he wants to [e] i. ([e] i = come)
b. Who does he admire [e] ?
i i
As these examples indicate, inserting empty categories into syntactic representations is a necessity.
Actually, it is a biological necessity since human beings are not clairvoyant. Elements cannot be left
out of sentences unless there are principles that can recover them, because human beings can’t read
minds. In fact, we can propose the following universal condition operative in all human languages:
(302) THE RECOVERABILITY OF DELETION CONDITION (RDC)
All gaps in sentences must be recoverable.
The ECC, the CBC and the RDC are natural conditions for a grammar to contain. Human beings
are not clairvoyant (generally). Since the ECC stipulates that an empty category can only occur in
a position where a fully specified category can occur, empty categories are associated with positions
where speakers expect something to occur. The RDC stipulates that an empty category in those
positions must be recoverable. The position occupied by [u] is one that is usually filled by a fully
specified phrase; the one occupied by [e] is the position occupied by its referent in a given sentence.
Contraction over empty categories would have the effect of eliminating those positions; in essence,
speakers would be unable to fill the empty positions with appropriate phrases because the empty
positions would have disappeared in contraction.

3.17 WORD ORDER VARIATIONS
In this section, we will consider three variations in word order which appear to seriously challenge
an analysis of syntax that does not allow movement rules, that is, a nontransformational grammar.
3.17.1 ADVERBS IN FRENCH AND ITALIAN
Let us consider first the apparent C–command violations in the position of manner adverbs and
temporal quantifiers in French and Italian. Basically, the problem is that these languages permit
manner adverb and temporal quantifiers to occur between a verb and its direct object, an apparent
violation of the C–command relation that exists between modifiers and complements as stated above.
Consider the following examples:
(303) a. Jean embrasse Marie souvent.
b. Giovanni bacia Maria spesso.
c. John kisses Mary often.
(304) a. Jean embrasse Marie passionnément.
b. Giovanni bacia Maria appassionatamente.
c. John kisses Mary passionately.
(305) a. Jean embrasse souvent Marie.
b. Giovanni bacia spesso Maria.
c. *John kisses often Mary.
(306) a. Jean embrasse passionnément Marie.
b. Giovanni bacia appassionatamente Maria.
c. *John kisses passionately Mary.
Examples (303), (304), (305c) and (306c) do not present any problems. They follow the
C–command relation we have articulated, namely, modifiers must C–command complements. The
question is how to analyze (305a), (305b), (306a) and (306b). Basically, we have two options given
the structures we have proposed.
First, following William’s analysis of French (1994: 174 ff.), we might say that the modifiers in
(305a), (305b), (306a) and (306b) reside in an “intraposed” position as follows, where the modifier
is bound to an [e] on the V2 level:
(307) [ V2 [ V1 [ V0 [ V0 verb] [ X3 modifier] i ] [ N3 complement] ] [e] i ]
An intraposition analysis, with or without an associated [e], is certainly justifiable for some English
sentences. Consider the following:
247

248
(308) a. John pushed the door open.
[ [ John] [ PST] [ [ push] [ the door]] [ open]]
V3 N3 C3 V1 V0 N3 C3
b. John pushed open the door.
[ [ John] [ PST] [ [ [ push] [ open]] [ the door]]
V3 N3 C3 V1 V0 V0 C3 N3
(309) a. John took the trash out.
[ [ John] [ PST] [ [ take] [ the trash]] [ out]]
V3 N3 C3 V1 V0 N3 C3
b. John took out the trash.
[ [ John] [ PST] [ [ [ take] [ out]] [ the trash]]
V3 N3 C3 V1 V0 V0 C3 N3
Notice that it is not necessary to assume a movement analysis in the above, that is, we do not have
to say that characterizers open and out are generated in one position from which they are moved by
transformation to the other. On the other hand, there is reason to assume that English resultative
expressions and particles have the “base home” in (308a) and (309a), respectively.
First, the positions in (308a) and (309a) are the most unconstrained. If the direct object is a pronoun
or the following characterizer phrase contains a modifier, then the structural alternatives in (308a)
and (309a) are the only ones possible:
(310) a. John pushed it open./John took it out.
b. *John pushed open it./*John took out it.
(311) a. John pushed the door right open./John took the trash right out.
b. *John pushed right open the door./*John took right out the trash.
Further, the “intraposition” of a C3 is only obligatory when the direct object is heavy:
(312) a. John heroically pushed open the door that was nailed shut from the inside by a
maniac on the run from the police for having committed some terrible crimes.
b. *John heroically pushed the door that was nailed shut from the inside by a maniac on
the run from the police for having committed some terrible crimes open.
(313) a. John has finally taken out the trash which had accumulated during his illness and had
become a health hazard for all of the people living with him.
b. *John has finally taken the trash which had accumulated during his illness and had
become a health hazard for all of the people living with him out.
Despite the availability of an intraposition analysis, the data suggest that we not use it to analyze the
apparent C–command violations in French and Italian for two reasons.

First, while English push open and take out seem like composite predicates, the French and Italian
examples above do not. Indeed, there may be a universal that prevents manner adverbs and temporal
quantifiers from forming composite predicates with a verb.
Second, while English does not tolerate multiword phrases in intraposition (311), French and Italian
do. Consider the following:
(314) a. Jean embrasse très passionnément Marie.
b. Giovanni bacia molto appassionatamente Maria.
As an alternative to an intraposition analysis, consider instead that the verb might be in a different
position than expected. Notice that auxiliary verbs is English are frequently positioned outside of
the verb phrase and in the tense characterizer. This occurs with negatives, emphatics, and quantifiers
and only with morphologically heavy verbs, that is, verbs inflected for person, number and tense like
the forms of the verb be (I am happy, she is happy, we are happy, he was happy, they were happy,
etc.). Compare the following examples noting the positions of the main verbs:
(315) a. [ V3 John [ C3 [ C0 modal
John will
[ C0TNS]] FUT
[ C3 C0]] [ V2[ V1 [ V0 verb] [ N3 complement]]]]
be happy
John did PST not become happy
John did PST so become happy
They will FUT all become happy
b. [ V3 John [ C3 [ C0 verb i [ C0TNS]] [ C3 C0]] [ V2[ V1 [ V0 [e] i] [ N3 complement]]]]
John isi PRS not [e] i happy
They werei PST all [e] i happy
They werei PST so [e] i happy
*They become PRS not [e] happy
i i
Again, notice that it is only morphologically heavy verb forms that occupy positions outside the verb
phrase. Here are other English examples in questions:
(316) a. Are they happy?
b. Will they become happy?
c. *Become they happy?
When there is no modal or auxiliary in English, a form of the “support verb” do must be added:
(317) a. They will be happy. They will not be happy. Will they be happy?
b. They are happy. They are not happy. Are they happy.
c. They become happy. They do not become happy. Do they become happy?
d. They become happy. *They become not happy. *Become they happy?
249

250
Returning to the main theme, suppose we assume that the French and Italian verbs actually reside
in the tense characterizer the way English auxiliaries do:
(318) [ V3 subject [ C3 [ C0 verb i [ C0 TNS]]] [ V2 [ X3 modifier] [ V1 [ V0 [e] i] [ N3 complement]]]]
If the French and Italian verbs are actually in the position of “verb” in the above structure, then the
“modifier” will follow the verb, and there is actually no C–command violation at all in any of the
examples we have given. This analysis not only solves the apparent problem, but also it seems
appropriate since main verbs in both French and Italian are all morphologically heavy (inflected for
PERSON, NUMBER, and TENSE) like the English verb be.
Given (318), we conclude that the French and Italian examples do not violate the universal
C–command relations we have proposed. Apparent counterexamples parallel English examples like
the following, where it only seems that the C–command relations have been violated.
(319) a. John is not/so/too/probably meditating.
b. The jars were (all) all broken.
We can extend the above analysis to account for the position of negatives and quantifiers in French
(see Pollock 1989), again not unlike English. Consider the following.
(320) a. Jean n’aime pas Marie.
John likes not Mary
‘John doesn’t like Mary.’
b. [ V3 [ N3 Jean] [ C3 ne [ C0 aime i [ C0 PRS]] [ C3 pas]] [ V1 [ V0 [e] i] [ N3 Marie]]]
As before, we do not need to assume any movement of the verb in the above structure. Similarly,
we can account for the position of tous ‘all’ to the right of the verb:
(321) a. Mes amis aiment tous Marie.
my friends like all Mary
‘My friends all like Mary.’
b. [ V3 [ N3 mes amis] [ C3 [ C0 aiment i [ C0 PRS]] [ N3 tous]] [ V1 [ V0 [e] i] [ N3 Marie]]]
Thus, we arrive at the same conclusion as Chomsky (1995:195): “French–type and English–type
languages now look alike...” Only we have achieved the same result as Chomsky without moving
anything anywhere. All the instances of [e] in the above structures are generated in place; they are
not the trace of movement.

3.17.2 NONCONFIGURATIONAL LANGUAGES
There seem to be clear cases of languages in which the order of sentential elements like subject (S),
verb (V), and object (O) cannot be related to phrasal architecture. Consider the following Latin
sentence which can occur in the six variations specified.
(322) Marcus Publium amat. ‘Marcus loves Publius.’
a. SVO: Marcus amat Publium.
b. SOV: Marcus Publium amat.
c. VOS: Amat Publium Marcus.
d. VSO: Amat Marcus Publium.
e. OVS: Publium amat Marcus.
f. OSV: Publium Marcus amat.
Allowing the variations for the subject and object in prehead and posthead positions, a basic three
level phrasal architecture in which specifiers must C–command modifiers and modifiers must
C–command complements will produce only the following orders, ignoring the tense characterizer
for the moment:
(323) a. SVO.
[ [ Marcus] – [ [ amat] – [ Publium]] ]
V3 N3 V1 V0 N3
b. SOV.
[ [ Marcus] – [ [ Publium] – [ amat]] ]
V3 N3 V1 N3 V0
c. VOS.
[ [ [ amat] – [ Publium]] – [ Marcus] ]
V3 V1 V0 N3 N3
d. OVS.
[ [ [ Publium] – [ amat]] – [ Marcus] ]
V3 V1 N3 V0 N3
However, if we allow the order of [+NML] and [–NML] categories to be free and further, following
the previous analysis for French and Italian, allow verbs to be embedded in the tense characterizer
when they are highly inflected, then all six of the possible orders can be represented in our system.
Notice that the V1 constituent in all of the alternatives in (324) is the same. Further, the subject is
always a V3 constituent and the internal structure of the tense characterizer is always the same.
(324) a. SVO.
[ [ Marcus] – [ [ [ [ amat] ][ PRS]]] – [ [ Publium][ [e] ]] ]
V3 N3 C3 C0 V3 V0 i C0 V1 N3 V0 i
b. SOV.
[ [ Marcus] – [ [ Publium][ [e] ]] – [ [ [ [ amat] ][ PRS]]] ]
V3 N3 V1 N3 V0 i C3 C0 V3 V0 i C0
251

252
c. VOS.
[ [ [ [ [ amat] ][ PRS]]] – [ [ Publium][ [e] ]] – [ Marcus] ]
V3 C3 C0 V3 V0 i C0 V1 N3 V0 i N3
d. VSO.
[ [ [ [ [ amat] ][ PRS]]] – [ Marcus] – [ [ Publium][ [e] ]] ]
V3 C3 C0 V3 V0 i C0 N3 V1 N3 V0 i
e. OVS.
[ [ [ Publium][ [e] ]] – [ [ [ [ amat] ][ PRS]]] – [ Marcus] ]
V3 V1 N3 V0 i C3 C0 V3 V0 i C0 N3
f. OSV.
[ [ [ Publium][ [e] ]] – [ Marcus] – [ [ [ [ amat] ][ PRS]]] ]
V3 V1 N3 V0 i N3 C3 C0 V3 V0 i C0
What is varying in the above structures is the order of the tense characterizer, the subject, and the
verb phrase. Earlier in this Chapter, we reduced all English phrases to variations on one PHRASE
STRUCTURE FILTER (PSF) which we repeat here as follows (see (231)):
n m
(325) [X ([+NML]) ([–NML]) X ([+NML]) ([–NML]) ]
where m n
For Latin, the left to right order of constituents in phrases is freer. Therefore, we can propose that
the language has the following PSF, where “” indicates that the order of adjacent items is free
n m
(326) [X ([+NML]) ([–NML]) X ]
where m n
We see examples of these variations in the following:
(327) Marcus Publium amat. ‘Marcus loves Publius.’
a. SVO: Marcus amat Publium. [ V3 N3 – C3 – V1]
b. SOV: Marcus Publium amat. [ V3 N3 – V1 – C3]
c. VOS: Amat Publium Marcus. [ V3 C3 – V1 – N3]
d. VSO: Amat Marcus Publium. [ V3 C3 – N3 – V1]
e. OVS: Publium amat Marcus. [ V3 V1 – C3 – N3]
f. OSV: Publium Marcus amat. [ V1 – N3 – C3]
Thus, we need not postulate any movement rules to obtain the six possible orders that occur in a
simple Latin clause. Further, by allowing [+NML] and [–NML] categories to freely occur in either
prehead or posthead positions we can also account for variations like the following:
V3

(328) ducentes fortes milites ‘those two–hundred brave soldiers’
a. ducentes fortes milites
b. ducentes milites fortes
c. fortes ducentes milites
d. fortes milites ducentes
e. milites fortes ducentes
f. milites ducentes fortes
Since all of the elements in the above alternatives are marked by the same case, number, and gender,
the integrity of the phrase is preserved in each variation. As we have seen, such conditions on the
integrity of phrases is demanded by human biology.
Similar remarks are appropriate for other languages with “free word order” like Japanese. Consider
the following examples from Kuno (1983: 3–4):
(329) a. John–ga Mary–o but–ta.
John–nominative Mary–accusative hit
‘John hit Mary.’
b. Mary–o John–ga but–ta.
c. *John–ga but–ta Mary–o.
d. *But–ta John–ga Mary–o.
According to Kuno (1983:4), (329c) and (329d) are ungrammatical “because they violate the
Verb–Final Constraint.” With regard to (329b), Kuno (1983, Page 4, Note 3) offers the following
comment: “According to a large–scale statistical study of sentence structure conducted by the
National Language Research Institute, Tokyo, Japan, the ratio of frequencies of occurrences between
the SOV word order and the OSV word order in Japanese is 17 to 1.”
Given Kuno’s remarks, suppose we say, first, that Japanese is a verb final language, and, second, that
(329b) is the result of a variation like topicalization in English. As we saw in the preceding sections,
variations in word order like those introduced by topicalization do not require a syntactic movement
rule. The PSF for Japanese is, therefore, like Latin, and no movement rules is necessary:
n m
(330) [X ([+NML]) ([–NML]) X ]
where m n
253

254
3.17.3 THE ORDERING OF V1 CONSTITUENTS IN ITALIAN AND HEBREW
A potentially more difficult problem for the nontransformational theory we have presented concerns
verbal complements in Italian and Hebrew. Basically, both languages allow structures of the form
(331b) as alternatives to (331a) (data in Italian and Hebrew from Belletti and Shlonsky 1995).
(331) a. [ VP V NP PP]
b. [ V PP NP]
VP
(332) a. ho messo [ NP quel libro] [ PP sul tavolo]
henaxti [ NP et ha–sefer ha–hu] [ PP al ha–šulxan]
I put [ the book] [ on the table]
NP PP
b. ho messo [ PP sul tavolo] [ NP quel libro]
henaxti [ PP al ha–šulxan] [ NP et ha–sefer ha–hu]
I put [ on the table] [ the book]
PP NP
Belletti and Shlonsky argue that (331a) is the base structure for verbal complements in both
languages. The initial independent motivation they give for this choice concerns double complement
idioms: in both Italian and Hebrew, such idioms only have the structure (331a):
(333) a. Ho messo i puntini sugli i
have put the dots on the i’s
‘I have clarified things.’
b. *Ho messo sugli i i puntini
have put on the i’s the dots
(334) a. sam et nafš–o be–xap–o
put acc life–his in–palm–his
‘He took his life in his hand.’
b. *sam be–xap–o et nafš–o
put in-palm–his acc life–his
Given such examples, Belletti and Shlonsky conclude that the ungrammatically of the PP NP order
in idioms “strongly suggests that there exists a linear order on the basis of which the idiom is
constructed and that this order corresponds to the one assumed” (Page 496), namely, (331a).
Actually, there are many alternative ways of accounting for the relationships between the above pairs
of nonidiomatic examples in Italian and Hebrew. Further, however one chooses to relate the two
structures in (331), the idioms must be treated exceptionally and tell us very little. We can outline

at least six possible ways of dealing with the data as follows, where each choice contains a statement
regarding the exceptional status of idioms in Italian and Hebrew:
(335) Choice I: (331a) is the base structure, and (331b) is derived from it by a rule which
moves one of the constituents:
Choice Ia: the NP moves right over the PP, except in idioms.
Choice Ib: the PP moves left over the NP, obligatorily in idioms.
Choice II: (331b) is the base structure, and (331a) is derived by a rule which moves
one of the constituents:
Choice IIa: the PP moves right over the NP, obligatorily in idioms.
Choice IIb: the NP moves left over the PP, except in idioms.
255
Choice III: both (331a) and (331b) are base structures, and the two variants are related
elsewhere in the grammar, say, by some lexical or interpretive rule; idioms
are marked somehow as only occurring in (331a).
Choice IV: neither (331a) nor (331b) is a basic structure; rather both structures are
projected from the lexicon in the manner we have described before and will
illustrate further below; idioms are marked so that they are only projected
as (331a).
Belletti and Shlonsky do not consider Choice III or Choice IV, which is understandable. They are
working within the framework of a theory which generally accounts for variations like those in (331)
via a movement rule of some kind. It appears that Belletti and Schlonsky consider neither Choice
III nor Choice IV, because the spirit of their work is to assume that surface variations are related to
base configurations, a position which strongly predisposes consideration of only Choice I and Choice
II. Unfortunately, this predisposition, forces an analyst to select one of the two variants as base
generated and to find arguments to support that selection.
Naturally, the approach we have taken elsewhere in this book predisposes us to Choice IV, namely,
that neither (347a) nor (347b) is more basic than the other. Adopting the approach suggested above
for Latin and Japanese, we can dispense entirely with the question of whether movement occurs to
the right or to the left by postulating that both Italian and Hebrew contain the following V1 frame
for nonidioms, where “” indicates, as before, that the order of adjacent items is free.
(336) THE ITALIAN/HEBREW V1 FRAME FOR NONIDIOMS
[ V1 V0 – N3 C3 ]

256
This frame says that the order of N3 and C3 in V1 is free in Italian and Hebrew. Such an approach
has the immediate advantage of not having to prove that one structure is more basic than the other.
Idioms are an exception to this frame, requiring instead the following structural analysis, which can
be stated in lexical redundancy rules (see Section3.7 beginning on page 188):
(337) THE ITALIAN/HEBREW V1 FRAME FOR IDIOMS
[ V1 V0 – N3 – C3 ]
Opting for Choice IV and accepting a nonmovement account, we can still accommodate other data
that Belletti and Shlonsky discuss such as the cliticization of ne in Italian. As they note, “a crucial
property of ne cliticization is that the NP from which the clitic is extracted must occupy the
canonical direct object position” (Page 498). We can account for that property of ne cliticization as
follows: cliticization of ne is possible only if (331a) and not (331b) is the structure projected from
the lexicon.
On the other had, the problem s more ore general than just these constructions in Italian and Hebrew.
If we turn to the behavior of idioms in English, preferences for the four choices itemized above
change. Before we examine the choices, it is important to consider two facts about idioms. First,
it is well–known that they are severely restricted. Nonetheless, they must be provided with
appropriate internal structure, since that structure often determines possible variants. To see this,
consider the phrasal structure of the following English sentences:
(338) a. He put [on] [that hat].
b. He sat [on that hat].
The difference between the intransitive (338a) and transitive (338b) use of prepositions determines
the following contrasts:
(339) a. *It was on that hat that he put.
b. It was on that hat that he sat.
(340) a. *Where did he put?
b. Where did he sit?
(341) a. He put that hat on.
b. *He sat that hat on.
In English, only idioms with the structure (342a) allow the preposition to occur after the noun phrase,
as (343) and (344) show.
(342) a. V – [ PP P] – NP
b. V – [ P – NP]
PP

(343) a. off: blow off a lot of steam/blow a lot of steam off
b. down: bring down the house/bring the house down
c. up: make up one’s mind/make one’s mind up
(344) a. off: go off the deep end/*go the deep end off
b. down: jump down one’s throat/*jump one’s throat down
c. up: bark up the wrong tree/*bark the wrong tree up
Further, it is important to note that some English idioms can freely occur in a variety of
configurations such as the passive even when they are composed of more than one complement and
when both complements contribute to the idiom’s meaning:
(345) a. John’s problem is that he puts all his eggs into one basket.
b. Unfortunately, all of John’s eggs have been put into one basket.
(346) a. John’s lack of enthusiasm really took the wind out of my sails.
b. The wind was taken right out of my sails by John’s lack of enthusiasm.
Second, although some idioms exhibit the same structural variations as their nonidiomatic
counterparts, others have a fixed internal structure which will present problems for theories that
attempt to relate variations by some movement rule. For example, early in the history of
Transformational grammar, textbooks routinely related the variants in (347) with a movement rule.
(347) a. He gave [ NP the book] [ PP to the boy].
b. He gave [ the boy] [ the book].
NP NP
Whether one assumes (347a) or (347b) to contain the base or initial structure, one runs into trouble
with English idioms since some require one, some require the other, and some allow both — exactly
the same variations that one sees in nonidiomatic indirect object constructions:
(348) a. *He gave the creeps to me./He gave me the creeps.
b. *He spared the embarrassment to me/.He spared me the embarrassment.
(349) a. *He gave a run for my money to me./He gave me a run for my money.
b. *He guaranteed a promotion to me./He guaranteed me a promotion.
(350) a. *He gave hell to me./He gave me hell.
b. *He wished success to me./He wished me success
(351) a. The approach gave rise to many problems./*The approach gave many problems rise.
b. He explained linguistics to her./*He explained her linguistics.
257

258
(352) a. He needs to give vent to his rage./*He needs to give his rage vent.
b. He conveys assurance to all he meets./*He conveys all he meets assurance.
(353) a. We should give thanks to our parents./We should give our parents thanks.
b. We should bring gifts to our parents./We should bring our parents gifts.
(354) a. He gives free rein to all his employees./He gives all his employees free rein.
b. He lends money to his employees./He lends his employees money.
Paralleling (335), there are again at least six choices for the variants in (347) as follows:
(355) Choice I: (347a) is the base structure, and (347b) is derived from it by a rule which
moves one of the constituents; the PP converts to NP:
Choice Ia: the NP moves right over the PP.
Choice Ib: the PP moves left over the NP.
Choice II: (347b) is the base structure, and (347a) is derived by a rule which moves
one of the constituents; the first NP converts to a PP:
Choice IIa: the first NP moves right over the second NP.
Choice IIb: the second NP moves left over the first NP.
Choice III: both (347a) and (347b) are base structures, and the two variants are related
elsewhere in the grammar, say, by some lexical or interpretive rule; both
idiomatic and nonautomatic expressions are marked somehow as occurring
in either one or the other or both of the structures.
Choice IV: neither (347a) nor (347b) is a basic structure; rather both structures are
projected from lexical specifications in both idiomatic and nonidiomatic
expressions.
In view of the above data and the multitude of other relevant examples in English, the fixed structure
of different idioms based on the same verb, e.g., give, strongly supports Choice IV, that is, the option
in which neither of the two structures in (347) are viewed as basic. In the other three choices,
whether one variant is viewed as basic or both are, certain English idioms and nonidioms must be
treated exceptionally somewhere in the grammar, since some only occur in one of the two variants
while others occur in both. Using idioms to support a particular construction as basic is misguided.
In the next chapter, we will discuss in more detail how the above facts can be included in lexical
entries in a grammar that has no movement rules. For an in-depth investigation of the lexicon and
a typology of predicate types, see Binkert 1996.

3.18 THE ENGLISH AUXILIARY
3.18.1 THE FACTS AND GENERALIZATIONS
Noam Chomsky’s original analysis of the English auxiliary, along with subsequent work in the
framework of transformational grammar (TG), revealed a number of important facts about the
English auxiliary, which are summarized in (356).
(356) a. Every English independent clause (S/V3) must contain a tense (TNS) marker:
1. PRS (PRESENT): hunts, has hunted, is hunting
2. PST (PAST): hunted, had hunted, was hunting
3. FUT (FUTURE): will hunt, will have hunted, will be hunting
4. CND (CONDITIONAL): could/would/should hunt
5. IMP (IMPERATIVE): hunt/do hunt
259
b. The verb that carries the TNS marker is always the first verb in the verbal sequence.
This verb also carries agreement: She is hunting/*They is being hunted/*She have
been hunting.
c. Every English clause must also contain a main verb, which is the last verb in the
verbal sequence. This main verb may be empty: The women have hunted, and the
men have [e] too.
d. Every English clause may also optionally contain a MODAL, PERFECTIVE,
PROGRESSIVE, and/or PASSIVE element in that order. Thus:
TNS (MOD) (PERF) (PROG) (PASS)
e. The PERF, PROG, and PASS each consist of an auxiliary verb and a participial affix
which always appears on the next verb in the sequence, whatever that verb is (the
PROG affix is always –ing; the PERF and PASS affixes are –ed in regular verbs):
PERF = have + PFP (PERF PARTICIPLE AFFIX): has hunt–ed
PROG = be + PGP (PROGRESSIVE PARTICIPLE AFFIX): is hunt–ing
PASS = be + PSP (PASSIVE PARTICIPLE AFFIX): is hunt–ed

260
f. Further examples of participles are:
PROGRESSIVE PARTICIPLES PASSIVE/PERFECTIVE PARTICIPLES
hunting hunted
going gone
weeping wept
being been
having had
doing done
These facts are realized in English in sixteen possible auxiliary sequences:
(357) a. Women hunted. TNS
b. Women could hunt. TNS + MOD
c. Women had hunted. TNS + PERF
d. Women were hunting. TNS + PROG
e. Men were hunted. TNS + PASS
f. Women could have hunted. TNS + MOD + PERF
g. Women could be hunting. TNS + MOD + PROG
h. Men could be hunted. TNS + MOD + PASS
i. Women have been hunting. TNS + PERF + PROG
j. Men had been hunted. TNS + PERF + PASS
k. Men were being hunted. TNS + PROG + PASS
l. Women could have been hunting. TNS + MOD + PERF + PROG
m. Men could have been hunted. TNS + MOD + PERF + PASS
n. Men could be being hunted. TNS + MOD + PROG + PASS
o. Men have been being hunted. TNS + PERF + PROG + PASS
p. Men could have been being hunted. TNS + MOD + PERF + PROG + PASS
Listing the possible combinations is hardly an adequate description of English: it attains only
observational adequacy. The TG analysis has gone much further. It notes that all the above
combinations are variations of the sequence in (3).
(358) TNS (MOD) (have + PFP) (be + PGP) (be + PSP) V
The major coup was realizing that the affixes (TNS, PFP, PGP, and PSP) are always associated with
the following verb, whichever happen to come next in the sequence. Given this, a transformational
analysis is necessary; it must take a phrase structure sequence like (3) and distribute the affixes to
the following verb, whatever that might be. Consider the following examples:

(359) TNS + V
PST + hunt
.)))0)))hunted
(360) TNS + MODAL + V
CND + can + hunt
.)))0))))could
hunt
(361) TNS + MODAL + PERF + PROG + V
CND + can + have + PFP + be + PGP + hunt
.)))0)))- .))0))- .)))0)))-
could have been hunting
(362) TNS + PERF + PROG + PASS V
PRS + have + PFP + be + PGP + be + PSP + hunt
.)))0))))- .))0))- .))0)))- .))0))has
been being hunted
261

262
3.18.2 PROBLEMS WITH THE CLASSIC TG ANALYSIS
Despite the elegance of this treatment of the English auxiliary, it was soon realized that it could not
be maintained in its original form as described above. Three major problems were encountered.
First, it was realized that the auxiliary elements (TNS, MOD, PERF, PROG, PASS) do not have the
same range of occurrence possibilities. In particular, the following split exists:
(363) TNS and MOD cannot occur in infinitives (to go) and participles (going):
a. I wanted to be seen.
*I wanted to was seen.
I wanted to be able to drive.
*I wanted to can drive.
b. Being seen with her is a sign of success.
*Wasing seen with her is a sign of success.
Being able to drive is a sign of success.
*Canning (to) drive is a sign of success.
(364) PERF, PROG, and PASS can occur in infinitives and participles:
a. PERF:
It was fun [to have done that].
[Having done that] he left.
b. PROG:
I want [to be working] when he arrives.
?[Being hunting] when you should be working is bad.
c. PASS:
(365) BASIC SPLIT:
He tried [to be seen].
[Being invited to such parties] is not important.
(MOD) TNS – (PERF) (PROG) (PASS) V
This split is confirmed in two ways:

(366) An negative (not) can modify either constituent:
a. He can not attend. (ambiguous)
b. He can’t just not attend.
(367) When the modal occurs, it inverts in questions:
a. Can he go? (cf. He can go.)
b. Won’t they go? (cf. They won’t go.)
Adopting either a TG or RG framework, we can account for the above split with the following
diagrams:
(368) You can not attend.
[ S [ NP you] [ AUX [ TNS [ MOD can]] PRS [ NEG not] ] [ VP [ V attend] ]]
[ [ you] [ [ [ can]] PRS [ not] ] [ attend] ]
V3 N3 C3 C0 V3 C3 V0
[ S [ NP you] [ AUX [ TNS [ MOD can]] PRS ] [ VP [ NEG not] [ V attend] ]]
[ [ you] [ [ [ can]] PRS ] [ [ not] [ attend] ]]
V3 N3 C3 C0 V3 V2 C3 V0
The second problem with the classic TG analysis of the auxiliary concerns the use of do. If a clause
contains only TNS, then the verb do must occur in questions:
(369) a. He will go. Will he go?
b. He has gone. Has he gone?
c. He is going. Is he going?
(370) a. He went. *Went he?
b. He went. Did he go?
This do acts like a modal for two reasons:
(371) a. It carries TNS, agreement, and NEG:
Doesn’t – he – want that?
Don’t – they – want that?
Did – they – want that?
b. It does not occur in infinitives and participles:
*I want to do go there.
*Doing go there is fun.
263

264
This do must be distinguished from the main verb do as follows:
(372) MODAL MAIN VERB
a. He does that.
b. He does see that.
c. He does do that.
d. He doesn’t do that.
e. He does so do that.
But, do (unlike the other modals, e.g., can) cannot occur with PERF, PROG, and/or PASS:
(373) a. He can be working.
b. *He does be working.
c. He could have gone.
d. *He did have gone.
If we treat do as a modal, we can add a constraint to rule out occurrences of do with PERF, PROG,
and/or PASS. Recalling that PERF, PROG, and PASS all consist of a verb, we can add filter (374)
to our grammar (where W, X, Y and Z are variables):
(374) a. TG: *[ S W do X TNS Y V Z]
b. RG: *[ W do X TNS Y V Z]
V3
This filter, like most syntactic filters in English, is under a clause constraint, meaning that the listed
morphological elements do and TNS must be in the same clause (S or V3). This proviso is
necessary to disallow *he does be going while allowing does he think he is going.
Given this analysis, do occupies the same residence as the modals generally do in addition to main
verb position:
(375) He does not do that.
[ S [ NP he] [ AUX [ TNS [ MOD does] PRS] [ NEG not] ] [ VP do that] ]
[ V3 [ N3 he] [ C3 [ C0 [ V3 does] PRS] [ C3 not] ] [ V0 do that] ]
The third problem with the classic TG analysis of the auxiliary is the most serious. It concerns the
constituency of the various elements. It is here that RG provides an analysis that is superior to the
classic TG analysis. In particular, the classic TG analysis leaves the following facts unexpressed:

(376) The PROG and PASS can occur, as participles, without their associated verbs:
a. The man lurking in the shadows is the culprit.
The man who is lurking in the shadows is the culprit.
b. The man discovered in the bushes is a flasher.
The man who was discovered in the bushes is a flasher.
c. We saw the man hunting.
d. We saw the man hunted.
(377) PGP and PSP forms are ambiguous:
a. an imposing Avon lady
an Avon lady imposing on her neighbors (PARTICIPLE)
an Avon lady imposing in size (ADJECTIVE)
b. a broken vase
a vase accidentally broken by a child (PARTICIPLE)
a vase so broken that it can’t be fixed (ADJECTIVE)
These facts can be accommodated within the RG framework by observing that participial markers
behave like characterizers, in particular, like prepositions. Thus, (378a), (378b), (378c), and (378d)
are analogous structures; significantly, the postverbal characterizers (the C3 off V2 to the right of
V0) are independent structures, thus allowing for participial examples like those in (376):
(378) a. The jars started to break.
[ V3 [ N3 the jars] [ C3 PST] [ V2 [ V0 started] [ C3 [ C1 [ C0 to ] [ V3 [ V0 break] ]]]]
b. The jars started breaking.
[ V3 [ N3 the jars] [ C3 PST] [ V2 [ V0 started] [ C3 [ C0 [ C0 PGP ] [ V3 [ V0 breaking] ]]]]
c. The jars were breaking.
[ V3 [ N3 the jars] [ C3 PST] [ V2 [ V0 were] [ C3 [ C0 [ C0 PGP ] [ V3 [ V0 breaking] ]]]]
d. The jars were broken.
[ V3 [ N3 the jars] [ C3 PST] [ V2 [ V0 were] [ C3 [ C0 [ C0 PSP ] [ V3 [ V0 broken] ]]]]
265

266
If we treat be as a main verb and treat PGP and PSP as characterizers, we solve the above problems
in (376) and (377). Specifically, we take advantage of the fact that RG makes transparent the many
left-of head/right-of-head alternations that occur in English.
English is a head–center language, that is, one that tolerates elements both before and after the head;
many phrases can occur alternatively in prehead or posthead position. For example, we have noun
phrase pairs like the following:
(379) a. [ N3 the [ N2 Spanish [ N1 math [ N0 teacher]]]]
b. [ N3 the [ N2 [ N1 [ N0 teacher] of math] from Spain]
We have other examples of prehead/posthead alternations like the following:
(380) a. V3: Ultimately, peace will prevail/Peace will prevail, ultimately.
That guy there, he’s my friend./He’s my friend, that guy there.
b. N3: enough money/money enough
as many too many marbles/as many marbles too many
c. C3: quite far down the road/down the road quite far
two miles down the road/down the road two miles
(381) a. V2: enthusiastically sang the aria/sang the aria enthusiastically
often sang the aria/sang the aria often
b. N2: a sleeping baby/a baby sleeping
a 5000 foot high mountain/a mountain 5000 feet high
a courageous man/a man of courage
c. C2: directly down the road for a mile/down the road directly for a mile
especially for Sue/for Sue especially
In English, complements are generally posthead, although there are productive alternatives in noun
phrases (history student versus student of history; *history studied versus studied history) and
participial phrases (French speaking students versus students speaking French). Depending on the
language (head center, head initial, head final), complements generally either precede or follow the
head.
When a verb is embedded in prehead position of PGP or PSP it forms a kind of compound adjective
with the affix; in posthead position, the embedded verb is the complement of the affix:

(382) a. The lady was imposing (PROGRESSIVE).
[ V3 [ N3 the lady] [ C3 PST] [ V2 [ V0 was] [ C3 [ C0 [ C0 PGP ] [ V3 [ V0 imposing] ]]]]
b. The lady was imposing ( DERIVED ADJECTIVE 1).
[ V3 [ N3 the lady] [ C3 PST] [ V2 [ V0 was] [ C3 [ C0 [ V3 [ V0 imposing] ] [ C0 DA1 ] ]]]
(383) a. The jars were broken (PASSIVE).
[ V3 [ N3 the jars] [ C3 PST] [ V2 [ V0 were] [ C3 [ C0 [ C0 PSP ] [ V3 [ V0 broken] ]]]]
b. The jars were broken (DERIVED ADJECTIVE 2).
[ V3 [ N3 the jars] [ C3 PST] [ V2 [ V0 were] [ C3 [ C0 [ V3 [ V0 broken] ] [ C0 DA2 ] ]]]
If we allow that a main verb can be empty [e] and bound to be in the TNS characterizer, we account
for the occurrence of be in questions:
(384) The jars all were broken.
[ V3 [ N3 the jars] [ C3 [ C0 PST] ] [ V2 [ N3 all] [ V0 were] [ C3 broken]] ]
(385) The jars were all broken (PASSIVE). [= All the jars were broken (PASSIVE).]
[ V3 [ N3 the jars] [ C3 [ C0 were i [ C0 PST]] ] [ V2 [ N3 all] [ V0 [e] i] [ C3 broken]] ]
In the process, we also account for the ambiguity of all.
(386) The jars were all broken (ADJ). [= All the jars were broken (ADJ).]
[ V3 [ N3 the jars] [ C3 [ C0 were i [ C0 PST]]] [ V2 [ N3 all] [ V0 [e] i] [ C3 broken]] ]
(387) The jars were all broken (ADJ). [= The jars were totally broken.]
[ V3 [ N3 the jars] [ C3 [ C0 were i [ C0 PST]]] [ V2 [ V0 [e] i] [ C3 [ N3 all] broken]] ]
(388) The jars were all all broken (ADJ). [= All of the jars were totally broken (ADJ).]
[ V3 [ N3 the jars] [ C3 [ C0 were i [ C0 PST]]] [ V2 [ N3 all] [ V0 [e] i] [ C3 [ N3 all] broken]] ]
267

268
The above analysis solves a multitude of problems for English syntax. Notice that the participial
characterizer phrases are all suspended from the V2 level. This is the residence in RG of predicate
nominatives: items suspended from this position refer back to the subject. Consider the following
sentences and their associated RG diagrams:
(389) The boys made good cooks. (make is copulative)
[ V3 [ N3 the boys] [ C3 [ C0 PST]] [ V2 [ V0 made] [ N3 good cooks] ] ]
(390) The boys made good cakes. (make is transitive)
[ V3 [ N3 the boys] [ C3 [ C0 PST]] [ V1 [ V0 made] [ N3 good cakes] ] ]
Observe that the only structural difference between (389) and (390, 391) is that the postverbal N3
in (389) comes off V2, while the one in (390, 391) comes off the V1. This is in accordance with the
RG residential requirements: predicate nominatives are V2 level elements; direct objects are V1
level elements. Since predicate nominatives refer back to the subject, they must agree with the
subject. In short, the syntactic hierarchy of RG accounts automatically for all of the following:
(391) Faulty agreement between subject and V2 level predicate nominative phrase:
a. *The boys became a good cook.
b. *The boy made good cooks.
c. *The boy is pregnant.
d. *The boys made (became) coffee.
e. *The coffee is broken.
(392) No need for agreement between subject and V1 level direct object phrase:
a. The boys made a good cake.
b. The boy made good cakes.
c. The boy made the girl pregnant.
d. The boys made coffee.
(393) There is often ambiguity between V1 and V2 level elements.
a. John left the house messy. (= John was messy when he left the house.)
[ V3 John PST [ V2 [ V1 [ V0 left] [ N3 the house]] [ C3 messy]]]
b. John left the house messy. (= The house was left messy by John.)
[ V3 John PST [ V1 [ V0 left] [ N3 the house] [ C3 messy]]

(394) The order must be V1 level elements before V2 level elements in post verbal position.
a. John [ [
V2 V1 left the house unlocked]
]
b. John [ V2 [ V1 left the house] dazed ]
c. John [ V2 [ V1 left the house unlocked] dazed ]
d. *John [ [ left the house dazed] unlocked ]
V2 V1
Since the structure of (394) parallels (389) as regards predicate nominative position, agreement
between participles and subjects is automatically accounted for:
(395) a. The terrorist was undaunted by threats.
[1] *Threats undaunted the terrorist.
[2] *The lamp was undaunted by threats.
b. The men were undivided.
[1] *That man was undivided.
[2] *We undivided those men.
c. The men were divided (ambiguously adjective or participle).
[1] I am divided. (adjective only)
[2] *These conflicts are dividing me.
Of course, an added plus for this analysis is that case/number/gender agreement between subjects
and predicate nominatives (including participles) in languages like Latin, French, etc., is directly
accounted for.
Given the above, it follows that we can analyze PERFECTIVE have in either of two ways (recall that
the emphatic so is part of the TNS characterizer):
(396) a. They [could have so] [gone].
b. They [could so] [have gone].
There is, in fact, support for assuming that have can occur in each one of the two above positions.
It must be in V3 when it occurs in infinitives and participles:
(397) a. For him [to have done that] is awful.
b. [Having done that] he left.
In cases like (397), we know that have cannot be in the tense characterizer, because TNS cannot
occur in infinitives and participles. However, have must be in the TNS characterizer in questions
and in certain elliptical clauses:
269

270
(398) a. Have they done that?
b. He has done that, and so has she _______.
The two positions for have in (396) are, in fact, fairly fixed positions in English syntax: the tense
characterizer and the main verb. Perhaps, for this reason, have seems to float back and forth between
the two possible structures, required in the one to account for (397), and in the other to account for
(398). We document this fluctuation with the following two representations for PERFECTIVE have:
(399) Perfective have is part of the TNS characterizer:
[ V3 [ N3 he] [ C3 [ C0 [ V3 [ V0 has]] PRS ]] [ V0 gone] ]
(400) Perfective have is part of the verb phrase:
[ V3 [ N3 he] [ C3 PRS ] [ V2 [ V3 [ V0 has]] [ V0 gone]] ]
3.18.3 RESOLUTION
The complete English auxiliary is, therefore, the following:
(401) The jars could so have been being broken.
[ V3 [ N3 the jars] [ C3 [ C0 could [ C0 CND]] so ]
[ V3 have [ V2 [ V0 been]
[ C3 [ C0 [ C0 PGP] [ V3 [ V2 [ V0 being]
[ C3 [ C0 [ C0 PSP] [ V3 [ V0 broken]]]] ]]]]]]]
Additional support for the above analysis comes from the behavior of so-called "floating" quantifiers
like all:
(402) a. All of the men could have gone fishing.
b. The men all could have gone fishing.
c. The men could all have gone fishing.
d. The men could have all gone fishing.
e. *The men could have gone all fishing.
Observe that floating quantifiers can "float" up to the main verb (gone in (402)).We see the same
paradigm in (402) in all of the following:

(403) a. All of the men could have been six-feet tall.
b. The men all could have been six-feet tall.
c. The men could all have been six-feet tall.
d. The men could have all been six-feet tall.
e. *The men could have been all six-feet tall.
(404) a. All of the men could have been fishing.
b. The men all could have been fishing.
c. The men could all have been fishing.
d. The men could have all been fishing.
e. *The men could have been all fishing.
(405) a. All of the men could have been seen.
b. The men all could have been seen.
c. The men could all have been seen.
d. The men could have all been seen.
e. *The men could have been all seen.
(406) a. All of a men could have been aware of a problem.
b. The men all could have been aware of a problem.
c. The men could all have been aware of a problem.
d. The men could have all been aware of a problem.
e. *The men could have been all aware of a problem.
(407) a. All of the jars could have been all broken.
b. The jars all could have been all broken.
c. The jars could all have been all broken.
d. The jars could have all been all broken.
e. *The jars could have been all all broken.
Of course, be is the main verb of all of these sentences in the RG analysis. Accordingly, all of the
data are accounted for. There is no version of TG that can make the same claim.
Notice, in addition, that the above analysis treats all uses of be (copular, progressive, and passive)
in the same way, which is necessary given the form of tag questions like the following:
(408) a. They were aware of the problem, weren’t they?
They weren’t aware of the problem, were they?
b. They were fishing, weren’t they?
They weren’t fishing, were they?
c. They were seen, weren’t they?
They weren’t seen, were they?
271

272
Given (408), we see that all uses of be repeat be in the tag with opposite polarity. In this regard, all
uses of be work the same way as modals and perfective have:
(409) a. They could have been aware of the problem, couldn’t they?
They couldn’t have been aware of the problem, could they?
b. They have been aware of the problem, haven’t they?
They haven’t been aware of the problem, have they?
Thus, all auxiliary verbs and all uses of be repeat those verbs in the tag, as opposed to nonauxiliary
verbs, which require a form of do:
(410) a. They went fishing, didn’t they?
They went fishing, did they not?
*They went fishing, weren’t they?
*They went fishing, went they not?
b. They have new cars, don’t they?
They have new cars, do they not?
*They have new cars, haven’t they?
In short, verbs that do not require do-support form their tags without do:
(411) a. *They do be aware of the problem.
b. *They do be fishing.
c. *They do be seen.
d. *They do have gone.
e. *They do can go.
(412) a. They do go fishing.
b. They do have new cars.
c. They do do a good job.
d. They do can tomatoes.
The data above suggest that modals, perfective have, and all uses of be form one class of verbs, say
[+AUX], while all other verbs form another class, say [–AUX].

3.18.4 SUMMARY OF NOMINALS AND VERBALS IN RG
(413) Progressive Participles (PGP) and Derived Adjectives (DA1).
a. Participle (the lady charming the man): [ C3 [ C0 [ C0 PGP] [ V3 [ V0 charming]]]]
b. Adjective (a very charming lady): [ C3 [ C0 [ V3 [ V0 charming]] [ C0 DA1]]]
(414) Passive Participles (PSP) and Derived Adjectives (DA2).
a. Participle (charmed by her flattery): [ C3 [ C0 [ C0 PSP] [ V3 [ V0 charmed]]]]
b. Derived Adjective (a very charmed life): [ C3 [ C0 [ V3 [ V0 charmed]] [ C0 DA2]]]
(415) Gerundial Nominals (GN) and Derived Nominals (DN).
a. GN (charming candy from a baby): [ N3 [ N0 [ N0 GN] [ V3 [ V0 charming]]]]
b. DN: (such charming of snakes): [ N3 [ N0 [ V3 [ V0 charming]] [ N0 DN]]]
(416) Summary of Verbal Structures.
a. Verb resides in posthead position.
[ C3 [ C0 [ C0 PGP] [ V3 [ V0 charming]] ]]
[ C3 [ C0 [ C0 PSP] [ V3 [ V0 charmed]] ]]
[ N3 [ N0 [ N0 GN] [ V3 [ V0 charming]] ]]
b. Verb resides in prehead position.
[ C3 [ C0 [ V3 [ V0 charming]] [ C0 DA1] ]]
[ C3 [ C0 [ V3 [ V0 charmed]] [ C0 DA2] ]]
[ N3 [ N0 [ V3 [ V0 charming]] [ N0 DN] ]]
273

274
3.18.5 PARSES ILLUSTRATING THE ENGLISH AUXILIARY
(417) Imperative Tense: interview them
(418) Present Tense: he interviews them
(419) Past Tense: he interviewed them

(420) Future Tense: he will interview them
(421) Conditional Tense: he would interview them
275

276
(422) Present Tense + Perfective Aspect: he has interviewed them
(423) Present Tense + Progressive Aspect: he is interviewing them

(424) Past Tense + Passive Voice: they were interviewed by him
277

278
(425) Conditional Tense + Perfective Progressive Aspect:
he should have been interviewing them
There are two possible positions in which perfective have occurs, either in the tense phrase or in the
verb phrase. It must be in the tense phrase in sentences like He should’ve been interviewing them;
it must be in the verb phrase in sentences like It was important for him to have been interviewing
them (see page 269 ff.). In the example here, perfective have can be in either position.

(426) Conditional Tense + Perfective Aspect + Passive Voice:
they might have been interviewed by him
279

280
(427) Conditional Tense + Perfective Progressive Aspect + Passive Voice:
they really couldn’t all have been being interviewed by him

(428) Conditional Tense + Perfective Progressive Aspect + Passive Voice:
they each could not actually have been being interviewed by him
Another parse has the word not specifying (and, therefore, negating) the modal, which would
underlie the sentence They each couldn’t actually have been being interviewed by him.
281

282
EXERCISES FOR CHAPTER THREE
1. Draw TG diagrams for the DEEP STRUCTURE of the following sentences:
a. The book may be in the library.
b. John will not read those magazines.
c. Mary will send a letter to the company.
d. John reads those magazines.
e. The roses border the fence.
f. The accident occurred at the corner.
g. The plant fell out the window.
h. The boy could take the garbage out.
I. The boy could take out the garbage.
j. The book was not in the library.
2. Collapse each of the following sets of rules into one rule using formal abbreviatory devices
(parentheses and curly brackets):
a. The symbol QNT = QUANTIFIER (more, all, half, etc.)
NP DET + N
NP QNT + DET + N
NP QNT + N
NP N
b. The symbol AP = ADJECTIVE PHRASE
VP V
VP V + PP
VP V + AP
VP V + NP
3. As we have noted in the text, the PS–Rules given in (35) are incomplete. They will have to
be revised to accommodate each new structure we consider. Such revisions must be justified
by argument, and the details of proposals must be carefully worked out. Which of the
following sentences can be generated by (107)? Which cannot and, therefore, will require that
(107) be revised? What sort of revisions are necessary?
a. The gardener should water those shrubs with a fine spray.
b. The nearsighted soprano fell into the orchestra pit.
c. They will probably read the assignment reluctantly.
d. A new cook from India prepared all the meals.
e. John thinks Bill will come.

4. All of the following sentences have the same basic phrase structure, that is, they contain the
same basic units. Keep this in mind as you provide TG diagrams for them.
a. The soprano sang the aria with skill.
b. The young soprano from a small town in the south of France sang the aria with skill.
c. The soprano sang the aria from the third act of an opera by Verdi with skill.
d. The soprano sang the aria with the skill of a seasoned veteran.
e. The young soprano from a small town in the south of France sang the aria from the third
act of an opera by Verdi with the skill of a seasoned veteran.
5. Structurally ambiguous sentences must be given a different diagram for each structural
ambiguity. For example, the sentence Bill may slip on the boots has the following two
diagrams in TG format:
a. For the meaning on the boots, Bill may slip:
b. For the meaning Bill may slip the boots on:
283

284
Provide disambiguating diagrams for each of the following ambiguous sentences:
a. The thief hit the lady with the hat.
b. The repair man will look up the street.
c. John left the house messy.
d. The prisoner of war spoke foolishly.
6. What problems, if any, do the following ambiguous sentences pose for the TG model of
grammar?
a. They hired Spanish teachers.
b. They are all finished.
c. They do so love a good meal.
d. We made them idols.
e. They made idols.
7. Draw RG diagrams for the following sentences:
a. The book may be in the library.
b. John will not read those magazines.
c. Mary will send a letter to the company.
d. John reads those magazines.
e. The roses border the fence.
f. The accident occurred at the corner.
g. The plant fell out the window.
h. The boy could take the garbage out.
I. The boy could take out the garbage.
j. The book was not in the library.
8. Draw diagrams for the following sentences in RG framework. What are the advantages of the
X–Bar notation?
a. The gardener should water those shrubs with a fine spray.
b. The nearsighted soprano fell into the orchestra pit.
c. They will probably read the assignment reluctantly.
d. A new cook from India prepared all the meals.
e. John thinks Bill will come.

9. Draw diagrams for each of the meanings of the following sentences in the RG framework.
Does the X–Bar notation allow you to reach a higher level of descriptive adequacy?
a. The thief hit the lady with the hat.
b. John left the house messy.
c. The repair man will look up the street.
d. The prisoner of war spoke foolishly.
e. They hired Spanish teachers.
f. They made idols.
10. Show that the following sentences must be assigned different structural descriptions no matter
which theory of grammar is adopted.
a. (1) The boys made good cooks.
(2) The boys made good cakes.
Do the same with the following pair:
b. (1) The cars passed in the tunnel.
(2) The boys passed in the tests.
285

286
APPENDIX A: ANSWERS TO EXERCISES
1. Draw TG diagrams for the following sentences:
a. The book may be in the library.
b. John will not read those magazines.
c. Mary will send a letter to the company.

d. John reads those magazines.
e. The roses border the fence.
f. The accident occurred at the corner.
g. The plant fell out the window.
287

288
h. The boy could take the garbage out.
I. The boy could take out the garbage.
j. The book was not in the library.
2. Collapse each of the following sets of rules into one rule using formal abbreviatory devices
(parentheses and curly brackets):
a. NP (QNT) + (DET) + N
b. VP V + ({ PP, AP, NP})

3. a. (107) can generate the structure for The gardener should water those shrubs with a fine
spray:
b. (107) cannot generate the structure for The nearsighted soprano fell into the orchestra
pit. The phrase orchestra pit is a compound noun made up of two nouns (orchestra and
pit). The rule (107d) for rewriting NP does not allow an NP to contain two nouns. Note
that orchestra is a noun, not an adjective: it can be made plural (orchestras) and it
cannot be made comparative (*more orchestra).
c. (107) cannot generate the structure for They will probably read the assignment
reluctantly. (107) contains no rule for adverbs. Probably is a sentence adverb (a
constituent of S), so (107a) must be revised to allow an adverb. Reluctantly is a manner
adverb (a constituent of VP), so (107b) must also be revised to allow an adverb.
d. (107) cannot generate the structure for A new cook from India prepared all the meals.
All is a quantifier, and (107d) does not allow for the generation of a quantifier in an NP.
The top line of (107d) would have to be revised as follows:
(QNT) + (DET) + (ADJ) + N + (PP)
e. (107) can generate the structure for John thinks Bill will come.
289

290
4. All of the following sentences have the same basic phrase structure, that is, they contain the
same basic units. Keep this in mind as you provide TG diagrams for them.
a. The soprano sang the aria with skill.
b. The young soprano from a small town in the south of France sang the aria with skill.
c. The soprano sang the aria from the third act of an opera by Verdi with skill.
d. The soprano sang the aria with the skill of a seasoned veteran.
e. The young soprano from a small town in the south of France sang the aria from the third
act of an opera by Verdi with the skill of a seasoned veteran.

5. Provide disambiguating diagrams for each of the following ambiguous sentences:
a. The thief hit the lady with the hat.
(1) =It was with the hat that the thief hit the lady.
(2) =It was the lady with the hat that the thief hit.
b. The (repair) man will look up the street.
291
Note: Repair man cannot be generated by the rules in (107); it is a compound noun like
orchestra pit in exercise (3b) above. Still, the sentence remains ambiguous without the
word repair in it.
(1) = It is up the street that the man will look.

292
(2) = The man will look the street up.
c. Ron left the house messy.
(1) = The house was messy when Ron left it.
(2) = Ron was messy when he left the house.
Note: While this diagram would appear to be satisfactory to account for the meaning
in which messy modifies Ron, it will prove to be descriptively inadequate when other
sentences are considered. For example, we know that a manner adverb like reluctantly
must be part of the VP (see exercise (5d) below and the explanation for (3c) above).
There are sentences like Ron left the house messy reluctantly, which are still ambiguous
as to whether it is Ron or the house that was left messy. Thus, the above diagram is
descriptively inadequate: since reluctantly must part of the VP in both readings of the
sentence Ron left the house messy reluctantly (it is a manner adverb in both readings),
then messy cannot be outside of the VP as in the above diagram. The standard theory
fails in this case.

d. The prisoner of war spoke foolishly.
(1) = The manner in which the prisoner of war spoke was foolish.
(2) = It was foolish of the prisoner of war to speak. (foolishly is a sentence adverb)
6. What problems, if any, do the following ambiguous sentences pose for TG?
Within the TG framework, there is no way to disambiguate any of the following sentences, that
is, PS-Rules, like those in (107), cannot handle the subtle distinctions made.
a. They hired Spanish teachers means either They hired teachers who are Spanish or They
hired teachers of Spanish. Notice that we have They hired Spanish Spanish teachers.
b. They are all finished means either All of them are finished or They are completely
finished. Notice that we have They are all all finished.
c. In They do so love a good meal, so emphasizes do as in They do too love a good meal
or it emphasizes love as in He really loves a good meal.
d. We made them idols means We made them into idols (like We elected them officers) or
We made idols for them (like We bought them presents).
e. They made idols means either They became idols or They created idols. The verb make,
therefore, can be used either as a transitive verb (one that takes a direct object) or as a
copulative verb (one that has a complement which refers back to the subject.
293

294
7. Draw RG diagrams for the following sentences:
a. The book may be in the library. (be is a copulative verb; note the V2)
b. John will not read those magazines. (read is a transitive verb; note the V1)
c. Mary will send a letter to the company. (send is a transitive verb; note the V1)

d. John reads those magazines. (read is a transitive verb; note the V1)
e. The roses border the fence. (border is a transitive verb; note the V1)
f. The accident occurred at the corner. (occur is an intransitive verb; note the V2)
g. The plant fell out the window. (fall is an intransitive verb; note the V2)
295

296
h. The boy could take the garbage out. (take is a transitive verb; note the V1)
i. The boy could take out the garbage. (take is a transitive verb; note the V1)
j. The book was not in the library. (be is a copulative verb; note the V2)

8. Draw diagrams for the following sentences in RG framework. What are the advantages of the
X–Bar notation?
a. The gardener should water those shrubs with a fine spray.
b. The nearsighted soprano fell into the orchestra pit.
297

298
c. They will probably read the assignment reluctantly.
d. A new cook from India prepared all the meals.
e. John thinks Bill will come.

9. a. The thief hit the lady with the hat.
(1) = It was with the hat that the thief hit the lady.
(2) = It was the lady with the hat that the thief hit.
299

300
b. Ron left the house messy.
(1) = The house was messy when he left it.
(2) = Ron was messy when he left the house.

c. The repair man will look up the street.
(1) = The repair man will look the street up.
(2) = It is up the street that the repair man will look.
301

302
d. The prisoner of war spoke foolishly.
(1) = It was foolish of the prisoner of war to speak.
(2) = The way the prisoner of war spoke was foolish.

e. They hired Spanish teachers.
(1) = teachers who were Spanish (2) = teachers of Spanish
f. They made idols
(1) = They became idols. (2) = They built idols.
10. a. The following data indicate that The boys made good cooks and The boys made good
cakes must have different structural descriptions:
(1) a. *The boy made good cooks.
b. The boy made good cakes.
(2) a. *Good cooks were made by the boy.
b. Good cakes were made by the boy.
b. The following data indicate that The cars passed in the tunnel and The boys passed in
the tests must have different structural descriptions:
(1) a. *The cars passed the tunnel in.
b. The boys passed the tests in.
(2) a. It was in the tunnel that the cars passed.
b. *It was in the tests that the boys passed.
303

304
APPENDIX B: LATIN SYNTAX
Latin verbs must agree with their subjects in number:
1. a. Agricola lupam necat. The farmer kills the wolf.
b. Agricola lupas necat. The farmer kills the wolves.
2. a. Agricolae lupam necant. The farmers kill the wolf.
b. Agricolae lupas necant. The farmers kill the wolves.
Agreement patterns between subjects and verbs:
3. a. Singular: Subject–a–Ø verb–a–t
regin–a–Ø voc–a–t
b. Plural: Subject–a–e verb–a–nt
regin–a–e voc–a–nt
4. a. *Agricola lupas necant.
b. *Agrocilae lupam necat.
Latin adjectives and numerals must agree with the nouns they modify in both number and case:
5. a. Laeta agricola malas lupas duas necat.
‘The happy farmer kills the two bad wolves.’
b. Amicae feminae duae aegras aegricolas duas curant.
‘The two friendly women take care of the two sick farmers.’
Phrase Structure Rules for Latin:
6. a. S NP + VP
b. VP NP + V
n
c. NP ADJ 0 + N + (NUM)

7.
Rules for Case Marking and Rules for Agreement:
8. NP dominated by S is subject; all elements of that NP are in the nominative case (–Ø if
singular, –e if plural).
9. NP dominated by VP is object; all elements of that NP are in the accusative case (–m if
singular, –s if plural).
10. If the subject is singular, verb ends in –t; if the subject is plural, verb ends in –nt.
After these rules apply to the above tree, the grammar generates the sentence Aminae feminae duae
aegras agricolas duas curant.
Vocabulary:
11. Nouns (quoted in the nominative singular feminine form):
lupa ‘wolf’ luna ‘moon’ regina ‘queen’ agricola ‘farmer’
stella ‘star’ casa ‘cottage’ fabula ‘story’ domina ‘mistress’
12. Adjectives (quoted in the nominative singular feminine form):
mala ‘bad’ pulchra ‘pretty’ multa ‘many’ clara ‘famous’
bona ‘good’ amica ‘friendly’ inimica ‘unfriendly’ laeta ‘happy’
13. Verbs (quoted in the infinitive form):
liberare ‘free’ vocare ‘call’ laudare ‘praise’ aedificare ‘build’
necare ‘kill’ curare ‘care for’ adorare ‘worship’ superare ‘conquer’
14. Others:
non ‘not’ et ‘and’
305

306
APPENDIX C: SUMMARY OF TREE STRUCTURES
1. Traditional grammar (confuses functional and structural categories):
2. Structural grammar and TG (not enough internal levels for phrases):
3. X–bar grammar and RG:
4. Sentence with a transitive verb (John studies Latin):

5. Noun phrase structure (John’s study of Latin...):
6. Noun phrase structure (all the/Ø study of Latin...):
307

308
7. Sentence with an intransitive (John died happy/happily) or copulative verb (John was happy):

SUPPLEMENT ONE: GRAMMAR REVIEW
SYNTACTIC CATEGORIES (THE PARTS OF SPEECH)
NOUN: Nouns are identifiable on the basis of several criteria. First, they can occur after definite
articles, e.g., the book, and after indefinite articles, e.g., a book. Second, noun are inflected to show
number distinctions, e.g., one book versus two books, and to indicate possession, e.g., the book’s
cover. Third, nouns fulfill a number of specific functions in sentences. For instance, they can be
used as the subject of a verb, e.g., The book ends sadly, as the object of a verb, e.g., John likes that
book, and as the object of a preposition, e.g., There are no pictures in the book. Nouns are variously
subcategorized as follows:
PROPER NOUNS are those which refer to specific individuals, places, or things, e.g., John,
Rome, and English; COMMON NOUNS are nonspecific in reference, e.g., man, city, and
language.
CONCRETE NOUNS are those which refer to tangible objects, e.g., cow, fence, and
painting; ABSTRACT NOUNS name intangible things, e.g., admiration, gravity, and
painting. Notice that some nouns have both concrete and abstract references.
COUNT NOUNS are those which are numerable, e.g., tree, leg, and linguist; MASS NOUNS
are not numerable, e.g., butter, water, and furniture. Notice that one does not ordinarily say
butters, waters, and furnitures (but, the waters of the Nile).
. COLLECTIVE NOUNS are those which refer to groups of items, e.g., herd, crowd, and mob.
PRONOUN: A pronoun is a word which is used in place of a noun phrase, e.g., he, she, who,
myself, etc. The various subclasses of pronouns are as follows:
PERSONAL PRONOUNS refer to the speaker, the hearer, or the person or thing spoken
about, e.g., I, me, you, and him.
REFLEXIVE PRONOUNS are object pronouns which refer back to the subject of the clause
in which they occur, e.g., myself, itself, and themselves.
INTENSIVE PRONOUNS emphasize a preceding noun or pronoun, e.g. myself, itself, and
themselves. The form of reflexive and intensive pronouns is the same. In I myself saw myself
in the mirror, the first myself is intensive, the second is reflexive.
INDEFINITE PRONOUNS refer to indefinite persons and things, e.g., someone, anybody,
and everything.
POSSESSIVE PRONOUNS are those which take the place of possessive nouns, e.g., her,
his, and its.

310
DEMONSTRATIVE PRONOUNS are used to designate particular persons or things, e.g.,
this, that, and these.
RELATIVE PRONOUNS are those which introduce relative clauses, e.g., who, whom, and
which.
INTERROGATIVE PRONOUNS are used in place of nouns in questions, e.g., who, which
and what.
RECIPROCAL PRONOUNS express mutual actions or relationships, e.g., each other and
one another.
VERB: A verb is a word which serves to express an action, an occurrence, or a state of being, e.g.,
shot in The inspector shot the bandit, blow up in The school blew up, be in That girl is clever. Verbs
are identifiable by a variety of inflectional distinctions, including tense, person, number, and voice.
The subclasses of verbs are as follows:
TRANSITIVE VERBS are those which take objects, e.g., break in John broke the vase, and
destroy in John destroyed the vase.
INTRANSITIVE VERBS are those which do not take objects, e.g., break in The vase broke,
and disappear in The vase disappeared.
COPULATIVE VERBS are those which link the subject with a nominal or adjectival
complement, e.g., appear in John appears healthy, and be in John was the king for two years.
ADJECTIVE: An adjective is a word which qualifies, limits, or otherwise modifies the meaning
of a noun, e.g. pretty, little, and white. In English, adjectives occur in two principal positions, either
before the nouns they modify, e.g., the rebellious student, or after verbs like be, e.g., the student is
rebellious. The two positions are called, respectively, the ATTRIBUTIVE POSITION and the
PREDICATE POSITION. Adjectives are inflected to show one of three degrees of comparison:
the POSITIVE DEGREE, the COMPARATIVE DEGREE, and the SUPERLATIVE DEGREE.
The comparative degree is used when two items are being compared; the superlative is used when
three or more items are being compared. For example: John is a good boy (positive degree); John
is a better boy than Bill (comparative degree); John is the best boy in the class (superlative degree).
ADVERB: An adverb is a word which modifies a verb or an adjective, e.g., quickly in John ran
quickly, and unbelievably in John is unbelievably clever. Like adjectives, adverbs are compared in
the positive, comparative, and superlative degree, e.g., Mary cleans her house thoroughly all the time
(positive), Mary cleans her house more thoroughly on weekdays than on weekends (comparative),
and Mary cleans her house (the) most thoroughly before her mother–in–law comes to visit
(superlative).

CONJUNCTION: A conjunction is a word which connects one grammatical category or
grammatical construction with another. There are two basic types of conjunctions:
COORDINATING and SUBORDINATING. A coordinating conjunction connects elements in
such a way as to indicate approximately equal order or rank amongst the elements, e.g., John and
Harry, on the table or on the desk, and Sit down and eat your spinach. A subordinating conjunction
on the other hand, connects elements in such a way as to place one in a lower order or rank with
respect to the other, e.g., Sit down, while you are eating your spinach.
PREPOSITION: A preposition is a word which introduces a phrase ending in a noun, or the
equivalent of a noun, and which typically serves as an adverbial or adjectival modifier, e.g., John ran
into the yard, and John is in the yard.
SYNTACTIC CONSTRUCTIONS
SENTENCE: A sentence is a structured expression consisting minimally of two main constituents,
a SUBJECT, which specifies the topic of the expression, and a PREDICATE, which specifies what
is asserted of the topic. There are four principal types of sentences: DECLARATIVE, those which
make statements, e.g., John wrote Helen a letter; INTERROGATIVE, those which ask questions,
e.g., Did John write Helen a letter; IMPERATIVE, those which give a command or make a request,
e.g., Please, write Helen a letter; and EXCLAMATORY, those which express strong or sudden
feeling, e.g., What a beautiful letter John wrote Helen!
CLAUSE: Like sentences, clauses are also expressions which contain a subject and a predicate. All
clauses must contain a tensed verb. Clauses are of two types: INDEPENDENT and DEPENDENT
(or SUBORDINATE). An independent clause is one which can stand alone; hence, a sentence is
an independent clause. A dependent clause is one which occurs in conjunction with an independent
clause. When a sentence consists of only one independent clause, it is called a SIMPLE sentence;
when it consists of two or more independent clauses and no dependent clauses, it is called a
COMPOUND sentence; and when it contains one or more dependent clauses, it is called a
COMPLEX sentence.
PHRASE: A phrase is any group of words which serves as a unit within a clause. Phrases are
usually named for their most distinctive element, called the HEAD, e.g., noun phrase, verb phrase,
prepositional phrase, and so on.
311

312
MORPHOLOGY
Most grammars divide the study of sentence structure into two areas, MORPHOLOGY, which
deals with the internal structure of words, and SYNTAX, which deals with the way words are
combined to form sentences. In morphology, the basic units under investigation are the morphemes
of a language. By definition, a MORPHEME is a minimal unit of meaning, that is, a meaningful
sequence of sounds which is not divisible into smaller meaningful units, e.g., ball. It is important
not to equate morphemes with either words or syllables; in both theory and practice, there is no
necessary coincidence between a morpheme and either of these units. For example, the English noun
hands is one word consisting of one syllable, but having two morphemes: hand, which defines a
specific body part, and s which indicates plural. Conversely, finger has two syllables but is only one
morpheme. One cannot divide the word finger into smaller meaningful units: the sequences fing,
inger, ger, etc. are all without content. Even the sequence fin, which does have meaning in English,
is not part of the morphological structure of finger, because its meaning is unrelated to the meaning
of the whole word.
Morphemes are variously classified by linguists as BOUND or FREE and as ROOTS or AFFIXES.
A BOUND MORPHEME is one which cannot occur as an independent word, e.g., the un– meaning
‘not’ in untrustworthy, unappreciative, unnecessary, etc. A FREE MORPHEME is one which can
stand alone, e.g., the morpheme trust. In English, bound morphemes constitute a relatively small
class of items, while the number of free morphemes is very large. In other languages, the situation
is often quite different. For example, Classical Greek has very few free morphemes, but a great
number of bound ones. In Vietnamese, almost all of the morphemes are classified as free.
A ROOT is a morpheme which constitutes the core element of a word or group of words of closely
related meaning. For example, the morpheme trust serves as the root in all of the following words:
trusted, untrustworthy, mistrustful, entrust, and so on. An AFFIX is a bound morpheme added to
the beginning of a word (in which case it is called a PREFIX) or to the end of a word (in which case
it is called a SUFFIX). Some common affixes in English are un–, mis–, con–, ex–, –ly, –ing, –ness,
and –er. That part of a word to which affixes are added is called a STEM. In English, the stem and
the root of a word are usually the same string; in other languages, they differ.
Affixes that are added to roots to indicate grammatical relationships are known as
INFLECTIONAL ENDINGS, and the process of combining these endings with roots is called
INFLECTION. The –s used to indicate the plural of English nouns is an example of an inflectional
ending, as is the –ed used to indicate past tense. Comparatively speaking, English is very poor in
inflectional endings. Other languages have highly developed inflectional systems. For example, in
Latin, Russian, and Finnish, the form of a noun changes depending on whether it is used as the
subject of a verb or as the object. In English, nouns are not inflected for this particular contrast,
although some pronouns are, e.g., I/me, she/her, he/him, we/us, they/them, who/whom. When applied
to SUBSTANTIVES (nouns, pronouns, and adjectives) inflection is referred to as DECLENSION;
when applied to verbs, it is called CONJUGATION.

SOME INFLECTIONAL CATEGORIES
PERSON: the inflectional category in pronouns and verbs that refers to the three–way distinction
between the speaker (FIRST PERSON), e.g., I am tall, the hearer (SECOND PERSON), e.g., You
are tall, and someone or something else (THIRD PERSON), e.g., He (she, it) is tall.
NUMBER: an inflectional category in substantives and verbs that refers to the distinction between
singular, e.g., the house IS big, and plural, e.g., the houses ARE big.
TENSE: an inflectional category in verbs which indicates distinctions in the TIME (PRESENT,
PAST, FUTURE) and the ASPECT (PROGRESSIVE, PERFECTIVE) of an action or state. For
example, the verb phrase is looking is in the present progressive tense, that is, it indicates an action
which is going on in the present; the verb phrase has looked is in the present perfective tense, that
is, it indicates an action that was completed in the past. Strictly speaking, tense is only partially an
inflectional category in English, since it is signaled both by independent words, e.g., forms of the
verbs be and have and by endings, e.g., –ing and –ed.
VOICE: an inflectional category in verbs that refers to the distinction between ACTIVE and
PASSIVE. An active verb is one whose subject is viewed as performing the action it defines, e.g.,
John killed the tigers; a passive verb is one whose subject is viewed as undergoing the action it
defines, e.g., the tigers were killed by John.
MODE/MOOD: an inflectional category in verbs that refers to the distinction between
INDICATIVE, CONDITIONAL, and IMPERATIVE. The indicative mode expresses ordinary
statements (he stopped); the conditional mode expresses conditions (he would stop); the imperative
mode expresses commands (stop!).
CASE: an inflectional category which indicates the relationship of substantives to other words in
a sentence. In English, the opposition he/him/his is a case opposition: he is used to indicate the
subject (NOMINATIVE CASE), him to indicate the object (OBJECTIVE or ACCUSATIVE
CASE), and his to indicate the possessor (POSSESSIVE or GENITIVE).
GENDER: an inflectional category in substantives that refers to the tripartite distinction between
MASCULINE, FEMININE, and NEUTER. In English, only the third person singular pronouns
show gender distinctions, e.g., he, she, and it; moreover, the distinction is based on sex. In other
languages, e.g., German, gender is a grammatical category of substantives and bears no relationship
to sex.
313

314
SUPPLEMENT TWO (PART I): TYPOLOGICAL CLASSIFICATION
ANALYTIC LANGUAGES (ISOLATING LANGUAGES): words consist of single
morphemes; most words consist only of a root. Mandarin Chinese, Vietnamese, Cantonese,
Cambodian. Examples in Mandarin adapted from Norman J., Chinese, Cambridge, 1988:
1. Ta chî fàn le.
he eat food past
‘He ate the food.’
2. Ta chî le fàn.
he eat past food
‘He ate the food.’
3. Júzi w chî le.
orange I eat past
‘I ate the orange.’
AGGLUTINATING LANGUAGES: words consist of a stem and one or more clearly
identifiable affixes. Finnish, Hungarian, Estonian, Swahili, Turkish. Examples in Estonian
adapted from Oinas, F., Basic Course in Estonian, Indiana University, 1966:
1. Ta on kohvikus.
he is coffee–house–in
‘He is in the coffee house.’
2. Lähme kohvikusse.
go–we coffee–house–into
‘Let us go into the coffee house.’
3. Ma tulen uuest kohvikust.
I come new–from coffee–house–from
‘I am coming out of the new coffee house.’
4. Cases:
nominative kohvik the coffee house
genitive kohviku of the coffee house
partitive kohvikut the coffee house
illative kohvikusse into the coffee house
inessive kohvikus in the coffee house
elative kohvikust out of the coffee house
allative kohvikule to the coffee house
adessive kohvikul by the coffee house
ablative kohvikult from the coffee house
translative kohvikuks for the coffee house
essive kohvikuna as the coffee house
terminative kohvikuni up to the coffee house
comitative kohvikuga with the coffee house
abessive kohvikuta without the coffee house

SUPPLEMENT TWO (PART II): TYPOLOGICAL CLASSIFICATION
INFLECTIONAL LANGUAGES (FUSIONAL LANGUAGES): words consist of stem and
affixes which often mark several grammatical categories simultaneously. Greek, Latin, Sanskrit,
Russian. Examples in Latin (NomSg=nominative singular, NomPl=nominative plural,
AccSg=accusative singular, AccPl=accusative plural, 3Sg=third person singular, 3Pl=third person
plural):
1. Custos fidelis consulem veterem ducit.
NomSg NomSg AccSg AccSg 3Sg
guard trusty consul old is leading
‘The trusty guard is leading the old consul.’
2. Custodes fideles consules veteres ducunt.
NomPl/AccPl NomPl/AccPl NomPl/AccPl NomPl/AccPl 3Pl
‘The trusty guards are leading the old consuls.’
‘The old guards are leading the trusty consuls.’
‘The trusty consuls are leading the old guards.’
‘The old consuls are leading the trusty guards.’
POLYSYNTHETIC LANGUAGES: words consist of long strings of stems and affixes, which
may translate as an entire English sentence. American Indian languages. Examples in Ayacucho
Quechua adapted from Parker, G., Ayacucho Quechua Grammar and Dictionary, Mouton, 1969:
Verbs can be inflected for both actor and object in different persons and number.
1. riku–yki ‘I see you.’
2. riku–yki–èik ‘I see you all.’
3. riku–yki–ku ‘We see you.’
riku: ‘see’
yki: first person singular actor and second person singular object
èik: marks object as plural
ku: marks actor as plural
4. riku–wanki ‘You see me.’
5. riku–wanki–èik ‘You all see me.’
6. riku–wanki–ku ‘We see you.’
wanki: second person singular actor and first person singular object
èik: marks actor as plural
ku: marks object as plural
315

316
GERMANIC:
SUPPLEMENT THREE (PART I): HISTORICAL CLASSIFICATION
LEXICON: INDO–EUROPEAN
love (n) mother father brother break (v) three five hundred heart
ENGLISH: love mother father brother break three five hundred heart
MID ENGLISH: love moder fader brother breke thre fîf hund(red) herte
OLD ENGLISH: lufu môdor fœder brôþor brecan þrî fîf(e) hund heorte
GERMAN: liebe mutter vater bruder brechen drei fünf hundert herz
DUTCH: liefde moeder vader broeder breken drie vijf honderd hart
GREEK:
ANCIENT GREEK: erôs mçtçr patçr adelphos hrçgnûmi treis pente hekaton kardia
CELTIC:
IRISH: searc mâthair athair brâthir brisim trî cóic cçt croidhe
WELSCH: serch mam tad brawd torri tri pump cant calon
ITALIC:
LATIN: amor mâter pater frater frangere trçs quinque centum cor
rumpere
ROMANCE:
SPANISH: amor madre padre hermano romper tres cinco ciento corazón
PORTUGUESE: amor mãe pai irmão quebrar três cinco cem coração
ITALIAN: amore madre padre fratello rompere tre cinque cento cuore
FRENCH: amour mère père frère rompre trois cinq cent coeur
RUMANIAN: amor mamã tatã frate frînge trei cinci sutã inimã
rupe
BALTO–SLAVIC:
RUSSIAN: ljubov' mat' otec brat lomat' tri pyat' sto serdce
CHURCH SLAVIC: ljuby mati otc brat lomiti trje petü sto srdce
INDO–IRANIAN:
SANSKRIT: kâma– mâtar– pitar– bhrâtar– bhañj– trayas páñca çatam hrd–
HINDI: muhabbat mátá báp bháí torná tîn pãc sau dil
AVESTAN: kan– mâtar– pitar– brâtar– sèand– râyô panèa satm zrd–

ALTAIC:
SUPPLEMENT THREE (PART II): HISTORICAL CLASSIFICATION
LEXICON: NON–INDO–EUROPEAN
love (n) mother father brother break (v) three five hundred heart
JAPANESE: ai haha chichi kyôdai kowareru san go hyaku shinzô
SINO–TIBETAN:
CHINESE: ai mchîsn fùchin syûngdì nùngpwò sân w bãi syîn
URALIC:
FINNISH: rakkaus äiti isä veli rikkoa kolme viisi sata sydän
AUSTRONESIAN:
SAMOAN: alofa tinâ tamâ 1 uso 4 talepe tolu lima selau fatu
HAWAIIAN: aloha makuahine makua 2 kua'ana 4 wâhi kolu lima hanele pu'uwai
AFRO–ASIATIC:
CHADIC:
5 3
HAUSA: so ûwa ôba wâ kária úku biál darí zútsia
SEMITIC:
ARABIC: hubb imm ab(u) ax yiksir talaate xams miyy alb
UTO–AZTECAN:
5
COMANCHE kamaku–ru– pia ahpu– samohpu– tahparu– pahitu– moobetu– makekitu– pihi
NOTES:
1
2
3
4
5
uso = brother of a man; tuagane = brother of a woman
kua'ana = older sibling of the same sex; kaina = younger sibling of the same sex; kuâne = sibling of a female;
etc.
wâ = older male sibling; kâne = younger male sibling
There are many verbs for break depending on the nature of the thing broken, e.g., hard things, brittle things,
things that break easily, bones, etc.
There is no noun meaning love; these are verb forms.
317

318
SUPPLEMENT FOUR (PART I): LANGUAGES OF AFRICA AND THE MID EAST*
Family Branch Group Representative Languages Principal Locations Number of Speakers
Afro–Asiatic
(241)
Chadic Hausa N Nigeria; Niger; Cameroon 39,000,000
Semitic Hebrew Isreal 5,000,000
Tigrinya S Eritrea; N Ethiopia 4,000,000
Amharic Ethopia 20,000,000
Arabic Egypt; Iraq; Syria; Morocco 230,000,000
Hamitic Berber Tamazight W Algeria; N Morocco 3,000,000
Kabyle W Kabylia; N Algeria 3,000,000
Shilha W Algeria; S Morocco 3,000,000
Cushitic Somali Somalia; Kenya; Ethiopia; Djibouti 4,000,000
Galla (Oromo) W Ethopia; N Kenya 9,000,000
Omotic Wolaytta SE Ethopia 2,000,000
Khoisan (31) Nama (Hottentot) Namibia 146,000
Niger–Congo
(1032)
Nilo–Sarahan
(138)
Sandawe Tanzania 70,000
Mande Mende Sierra Leone 2,000,000
West Atlantic Fula Senegal; Nigeria; Cameroon 13,000,000
Wolof Senegal 7,000,000
Kwa Akan Ghana; Côte d’Ivoire 7,000,000
Yoruba SW Nigeria; Benin 20,000,000
Igbo (Ibo) Nigeria 17,000,000
Ewe S Togo; SE Ghana 3,000,000
Benue–Congo Efik SE Nigeria 6,000,000
Tiv SE Nigeria; Cameroon 2.000,000
Bantu Swahili Tanzania; Kenya; Zaire; Uganda 49,000,000
Zulu S Africa; Lesotho 9,000,000
Xhosa S Africa 8,000,000
Gur (Voltaic) More Burkina Faso 4,000,000
Adamawa–Ubangian Sango Central African Republic 4,000,000
Nilo–Hamitic Maasai Kenya; Tanzania 689,000
Nilotic Luo Kenya; Nyanza; Tanzania 4,000,000
* There is much scholarly debate on the classification of languages. The various subdivisions in Figure Four I, II, III, IV, and V reflect only
major distinctions generally recognized. The approximate number of languages in each family is in parentheses (data from Ruhlen 1987).
Statistics for individual languages with more than 1 million speakers are from The World Almanac and Book of Facts, 1997, Mahwah, NJ:
World Almanac Books. Statistics for individual languages with less than 1 million speakers are from Grimes 1992.

SUPPLEMENT FOUR (PART II): AMERICAN INDIAN LANGUAGES
Family Branch Group Representative Languages Principal Locations Number of Speakers
Algonquian (15) Cree Canada; USA 46,700
Andean–Equatorial
(163)
Ojibwa Canada; USA 51,000
Arapaho Wyoming; W Oklahoma 1,500
Blackfoot S Alberta; Montana 9,000
Andean Quechua Peru; Equador; Bolivia; Argentina 8,000,000
Aymara Bolivia; Peru 2,000,000
Equatorial Guarani Paraguay 4,000,000
Eskimo–Aleut (9) Eskimo Inuit Canadian Arctic; Greenland 68,500
Yupik Alaska; Siberia 17,000
Aleut Aleut Aleutian Islands 700
Ge–Pano–Carib (117) Carib Northern South America 100,000
Hokan (28) Yuman Diegueño Baja California less than 400
Tequistlatecan Tequistlatec Mexico 5,000
Iroquoian (7) Mohawk N New York 3,000
Cherokee E Oklahoma; W North Carolina 22,500
Caddoan (4) Pawnee N Central Oklahoma 200
Na–Dené (34) Tlingit SE Alaska 2,000
Athapaskan Navajo New Mexico; Arizona; Utah 130,000
Apache Oklahoma; New Mexico; Arizona 14,300
Chipewyan N Alberta; Saskatchewan 4,000
Oto–Manguean (17) Otomian Zapotec S Mexico 500,000
Mixtecan Mixtec S Mexico 250,000
Penutian (68) Nez Perce N Idaho 500
Mayan Quiche Guatemala 885,000
Mosan (27) Squamish SW British Columbia 20
Siouan–Yuchi (11) Dakota (Sioux) S Manitoba; Nebraska; Minnesota 19,000
Crow S Montana 5,500
Winnebago Wisconsin; E Nebraska 1,500
Uto–Aztecan (25) Hopi NE Arizona 5,000
Shoshonean Comanche W Oklahoma 500
Takic Luiseño S California 100
Sonoran Huichol Mexico 12,500
Aztecan Nahuatl (Aztec) Southern Mexico 1,000,000
319

320
SUPPLEMENT FOUR (PART III): INDO–EUROPEAN LANGUAGES
Family Branch Group Representative Languages Principal Locations Number of Speakers
Indo–European
(144)
Albanian Albanian Albania; Yugoslavia 5,000,000
Anatolian Hittite Ancient Asia Minor Extinct
Armenian Armenian Armenia; Iran; Syria; Lebanon 5,000,000
Balto–Slavic Baltic Lithuanian Lithuania 3,000,000
Latvian Latvia 2,000,000
Slavic Russian Russia 280,000,000
Serbo–Croatian Serbia; Croatia; Bosnia 20,000,000
Czech Czechoslovakia 12,000,000
Polish Poland 44,000,000
Celtic Brythonic Welsh Wales 580,000
Goidelic Irish Ireland 120,000
Germanic East Gothic Eastern Europe Extinct
North Danish Denmark 5,000,000
Norwegian Norway 5,000,000
Swedish Sweden; Finland 9,000,000
West Dutch Netherlands; Belgium 21,000,000
English England, America 487,000,000
German Germany 124,000,000
Greek Greek Greece 12,000,000
Indo–Iranian Indic Sanskrit Ancient India Extinct
Bengali Bangladesh; India 204,000,000
Hindi India 457,000,000
Urdu Pakistan; India 104,000,000
Punjabi India; Pakistan 95,000,000
Iranian Pashto Afghanistan; Pakistan 21,000,000
Persian Iran; Afghanistan 35,000,000
Italic Latin Roman Empire Extinct
Romance French France 125,000,000
Romanian Romania; Moldova 26,000,000
Italian Italy 62,000,000
Portuguese Portugal; Brazil 186,000,000
Spanish Spain; Central and South America 401,000,000

SUPPLEMENT FOUR (PART IV): LANGUAGES OF EAST ASIA & THE PACIFIC
Family Branch Group Representative Languages Principal Locations Number of Speakers
Australian (170) Pama–Nyungan Walbiri Australia 3,000
Austro–Asiatic
(155)
Dyirbal Australia 40 to 50
Mon–Khmer Khmer Kampuchea; Vietnam; Thailand 9,000,000
Vietnamese Vietnam 63,000,000
Nung Vietnam; China 2,000,000
Munda Santali E India; Nepal 5,000,000
Nicobarese Central Nicobarese Nicobar Islands 5,000
Austronesian (959) Formosan Atayal Taiwan 41,000
Western Malay–Indonesian Indonesia; Thailand 164,000,000
Toba Batak Indonesia 4,000,000
Javanese Java 64,000,000
Sundanese Java 26,000,000
Madurese Madura 10,000,000
Tagalog Philippines 56,000,000
Cebuano Philippines 12,000,000
Malagasy Madagascar 12,000,000
Oceanic Samoan Samoa 328,000
Maori New Zealand 100,000
Tahitian Tahiti 125,000
Hawaiian Hawaii 2,000
Indo–Pacific (731) Motu Papua New Guinea 15,000
Sino–Tibetan (258) Tibeto–Burman Tibetan SW China; N India; Nepal 5,000,000
Burmese Mayanmar (Burma) 31,000,000
Newari Nepal 500,000
Sinitic Mandarin Mandarin China 999,000,000
Wu Wu Chiang Jiang delta, China 65,000,000
Min Taiwanese Taiwan 50,000,000
Yue Cantonese Guangdong province, China 70,000,000
Hakka Hakka SE China 34,000,000
Tai (57) Thai Thailand 52,000,000
Shan E Mayanmar (Burma) 3,000,000
Lao Thailand; Laos 4,000,000
321

322
SUPPLEMENT FOUR (PART V): OTHER LANGUAGES OF EUROPE AND ASIA
Family Branch Group Representative Languages Principal Locations Number of Speakers
Altaic (63) Turkic Turkish Turkey 61,000,000
Azerbaijani Azerbaijan; Iran; Turkey 15,000,000
Tartar Russia 8,000,000
Uzbek Uzbekistan 14,000,000
Mongolian Mongol Mongolian People’s Republic 6,000,000
Manchu–Tungus Evenki Russia 24,000
Manchu Northeast China less than 1,000
Japanese Japanese Japan 126,000,000
Korean Korean North Korea; South Korea 76,000,000
Caucasian (38) Kartvelian Georgian Georgia; Azerbaijan; Turkey 4,000,000
Svan Georgia 35,000
Northwest Kabardian Northwest Caucasus; Russia 370,00
Northeast Dagestanian Avar Northeast Caucasus 483,000
Lezgian (Lezgi) Russia; Azerbaijan 383,000
Nakh Chechen Chechnia 945,000
Dravidian (28) Telugu Southeast India 75,000,000
Tamil India; Sri Lanka 73,000,000
Kannada S India 45,000,000
Malayalam S India 36,000,000
Paleosiberian (5) Chukchi Northeast Siberia 10,900
Uralic (24) Finno–Ugric Finnic Finnish Finland 6,000,000
Estonian Estonia 1,000,000
Ugric Hungarian Hungary 14,000,000
Samoyedic Nenets Arctic Russia 27,000

SUPPLEMENT FIVE: LANGUAGES AND LANGUAGE GROUPS
(Roman numerals in square brackets refer to the Parts in Supplement Four)
Adamawa–Ubangian [I]
Afro–Asiatic [I]
Akan [I]
Albanian [III]
Aleut [II]
Algonquian [II]
Altaic [V]
Amharic [I]
Anatolian [III]
Andean [II]
Andean–Equatorial [II]
Apache [II]
Arabic [I]
Arapaho [II]
Armenian [III]
Atayal [IV]
Athapaskan [II]
Australian [IV]
Austronesian [IV]
Austro–Asiatic [IV]
Avar [V]
Aymara [II]
Azerbaijani [V]
Aztec [II]
Aztecan [II]
Baltic [III]
Balto–Slavic [III]
Bantu [I]
Bengali [III]
Benue–Congo [I]
Berber [I]
Blackfoot [II]
Brythonic [III]
Burmese [IV]
Caddoan [II]
Cantonese [IV]
Carib [II]
Caucasian [V]
Cebuano [IV]
Celtic [III]
Central Nicobarese [IV]
Chadic [I]
Chechen [V]
Cherokee [II]
Chipewyan [II]
Chukchi [V]
Comanche [II]
Cree [II]
Crow [II]
Cushitic [I]
Czech [III]
Dagestanian [V]
Dakota [II]
Danish [III]
Diegueño [II]
Dravidian [V]
Dutch [III]
Dyirbal [IV]
East Germanic [III]
Efik [I]
English [III]
Equatorial [II]
Eskimo [II]
Eskimo–Aleut [II]
Estonian [V]
Evenki [V]
Ewe [I]
Finnic [V]
Finnish [V]
Finno–Ugric [V]
Formosan [IV]
French [III]
Fula [I]
Galla [I]
Georgian [V]
German [III]
Germanic [III]
Ge–Pano–Carib [II]
Goidelic [III]
Gothic [III]
Greek [III]
Guarani [II]
Gur [I]
Hakka [IV]
Hamitic [I]
Hausa [I]
Hawaiian [IV]
Hebrew [I]
Hindi [III]
Hittite [III]
Hokan [II]
Hopi [II]
Hottentot [I]
Huichol [II]
Hungarian [V]
Ibo [I]
Igbo [I]
Indic [III]
Indo–European [III]
Indo–Iranian [III]
Indo–Pacific [IV]
Inuit [II]
Iranian [III]
Irish [III]
Iroquoian [II]
Italian [III]
Italic [III]
Japanese [V]
Javanese [IV]
Kabardian [V]
Kabyle [I]
Kannada [V]
Kartvelian [V]
Khmer [IV]
Khoisan [I]
Korean [V]
Kwa [I]
Lao [IV]
Latin [III]
Latvian [III]
Lezgi [V]
Lezgian [V]
Lithuanian [III]
323

324
Luiseño [II]
Luo [I]
Maasai [I]
Madurese [IV]
Malagasy [IV]
Malayalam [V]
Malay–Indonesian [IV]
Manchu [V]
Manchu–Tungus [V]
Mandarin [IV]
Mande [I]
Maori [IV]
Mayan [II]
Mende [I]
Min [IV]
Mixtec [II]
Mixtecan [II]
Mohawk [II]
Mongol [V]
Mongolian [V]
Mon–Khmer [IV]
More [I]
Mosan [II]
Motu [IV]
Munda [IV]
Nahuatl [II]
Nakh [V]
Nama [I]
Navajo [II]
Na–Dené [II]
Nenets [V]
Newari [IV]
Nez Perce [II]
Nicobarese [IV]
Niger–Congo [I]
Nilotic [I]
Nilo–Hamitic [I]
Nilo–Sarahan [I]
North Germanic [III]
Northeast Caucasian [V]
Northwest Caucasian [V]
Norwegian [III]
Nung [IV]
Oceanic [IV]
Ojibwa [II]
Omotic [I]
Oromo [I]
Otomian [II]
Oto–Manguean [II]
Paleosiberian [V]
Pama–Nyungan [IV]
Pashto [III]
Pawnee [II]
Penutian [II]
Persian [III]
Polish [III]
Portuguese [III]
Punjabi [III]
Quechua [II]
Quiche [II]
Romance [III]
Romanian [III]
Russian [III]
Samoan [IV]
Samoyedic [V]
Sandawe [I]
Sango [I]
Sanskrit [III]
Santali [IV]
Semitic [I]
Serbo–Croatian [III]
Shan [IV]
Shilha [I]
Shoshonean [II]
Sinitic [IV]
Sino–Tibetan [IV]
Siouan–Yuchi [II]
Sioux [II]
Slavic [III]
Somali [I]
Sonoran [II]
Spanish [III]
Squamish [II]
Sundanese [IV]
Svan [V]
Swahili [I]
Swedish [III]
Tagalog [IV]
Tahitian [IV]
Tai [IV]
Taiwanese [IV]
Takic [II]
Tamazight [I]
Tamil [V]
Tartar [V]
Telugu [V]
Tequistlatec [II]
Tequistlatecan [II]
Thai [IV]
Tibetan [IV]
Tibeto–Burman [IV]
Tigrinya [I]
Tiv [I]
Tlingit [II]
Toba Batak [IV]
Turkic [V]
Turkish [V]
Ugric [V]
Uralic [V]
Urdu [III]
Uto–Aztecan [II]
Uzbek [V]
Vietnamese [IV]
Voltaic [I]
Walbiri [IV]
Welsh [III]
West Germanic [III]
West Atlantic [I]
Western Austronesian [IV]
Winnebago [II]
Wolaytta [I]
Wolof [I]
Wu [IV]
Xhosa [I]
Yoruba [I]
Yue [IV]
Yuman [II]
Yupik [II]
Zapotec [II]
Zulu [I]

SUPPLEMENT SIX: SOME ALPHABETS
ROMAN GREEK CYRILLIC HEBREW ARABIC
A a alpha a aleph,alef alif
B b beta be bth b
C c gamma ve gimel t
D d delta ge dleth th
E e epsilon de h jim
F f zeta je vav, waw h
G g eta jo zayin kh
H h theta že heth di
I i iota ze teth dhi
J j kappa i yod, yodh r
K k lambda í krátkj , kph zy
L l mu ka lamedh sin
M m nu el , mm shin
N n xi em nn sd
O o omicron en samekh dd
P p pi o ayin t
Q q rho pe , p z
R r , sigma er , sade, sadhe yn
S s tau es qph ghayn
T t upsilon te rsh f
U u phi u sin qf
V v chi ef shin kaf
W w psi xa tv, tw lam
X x omega ce mim
Y y a nun
Z z ša h
š a ww
tv órd–ij znák y
–i
m áxkj znák
e
ju
ja
325

326

SUPPLEMENT SEVEN: SAMPLE PARSES FROM THE LANGTECH PARSER
SAMPLE PARSES ILLUSTRATING BASIC SENTENCE PATTERNS
(1) INTRANSITIVE:
the senator resigned
The clause “the senator resigned” is a statement.
The phrase “the senator” is a noun phrase and the subject of the verb “resigned.”
The word “resigned” is a past tense verb.
The word “resigned” is the main verb of “the senator resigned.”
The word “senator” is a singular count noun.
The word “senator” is the head of the noun phrase “the senator.”
The word “the” is a determiner specifying the noun “senator.”

328
(2) TRANSITIVE (DIRECT OBJECT COMPLEMENT ON V1):
the senator admired the president
The clause “the senator admired the president” is a statement.
The phrase “the president” is a noun phrase and the direct object of the verb “admired.”
The phrase “the senator” is a noun phrase and the subject of the verb “admired.”
The word “admired” is a past tense verb.
The word “admired” is the main verb of “the senator admired the president.”
The word “president” is a singular count noun.
The word “president” is the head of the noun phrase “the president.”
The word “senator” is a singular count noun.
The word “senator” is the head of the noun phrase “the senator.”
The word “the” is a determiner specifying the noun “president.”
The word “the” is a determiner specifying the noun “senator.”

(3) COPULATIVE (PREDICATE COMPLEMENT ON V2):
the senator became the president
The clause “the senator became the president” is a statement.
The phrase “the president” is a noun phrase describing the subject “senator.”
The phrase “the senator” is a noun phrase and the subject of the verb “became.”
The word “became” is a past tense verb.
The word “became” is the main verb of “the senator became the president.”
The word “president” is a singular count noun.
The word “president” is the head of the noun phrase “the president.”
The word “senator” is a singular count noun.
The word “senator” is the head of the noun phrase “the senator.”
The word “the” is a determiner specifying the noun “senator.”
The word “the” is a determiner specifying the noun “president.”
329

330
(4) COPULATIVE (PREDICATE COMPLEMENT ON V2):
the senator turned red
The clause “the senator turned red” is a statement.
The phrase “red” is an adjective phrase describing the subject “senator.”
The phrase “the senator” is a noun phrase and the subject of the verb “turned.”
The word “red” is an adjective.
The word “senator” is a singular count noun.
The word “senator” is the head of the noun phrase “the senator.”
The word “the” is a determiner specifying the noun “senator.”
The word “turned” is a past tense verb.
The word “turned” is the main verb of “the senator turned red.”

(5) INTRANSITIVE WITH A MANNER ADVERB:
the senator resigned angrily
The clause “the senator resigned angrily” is a statement.
The phrase “angrily” is an adverb phrase modifying the verb “resigned.”
The phrase “the senator” is a noun phrase and the subject of the verb “resigned.”
The word “angrily” is a manner adverb.
The word “resigned” is a past tense verb.
The word “resigned” is the main verb of “the senator resigned angrily.”
The word “senator” is a singular count noun.
The word “senator” is the head of the noun phrase “the senator.”
The word “the” is a determiner specifying the noun “senator.”
331

332
(6) INTRANSITIVE WITH PREDICATE ADJECTIVE:
the senator resigned angry
The clause “the senator resigned angry” is a statement.
The phrase “angry” is a predicate adjective referring back to the subject “senator.”
The phrase “the senator” is a noun phrase and the subject of the verb “resigned.”
The word “angry” is an adjective.
The word “resigned” is a past tense verb.
The word “resigned” is the main verb of “the senator resigned angry.”
The word “senator” is a singular count noun.
The word “senator” is the head of the noun phrase “the senator.”
The word “the” is a determiner specifying the noun “senator.”

(7) TRANSITIVE (DIRECT OBJECT + MANNER ADVERB):
fred left the country quickly
The clause “fred left the country quickly” is a statement.
The phrase “fred” is a noun phrase and the subject of the verb “left.”
The phrase “quickly” is an adverb phrase modifying the verb “left.”
The phrase “the country” is a noun phrase and the direct object of the verb “left.”
The word “country” is a singular count noun.
The word “country” is the head of the noun phrase “the country.”
The word “fred” is a proper noun.
The word “left” is a past tense verb.
The word “left” is the main verb of “fred left the country quickly.”
The word “quickly” is a manner adverb.
The word “the” is a determiner specifying the noun “country.”
333

334
(8) TRANSITIVE (DIRECT OBJECT + MANNER ADVERB):
fred left the country happily
The clause “fred left the country happily” is a statement.
The phrase “fred” is a noun phrase and the subject of the verb “left.”
The phrase “happily” is an adverb phrase modifying the verb “left.”
The phrase “the country” is a noun phrase and the direct object of the verb “left.”
The word “country” is a singular count noun.
The word “country” is the head of the noun phrase “the country.”
The word “fred” is a proper noun.
The word “happily” is a manner adverb.
The word “left” is a past tense verb.
The word “left” is the main verb of “fred left the country happily.”
The word “the” is a determiner specifying the noun “country.”

(9) fred left the country happy
a. TRANSITIVE (DIRECT OBJECT + OBJECTIVE COMPLEMENT):
The clause “fred left the country happy” is a statement.
The phrase “fred” is a noun phrase and the subject of the verb “left.”
The phrase “happy” is an adjective phrase describing “country.”
The phrase “the country” is a noun phrase and the direct object of the verb “left.”
The word “country” is the head of the noun phrase “the country.”
The word “country” is a singular count noun.
The word “fred” is a proper noun.
The word “happy” is an adjective.
The word “left” is a past tense verb.
The word “left” is the main verb of “fred left the country happy.”
The word “the” is a determiner specifying the noun “country.”
335

336
b. TRANSITIVE (DIRECT OBJECT + SUBJECTIVE COMPLEMENT):
The clause “fred left the country happy” is a statement.
The phrase “fred” is a noun phrase and the subject of the verb “left.”
The phrase “happy” is a predicate adjective referring back to the subject “fred.”
The phrase “the country” is a noun phrase and the direct object of the verb “left.”
The word “country” is the head of the noun phrase “the country.”
The word “country” is a singular count noun.
The word “fred” is a proper noun.
The word “happy” is an adjective.
The word “left” is a past tense verb.
The word “left” is the main verb of “fred left the country happy.”
The word “the” is a determiner specifying the noun “country.”

(10) TRANSITIVE (DIRECT OBJECT + INDIRECT OBJECT):
she gave them to us
The clause “she gave them to us” is a statement.
The phrase “she” is a noun phrase and the subject of the verb “gave.”
The phrase “them” is a noun phrase and the direct object of the verb “gave.”
The phrase “to us” is a prepositional phrase and the indirect object of the verb “gave.”
The phrase “us” is a noun phrase and the object of the preposition “to.”
The word “gave” is a past tense verb.
The word “gave” is the main verb of “she gave them to us.”
The word “she” is a third person singular feminine subject pronoun.
The word “them” is a third person plural object pronoun.
The word “to” is a preposition introducing the prepositional phrase “to us"
The word “us” is a first person plural object pronoun.
337

338
(11) TRANSITIVE (DIRECT OBJECT + PREPOSITIONAL PHRASE):
she put them near us
The clause “she put them near us” is a statement.
The phrase “near us” is a prepositional phrase and a complement of the verb “put.”
The phrase “she” is a noun phrase and the subject of the verb “put.”
The phrase “them” is a noun phrase and the direct object of the verb “put.”
The phrase “us” is a noun phrase and the object of the preposition “near.”
The word “near” is a preposition introducing the prepositional phrase “near us"
The word “put” is a past tense verb.
The word “put” is the main verb of “she put them near us.”
The word “she” is a third person singular feminine subject pronoun.
The word “them” is a third person plural object pronoun.
The word “us” is a first person plural object pronoun.

(12) TRANSITIVE (DOUBLE OBJECT):
she gave us them
The clause “she gave us them” is a statement.
The phrase “she” is a noun phrase and the subject of the verb “gave.”
The phrase “them” is a noun phrase and the direct object of the verb “gave.”
The phrase “us” is a noun phrase and the indirect object of the verb “gave.”
The word “gave” is a past tense verb.
The word “gave” is the main verb of “she gave us them.”
The word “she” is a third person singular feminine subject pronoun.
The word “them” is a third person plural object pronoun.
The word “us” is a first person plural object pronoun.
339

340
(13) TRANSITIVE (DOUBLE OBJECT):
they elected a democrat the president
The clause “they elected a democrat the president” is a statement.
The phrase “a democrat” is a noun phrase and the direct object of the verb “elected.”
The phrase “the president” is a noun phrase and the object complement of the verb “elected.”
The phrase “they” is a noun phrase and the subject of the verb “elected.”
The word “a” is a determiner specifying the noun “democrat.”
The word “democrat” is a singular count noun.
The word “democrat” is the head of the noun phrase “a democrat.”
The word “elected” is a past tense verb.
The word “elected” is the main verb of “they elected a democrat the president.”
The word “president” is a singular count noun.
The word “president” is the head of the noun phrase “the president.”
The word “the” is a determiner specifying the noun “president.”
The word “they” is a third person plural subject pronoun.

(14) TRANSITIVE (SENTENTIAL COMPLEMENT; INDIRECT STATEMENT):
I know that she went
The clause “I know that she went” is a statement.
The clause “she went” is an indirect statement.
The clause “that she went” is a complement clause of the verb “know.”
The phrase “I” is a noun phrase and the subject of the verb “know.”
The phrase “she” is a noun phrase and the subject of the verb “went.”
The word “I” is a first person singular subject pronoun.
The word “know” is a present tense verb.
The word “know” is the main verb of “I know that she went.”
The word “she” is a third person singular feminine subject pronoun.
The word “that” is a complementizer specifying the complement clause “she went.”
The word “went” is a past tense verb.
The word “went” is the main verb of “that she went.”
341

342
(15) TRANSITIVE (SENTENTIAL COMPLEMENT; INDIRECT STATEMENT):
I know she went
The clause “I know she went” is a statement.
The clause “she went” is an indirect statement.
The phrase “I” is a noun phrase and the subject of the verb “know.”
The phrase “she went” is a complement clause of the verb “know.”
The word “I” is a first person singular subject pronoun.
The word “know” is a present tense verb.
The word “know” is the main verb of “I know she went.”
The word “Ø” is a complementizer specifying the complement clause “she went.”
The word “she” is a third person singular feminine subject pronoun.
The word “went” is a past tense verb.
The word “went” is the main verb of “she went.”

(16) TRANSITIVE (SENTENTIAL COMPLEMENT; INDIRECT STATEMENT):
I know that man is mortal
The clause “I know that man is mortal” is a statement.
The clause “man is mortal” is an indirect statement.
The clause “that man is mortal” is a complement clause of the verb “know.”
The phrase “I” is a noun phrase and the subject of the verb “know.”
The phrase “man” is a noun phrase and the subject of the verb “is.”
The phrase “mortal” is a predicate adjective referring back to the subject “man.”
The word “I” is a first person singular subject pronoun.
The word “is” is a third person singular present tense verb.
The word “is” is the main verb of “man is mortal.”
The word “know” is a present tense verb.
The word “know” is the main verb of “I know that man is mortal.”
The word “man” is the head of the noun phrase “man.”
The word “mortal” is an adjective.
The word “Ø” is an abstract empty determiner specifying the noun “man.”
The word “that” is a complementizer specifying the complement clause “man is mortal.”
343

344
The clause “I know that man is mortal” is a statement.
The clause “that man is mortal” is a complement clause of the verb “know.”
The clause “that man is mortal” is an indirect statement.
The phrase “I” is a noun phrase and the subject of the verb “know.”
The phrase “mortal” is a predicate adjective referring back to the subject “man.”
The phrase “that man” is a noun phrase and the subject of the verb “is.”
The word “I” is a first person singular subject pronoun.
The word “is” is a third person singular present tense verb.
The word “is” is the main verb of “that man is mortal.”
The word “know” is a present tense verb.
The word “know” is the main verb of “I know that man is mortal.”
The word “man” is the head of the noun phrase “that man.”
The word “mortal” is an adjective.
The word “Ø” is a complementizer specifying the complement clause “that man is mortal.”
The word “that” is a determiner specifying the noun “man.”

(17) TRANSITIVE (SENTENTIAL COMPLEMENT; INDIRECT QUESTION):
I know whether she went
The clause “I know whether she went” is a statement.
The clause “whether she went” is an indirect question.
The phrase “I” is a noun phrase and the subject of the verb “know.”
The phrase “she” is a noun phrase and the subject of the verb “went.”
The word “I” is a first person singular subject pronoun.
The word “know” is a present tense verb.
The word “know” is the main verb of “I know whether she went.”
The word “she” is a third person singular feminine subject pronoun.
The word “went” is a past tense verb.
The word “went” is the main verb of “whether she went.”
The word “whether” is an interrogative pronoun.
345

346
(18) TRANSITIVE (SENTENTIAL COMPLEMENT; INDIRECT QUESTION):
I know where she went
The clause “I know where she went” is a statement.
The clause “where she went” is an indirect question.
The phrase “I” is a noun phrase and the subject of the verb “know.”
The phrase “she” is a noun phrase and the subject of the verb “went.”
The phrase “[e]” is the referent of “where” and the empty modifier phrase of the verb “went.”
The word “I” is a first person singular subject pronoun.
The word “know” is a present tense verb.
The word “know” is the main verb of “I know where she went.”
The word “she” is a third person singular feminine subject pronoun.
The word “went” is a past tense verb.
The word “went” is the main verb of “where she went.”
The word “where” is an interrogative pronoun.

(19) TRANSITIVE (SENTENTIAL COMPLEMENT; INDIRECT QUESTION):
I know who went
The clause “I know who went” is a statement.
The clause “who went” is an indirect question.
The phrase “I” is a noun phrase and the subject of the verb “know.”
The phrase “[e]” is the referent of “who” and the empty subject noun phrase of the verb “went.”
The word “I” is a first person singular subject pronoun.
The word “know” is a present tense verb.
The word “know” is the main verb of “I know who went.”
The word “went” is a past tense verb.
The word “went” is the main verb of “who went.”
The word “who” is an interrogative pronoun.
347

348
(20) TRANSITIVE (SENTENTIAL COMPLEMENT; INDIRECT QUESTION):
I know who he admires
The clause “I know who he admires” is a statement.
The clause “who he admires” is an indirect question.
The phrase “I” is a noun phrase and the subject of the verb “know.”
The phrase “[e]” is an abstract empty noun phrase and the direct object of the verb “admires.”
The word “admires” is a third person singular present tense verb.
The word “admires” is the main verb of “who he admires.”
The word “he” is a third person singular masculine subject pronoun.
The word “I” is a first person singular subject pronoun.
The word “know” is a present tense verb.
The word “know” is the main verb of “I know who he admires.”
The word “who” is an interrogative pronoun.

(21) TRANSITIVE (DIRECT OBJECT + S COMPLEMENT; INDIRECT STATEMENT):
he told me that she went
The clause “he told me that she went” is a statement.
The clause “that she went” is a complement clause of the verb “told.”
The clause “she went” is an indirect statement.
The phrase “he” is a noun phrase and the subject of the verb “told.”
The phrase “me” is a noun phrase and the direct object of the verb “told.”
The phrase “she” is a noun phrase and the subject of the verb “went.”
The word “he” is a third person singular masculine subject pronoun.
The word “me” is a first person singular object pronoun.
The word “she” is a third person singular feminine subject pronoun.
The word “that” is a complementizer specifying the complement clause “she went.”
The word “told” is a past tense verb.
The word “told” is the main verb of “he told me that she went.”
The word “went” is a past tense verb.
The word “went” is the main verb of “that she went.”
349

350
(22) TRANSITIVE (DIRECT OBJECT + S COMPLEMENT; INDIRECT STATEMENT):
he told me she went
The clause “he told me she went” is a statement.
The clause “that she went” is a complement clause of the verb “told.”
The clause “she went” is an indirect statement.
The phrase “he” is a noun phrase and the subject of the verb “told.”
The phrase “me” is a noun phrase and the direct object of the verb “told.”
The word “he” is a third person singular masculine subject pronoun.
The word “me” is a first person singular object pronoun.
The word “Ø” is a complementizer specifying the complement clause “she went.”
The word “she” is a third person singular feminine subject pronoun.
The word “told” is a past tense verb.
The word “told” is the main verb of “he told me she went.”
The word “went” is a past tense verb.
The word “went” is the main verb of “that she went.”

(23) TRANSITIVE (DIRECT OBJECT + S COMPLEMENT; INDIRECT QUESTION):
he told me whether she went
The clause “he told me whether she went” is a statement.
The clause “whether she went” is a complement clause of the verb “told.”
The clause “whether she went” is an indirect question.
The phrase “he” is a noun phrase and the subject of the verb “told.”
The phrase “me” is a noun phrase and the direct object of the verb “told.”
The phrase “she” is a noun phrase and the subject of the verb “went.”
The word “he” is a third person singular masculine subject pronoun.
The word “me” is a first person singular object pronoun.
The word “she” is a third person singular feminine subject pronoun.
The word “told” is a past tense verb.
The word “told” is the main verb of “he told me whether she went.”
The word “went” is a past tense verb.
The word “went” is the main verb of “whether she went.”
The word “whether” is an interrogative pronoun.
351

352
(24) TRANSITIVE (DIRECT OBJECT + S COMPLEMENT; INDIRECT QUESTION):
he told me where she went
The clause “he told me where she went” is a statement.
The clause “where she went” is a complement clause of the verb “told.”
The clause “where she went” is an indirect question.
The phrase “he” is a noun phrase and the subject of the verb “told.”
The phrase “me” is a noun phrase and the direct object of the verb “told.”
The phrase “she” is a noun phrase and the subject of the verb “went.”
The phrase “[e]” is the referent of “where” and the empty modifier phrase of the verb “went.”
The word “he” is a third person singular masculine subject pronoun.
The word “me” is a first person singular object pronoun.
The word “she” is a third person singular feminine subject pronoun.
The word “told” is a past tense verb.
The word “told” is the main verb of “he told me where she went.”
The word “went” is a past tense verb.
The word “went” is the main verb of “where she went.”
The word “where” is an interrogative pronoun.

(25) TRANSITIVE (DIRECT OBJECT + S COMPLEMENT; INDIRECT QUESTION):
he told me who [e] went
The clause “he told me who went” is a statement.
The clause “who went” is an indirect question.
The phrase “me” is a noun phrase and the direct object of the verb “told.”
The phrase “he” is a noun phrase and the subject of the verb “told.”
The phrase “[e]” is the referent of “who” and the empty subject noun phrase of the verb “went.”
The word “me” is a first person singular object pronoun.
The word “he” is a third person singular masculine subject pronoun.
The word “told” is a past tense verb.
The word “told” is the main verb of “he told me who went.”
The word “went” is a past tense verb.
The word “went” is the main verb of “who went.”
The word “who” is an interrogative pronoun.
353

354
(26) TRANSITIVE (DIRECT OBJECT + S COMPLEMENT; INDIRECT QUESTION):
he told me who she admires
The clause “he told me who she admires” is a statement.
The clause “who she admires” is an indirect question.
The phrase “me” is a noun phrase and the direct object of the verb “told.”
The phrase “he” is a noun phrase and the subject of the verb “told.”
The phrase “[e]” is an abstract empty noun phrase and the direct object of the verb “admires.”
The word “admires” is a third person singular present tense verb.
The word “admires” is the main verb of “who she admires.”
The word “he” is a third person singular masculine subject pronoun.
The word “me” is a first person singular object pronoun.
The word “she” is a third person singular feminine subject pronoun.
The word “told” is a past tense verb.
The word “told” is the main verb of “he told me who she admires.”
The word “who” is an interrogative pronoun.

(27) TRANSITIVE (PREPOSITIONAL PHRASE OBJECT):
he relies on her advice
The clause “he relies on her advice” is a statement.
The phrase “he” is a noun phrase and the subject of the verb “relies.”
The phrase “her advice” is a noun phrase and the object of the preposition “on.”
The phrase “her” is a possessive phrase specifying “advice.”
The phrase “on her advice” is a prepositional phrase and a complement of the verb “relies.”
The word “advice” is a singular mass noun.
The word “advice” is the head of the noun phrase “her advice.”
The word “he” is a third person singular masculine subject pronoun.
The word “her” is a third person singular feminine possessive pronoun.
The word “on” is a preposition introducing the prepositional phrase “on her advice.”
The word “POS” is a determiner specifying the noun “advice.”
The word “relies” is a third person singular present tense verb.
The word “relies” is the main verb of “he relies on her advice.”
355

356
(28) INTRANSITIVE (PREPOSITIONAL PHRASE MODIFIER):
he went on her advice
The clause “he went on her advice” is a statement.
The phrase “he” is a noun phrase and the subject of the verb “went.”
The phrase “her advice” is a noun phrase and the object of the preposition “on.”
The phrase “her” is a possessive phrase specifying “advice.”
The phrase “on her advice” is a prepositional phrase and a modifier of the verb “went.”
The word “advice” is a singular mass noun.
The word “advice” is the head of the noun phrase “her advice.”
The word “he” is a third person singular masculine subject pronoun.
The word “her” is a third person singular feminine possessive pronoun.
The word “on” is a preposition introducing the prepositional phrase “on her advice.”
The word “POS” is a determiner specifying the noun “advice.”
The word “went” is a past tense verb.
The word “went” is the main verb of “he went on her advice.”

(29) TRANSITIVE (PREPOSITIONAL PHRASE OBJECT):
she sat on the hat
The clause “she sat on the hat” is a statement.
The phrase “on the hat” is a prepositional phrase and a complement of the verb “sat.”
The phrase “she” is a noun phrase and the subject of the verb “sat.”
The phrase “the hat” is a noun phrase and the object of the preposition “on.”
The word “hat” is a singular count noun.
The word “hat” is the head of the noun phrase “the hat.”
The word “on” is a preposition introducing the prepositional phrase “on the hat.”
The word “sat” is the main verb of “she sat on the hat.”
The word “sat” is a past tense verb.
The word “she” is a third person singular feminine subject pronoun.
The word “the” is a determiner specifying the noun “hat.”
357

358
(30) TRANSITIVE (PARTICLE + DIRECT OBJECT):
she put on the hat
The clause “she put on the hat” is a statement.
The phrase “she” is a noun phrase and the subject of the verb “put.”
The phrase “the hat” is a noun phrase and the direct object of the verb “put.”
The word “hat” is a singular count noun.
The word “hat” is the head of the noun phrase “the hat.”
The word “on” is a particle (intransitive preposition).
The word “put” is the main verb of “she put on the hat.”
The word “put” is a past tense verb.
The word “she” is a third person singular feminine subject pronoun.
The word “the” is a determiner specifying the noun “hat.”

(31) TRANSITIVE (DIRECT OBJECT + PARTICLE):
she put the hat on
The clause “she put the hat on” is a statement.
The phrase “she” is a noun phrase and the subject of the verb “put.”
The phrase “the hat” is a noun phrase and the direct object of the verb “put.”
The word “hat” is a singular count noun.
The word “hat” is the head of the noun phrase “the hat.”
The word “on” is a particle (intransitive preposition).
The word “put” is a past tense verb.
The word “put” is the main verb of “she put the hat on.”
The word “she” is a third person singular feminine subject pronoun.
The word “the” is a determiner specifying the noun “hat.”
359

360
(32) TRANSITIVE (DIRECT OBJECT + PARTICLE):
she put the hat right on
The clause “she put the hat right on” is a statement.
The phrase “right” is an adverb phrase modifying the particle “on.”
The phrase “she” is a noun phrase and the subject of the verb “put.”
The phrase “the hat” is a noun phrase and the direct object of the verb “put.”
The word “hat” is a singular count noun.
The word “hat” is the head of the noun phrase “the hat.”
The word “on” is a particle (intransitive preposition).
The word “put” is a past tense verb.
The word “put” is the main verb of “she put the hat right on.”
The word “she” is a third person singular feminine subject pronoun.
The word “right” is an adverb.
The word “the” is a determiner specifying the noun “hat.”

(33) TRANSITIVE (COMPLEMENTARY INFINITIVE):
she asked to go
The clause “she asked to go” is a statement.
The phrase “she” is a noun phrase and the subject of the verb “asked.”
The phrase “to go” is an infinitive complement of the verb “asked.”
The word “asked” is a past tense verb.
The word “asked” is the main verb of “she asked to go.”
The word “go” is an infinitive verb.
The word “she” is a third person singular feminine subject pronoun.
The word “to” is the infinitive marker introducing the infinitive phrase “to go.”
361

362
(34) TRANSITIVE (DIRECT OBJECT + COMPLEMENTARY INFINITIVE):
she asked him to go
The clause “she asked him to go” is a statement.
The phrase “him” is a noun phrase and subject of the infinitive verb “go.”
The phrase “him” is a noun phrase and the direct object of the verb “asked.”
The phrase “she” is a noun phrase and the subject of the verb “asked.”
The phrase “to go” is an infinitive complement of the verb “asked.”
The word “asked” is a past tense verb.
The word “asked” is the main verb of “she asked him to go.”
The word “go” is an infinitive verb.
The word “him” is a third person singular masculine object pronoun.
The word “she” is a third person singular feminine subject pronoun.
The word “to” is the infinitive marker introducing the infinitive phrase “to go.”

(35) TRANSITIVE (PREPOSITIONAL OBJECT WITH INFINITIVE):
she prayed for him to win
The clause “she prayed for him to win” is a statement.
The phrase “for him to win” is a prepositional phrase and a complement of the verb “prayed.”
The phrase “him” is a noun phrase and the subject of the verb “win.”
The phrase “him” is a noun phrase and the object of the preposition “for.”
The phrase “she” is a noun phrase and the subject of the verb “prayed.”
The word “for” is a preposition introducing the prepositional phrase “for him to win.”
The word “him” is a third person singular masculine object pronoun.
The word “prayed” is a past tense verb.
The word “prayed” is the main verb of “she prayed for him to win.”
The word “she” is a third person singular feminine subject pronoun.
The word “to” is the infinitive marker introducing the infinitive phrase “to win.”
The word “win” is an infinitive verb.
363

364
(36) TRANSITIVE (PREPOSITIONAL OBJECT WITH INFINITIVE):
she relied on him to finish
The clause “she relied on him to finish” is a statement.
The phrase “him” is a noun phrase and the object of the preposition “on.”
The phrase “him” is a noun phrase and the subject of the verb “finish.”
The phrase “on him to finish” is a prepositional phrase and a complement of the verb “relied.”
The phrase “she” is a noun phrase and the subject of the verb “asked.”
The word “finish” is an infinitive verb.
The word “him” is a third person singular masculine object pronoun.
The word “on” is a preposition introducing the prepositional phrase “on him to finish.”
The word “relied” is a past tense verb.
The word “relied” is the main verb of “she relied on him to finish.”
The word “she” is a third person singular feminine subject pronoun.
The word “to” is the infinitive marker introducing the infinitive phrase “to finish.”
What are the TG equivalents of these two trees?

SUPPLEMENT EIGHT: SAMPLE PARSES FROM THE LANGTECH PARSER
SAMPLE PARSES ILLUSTRATING NOUN PHRASES
(1) The Spanish American history teacher resigned.
history teacher who is Spanish and American teacher of American history who is Spanish
history teacher who is Spanish American teacher of Spanish American history
365

366
(2) The kitchen appliance factory guarantee expired.
(3) All the men’s money disappeared.

(4) All those many nosey house guests with excessive demands are a pain in the neck.
(5) The quick brown fox jumped over the lazy dog.
367

368
(6) The soprano from a small town in the south of France collapsed.

SUPPLEMENT NINE: SAMPLE PARSES FROM THE LANGTECH PARSER
(1) Is this the violin Kevin Played?
EMPTY CATEGORIES: Ø, [u], and [e]
a. This violin is the violin.
this violin PRS is the violin
b. This is the violin.
this [u] PRS is the violin
c. Is this the violin?
is PRS this [u] [e] [e] the violin
i j j i
d. Is this the violin which Kevin played?
is PRS this [u] [e] [e] the violin which Ø Kevin played [e]
i j j i k k k
e. Is this the violin that Kevin played?
is PRS this [u] [e] [e] the violin [e] that Kevin played [e]
i j j i k k k
f. Is this the violin Kevin played?
is PRS this [u] [e] [e] the violin [e] Ø Kevin played [e]
i j j i k k k
(2) X3 Representaton.
369

370
(2) We believe John to be completely dominated by his wife.
An [e] is coindexed with its closest C-commanding N3.

(3) Saturn’s rings are believed to be completely dominated by its moons.
An [e] is coindexed with its closest C-commanding N3.
371

372
SUPPLEMENT TEN: SAMPLE PARSES FROM THE LANGTECH PARSER
SAMPLE PARSES ILLUSTRATING VERBALS
The internal structure of all –ing and –ed/en forms are as follows:
(1) Progressive Participles (-ing) and Derived Adjectives (DA1).
a. Participle (the lady charming the man): [ C3 [ C0 [ C0 -ing] [ V3 [ V0 charming]]]]
b. Adjective (a very charming lady): [ [ [ [ charming]] [ DA1]]]
C3 C0 V3 V0 C0

(2) Passive Participles (-en) and Derived Adjectives (DA2).
a. Participle (charmed by her flattery): [ C3 [ C0 [ C0 -en] [ V3 [ V0 charmed]]]]
b. Derived Adjective (a very charmed life): [ [ [ [ charmed]] [ DA2]]]
C3 C0 V3 V0 C0
373

374
(3) Gerundial Nominals (-ing) and Derived Nominals (DN).
a. -ing (charming candy from a baby): [ N3 [ N0 [ N0 -ing] [ V3 [ V0 charming]]]]
b. DN: (such charming of snakes): [ [ [ [ charming]] [ DN]]]
N3 N0 V3 V0 N0

(4) Summary of Verbal Structures.
a. Verb resides in posthead position.
[ C3 [ C0 [ C0 -ing] [ V3 [ V0 charming]] ]]
[ C3 [ C0 [ C0 -en] [ V3 [ V0 charmed]] ]]
[ [ [ -ing] [ [ charming]] ]]
N3 N0 N0 V3 V0
b. Verb resides in prehead position.
[ C3 [ C0 [ V3 [ V0 charming]] [ C0 DA1] ]]
[ C3 [ C0 [ V3 [ V0 charmed]] [ C0 DA2] ]]
[ [ [ [ charming]] [ DN] ]]
N3 N0 V3 V0 N0
Notice that the four possible ing–forms in English have the feature specification [–VBL] (they are
subjuncts); two are [–NML] (the derived adjective and the present participle); two are [+NML] (the
derived nominal and the gerundial nominal). Further, the two members of each pair are mirror
images of each other. Thus, all are subjuncts with a V3 embedded on their X0 level.
The nominal nature of gerunds is accounted for by embedding that V3 into an N3, and the adjectival
nature of participles is accounted for by embedding that V3 into a C3.
375

376
(5) TRANSITIVE (DIRECT OBJECT + BARE INFINITIVE VERB):
i saw him cheer
The clause “I saw him cheer” is a statement.
The phrase “cheer” is the bare infinitive complement of the verb “saw.”
The phrase “him” is a noun phrase and the direct object of the verb “saw.”
The phrase “I” is a noun phrase and the subject of the verb “saw.”
The word “cheer” is a bare infinitive verb.
The word “him” is a third person singular masculine object pronoun.
The word “I” is a first person singular subject pronoun.
The word “saw” is a past tense verb.
The word “saw” is the main verb of “I saw him cheer.”

(6) TRANSITIVE (DIRECT OBJECT + PASSIVE PARTICIPLE):
i saw him cheered
The clause “I saw him cheered” is a statement.
The phrase “cheered” is a passive participle phrase describing “him.”
The phrase “him” is a noun phrase and the direct object of the verb “saw.”
The phrase “I” is a noun phrase and the subject of the verb “saw.”
The phrase “[e]” is an empty noun phrase and the direct object of the verb “cheered.”
The word “cheered” is a passive participle.
The word “him” is a third person singular masculine object pronoun.
The word “I” is a first person singular subject pronoun.
The word “saw” is a past tense verb.
The word “saw” is the main verb of “I saw him cheered.”
The word “[-ed]” is the abstract marker for the passive participle.
377

378
(7) TRANSITIVE (DIRECT OBJECT + PROGRESSIVE PARTICIPLE):
i saw him cheering
The clause “I saw him cheering” is a statement.
The phrase “cheering” is a progressive participle phrase describing “him.”
The phrase “him” is a noun phrase and the direct object of the verb “saw.”
The phrase “I” is a noun phrase and the subject of the verb “saw.”
The word “cheering” is a progressive participle.
The word “him” is a third person singular masculine object pronoun.
The word “I” is a first person singular subject pronoun.
The word “saw” is a past tense verb.
The word “saw” is the main verb of “I saw him cheering.”
The word “[-ing]” is the abstract marker for the progressive participle.
The clause “I saw him cheering” is a statement.
The phrase “cheering” is a progressive participle phrase describing “I.”
The phrase “him” is a noun phrase and the direct object of the verb “saw.”
The phrase “I” is a noun phrase and the subject of the verb “saw.”
The word “cheering” is a progressive participle.
The word “him” is a third person singular masculine object pronoun.
The word “I” is a first person singular subject pronoun.
The word “saw” is a past tense verb.
The word “saw” is the main verb of “I saw him cheering.”
The word “[-ing]” is the abstract marker for the progressive participle.

(8) AMBIGUITY:
i heard the tenor singing
The clause “I heard the tenor singing” is a statement.
The phrase “I” is a noun phrase and the subject of the verb “heard.”
The phrase “the tenor singing” is a noun phrase and the direct object of the verb “heard.”
The word “heard” is a past tense verb.
The word “heard” is the main verb of “I heard the tenor singing.”
The word “I” is a first person singular subject pronoun.
The word “singing” is a singular count noun.
The word “singing” is the head of the noun phrase “the tenor singing.”
The word “tenor” is a noun forming a compound noun with “singing.”
The word “tenor” is a singular count noun.
The word “tenor” is the head of the noun phrase “tenor.”
The word “the” is a determiner specifying the noun “singing.”
The clause “I heard the tenor singing” is a statement.
The phrase “I” is a noun phrase and the subject of the verb “heard.”
The phrase “singing” is a progressive participle phrase describing “tenor.”
The phrase “the tenor singing” is a noun phrase and the direct object of the verb “heard.”
The word “heard” is a past tense verb.
The word “heard” is the main verb of “I heard the tenor singing.”
The word “I” is a first person singular subject pronoun.
The word “singing” is a progressive participle.
The word “tenor” is a singular count noun.
The word “tenor” is the head of the noun phrase “the tenor singing.”
The word “the” is a determiner specifying the noun “tenor.”
The word “[-ing]” is the abstract marker for the progressive participle.
379

380
The clause “I heard the tenor singing” is a statement.
The phrase “I” is a noun phrase and the subject of the verb “heard.”
The phrase “singing” is a progressive participle phrase describing “tenor.”
The phrase “the tenor” is a noun phrase and the direct object of the verb “heard.”
The word “heard” is a past tense verb.
The word “heard” is the main verb of “I heard the tenor singing.”
The word “I” is a first person singular subject pronoun.
The word “singing” is a progressive participle.
The word “tenor” is a singular count noun.
The word “tenor” is the head of the noun phrase “the tenor.”
The word “the” is a determiner specifying the noun “tenor.”
The word “[-ing]” is the abstract marker for the progressive participle.
The clause “I heard the tenor singing” is a statement.
The phrase “I” is a noun phrase and the subject of the verb “heard.”
The phrase “singing” is a progressive participle phrase describing “i.”
The phrase “the tenor” is a noun phrase and the direct object of the verb “heard.”
The word “heard” is a past tense verb.
The word “heard” is the main verb of “I heard the tenor singing.”
The word “I” is a first person singular subject pronoun.
The word “singing” is a progressive participle.
The word “tenor” is a singular count noun.
The word “tenor” is the head of the noun phrase “the tenor.”
The word “the” is a determiner specifying the noun “tenor.”
The word “[-ing]” is the abstract marker for the progressive participle.

OUTLINE OF TECHNICAL TERMS
(1) Phonological Feature and Phonological Category (Natural Class of sounds).
a. Phonological Feature: a phonological feature ó is a feature from the set of
primitives {CONSONANTAL, VOCALIC, NASAL,
VOICED...}.
b. Phonological Category: a phonological category is a phonological feature matrix Ó
consisting of pairs óì, where ó is a phonological feature and
where ì å {+, –}.
1. CONSONANT: [+CONSONANTAL, –VOCALIC]
2. VOWEL: [–CONSONANTAL, +VOCALIC]
3. GLIDE: [–CONSONANTAL, –VOCALIC]
4. LIQUID: [+CONSONANTAL, +VOCALIC]
(2) Semantic Feature and Semantic Category (Thematic Relation).
a. Semantic Feature: a semantic feature ó is a feature from the set of primitives
{POSITIONAL, DISJUNCTIVE, CONJUNCTIVE...}.
b. Semantic Category: a semantic category is a semantic feature matrix Ó consisting
of pairs óì, where ó is a semantic feature and where ì å {+,
–}.
1. SOURCE: [+DISJUNCTIVE, –CONJUNCTIVE]
2. GOAL: [–DISJUNCTIVE, +CONJUNCTIVE]
3. LOCATION: [–DISJUNCTIVE, –CONJUNCTIVE]
4. NONLOCATION: [+DISJUNCTIVE, +CONJUNCTIVE]
(3) Syntactic Feature and Syntactic Category (Part of Speech).
a. Syntactic Feature: a syntactic feature ó is a feature from the set of primitives
{NUMBER, TENSE,...}.
b. Syntactic Category: a syntactic category is a syntactic feature matrix Ó consisting
of pairs ìó, where ì å {+, –, 1, 2, 3, etc.} and where ó is a
syntactic feature.
1. Nouns: [+NOMINAL, –VERBAL]
2. Verbs: [–NOMINAL, +VERBAL]
3. Characterizers: [–NOMINAL, –VERBAL]
4. Gerunds: [+NOMINAL, +VERBAL]

382
(4) Structural Representation: a labeled bracketing of the form [...X...], where and X are
syntactic feature matrices.
a. Abstract Representation (AR): the structural representation containing all categories
(including empty categories) before the application of phonological rules, e.g., rules that
distribute affixes properly.
b. Phonetic Representation (PR): the structural representation after the application of
phonological rules.
(5) Projection, Phrase, Unit, Level Designator and Head.
a. Projection: a projection of a syntactic feature matrix Ó0, specified for a lexical entry,
is a syntactic hierarchy that includes Ó0 and (potentially) other syntactic feature
matrices. In English, the maximal projection of a syntactic feature matrix is Ó3, and the
minimal projection is Ó0.
b. Phrase: a syntactic feature matrix Ó3 representing the maximal projection of Ó0.
c. Unit: any sequence of syntactic feature matrices that is exhaustively dominated by some
Ón, where Ó is the projection of a syntactic feature matrix, where n is the Level
Designator, and where n å {0, 1, 2, 3,}.
d. Head: the terminal minimal projection 0.
(6) Dominance, Precedence, and Recursion.
a. Dominance: In a structure of the form [...X...], if and X are syntactic feature
matrices and if X is contained within the hierarchy , then dominates X. In such a
case, one also says X is bounded by .
b. Immediate Dominance: In a structure of the form [...X...], if dominates X and if
there is no syntactic matrix which dominates X and is dominated by , then
immediately dominates X.
c. Precedence: In a structure of the form [...X...Y...], if and X and Y are syntactic
feature matrices and if X is to the left of Y, then X precedes Y.
d. Recursion: In a structure of the form [...X...], if dominates X and if = X, then
is a recursion of X and, equivalently, X is a recursion of .

(7) Command.
a. C–command: A syntactic category C–commands a syntactic category if does not
dominate , if does not dominate , and if the first branching syntactic category n
dominating also dominates .
b. Immediate C–command: A syntactic category immediately C–commands a syntactic
category if C–commands and there is no syntactic category that C–commands
and is C–commanded by .
c. L–command: a category L–commands another category if the first branching
category n above dominates and if is an anaphor to the left of the head of n.
d. R–command: a category R–commands another category if the first branching
category n above dominates and if is an anaphor to the right of the head of n.
(8) Residence, Residential Domains, and Neighborhood.
a. Residence: A residence is a specific position in syntactic structure associated with a
specific grammatical function (Subject, Object, Complement, etc.) or semantic relation
(Quantification, Specification, Modification, etc.). A resident is a phrase occupying a
residence. Some important residents for English are:
X3: specifiers (determiners, possessive case, tense)
X2: modifiers (adjectives, manner adverbs, predicate nominatives)
X1: complements (direct objects, secondary complements)
X0: head
1. Free Resident: freely overt or empty, e.g., I believe (that) he went.
2. Permanent Resident: obligatorily overt, e.g., I whispered (*that) he went.
b. Residential Domain (RD).
1. The RD of a free resident includes all items it L–commands.
2. The RD of a permanent resident includes all items it C–commands.
c. Immediate Neighborhood: An immediate neighborhood embraces all phrases that are
immediately C–commanded by the same C–Specifier (DET or POS in N3; CPL or TNS
in V3).
d. Extended Neighborhood: An extended neighborhood embraces all X3 recursions
dominating the immediate neighborhood.
383

384
(9) Binding Relations:
a. Chain: A syntactic category and a syntactic category form a chain (of binding) if
and have equivalent syntactic feature matrices, if dominates a fully specified lexical
item or WH–phrase, if is an anaphor, and if can occur in the same context as .
b. Linked Chain: Syntactic categories , , and form a linked chain (of binding) if and
form a chain and if, when is substituted for , and form a chain.
(10) Binding Resident (BR), Barrier, and Barrier BR.
a. A Binding Resident (BR) is a C-Specifier with a fixed position in abstract
representations. This includes DET, POS, CPL, and TNS.
b. A barrier is any category which blocks the formation of binding relations.
[1] If a barrier is a free resident (optional), its residential domain includes all items
it L–commands.
[2] If a barrier is a permanent resident (obligatory), its residential domain includes all
items it either L–commands or R–commands.
[3] A binding relationship cannot be formed with an empty category that falls within
a residential domain, that is, [e] can be anywhere except in a residential domain.
c. A barrier BR is a BR with a fixed position in both abstract and superficial
representations ([+X3l, –X2L, –X1L, +PRH, –PSH, ENH, –EVH, –ECH]). This
includes DET, POS, and CPL. A barrier BR tolerates no prehead or posthead elements
on any of its own internal levels ([–OPH]) and belongs to a closed class of morphemes
([–OCL]). The structure of a barrier BR is simply [ C0].
(11) General Conditions on Representations.
a. The Empty Category Condition (ECC): An empty category and its referent must have
the same syntactic feature matrix, that is, they must be categories of the same type.
b. The Extended Neighborhood Condition (ENC): A WH-phrase or focused phrase in the
extended neighborhood must be bound to an [e] that it C–commands.
c. The Recoverability of Deletion Condition (RDC): All gaps in sentences must be
recoverable.
C3

d. The Contraction Block Condition (CBC):
Contraction is not possible over an empty category.
e. The Specifier Heaviness Condition (SHC): The heavier a specifier, the more it is a
barrier to binding. The heaviness of specifiers increases as one proceeds down the list
(“a” to “c”) for N3 and for V3 as follows:
N3: a. The indefinite determiners Ø and a/an.
b. Definite determiners (the, that, etc.)
c. POS.
V3: a. The complementizer Ø.
b. The complementizer that.
c. The complementizer if and WH–phrases.
f. The Distinct Reference Condition (DRC).
No N3–[e] chain can fulfill more than one grammatical function for the same verb.
g. The Double Trigger Condition (DTC).
[1] Clause triggers (WH–phrases and complementizers) must reside in the extended
neighborhood.
[2] The extended neighborhood can contain only one overt clause trigger.
h. The [e]–Binding Condition (EBC).
[e] (=trace) cannot reside in an RD bound to a referent outside that RD.
i. The One Affix Condition (OAC).
385
A word cannot be marked by more than one grammatical marker from the same set of
mutually exclusive grammatical markers, i.e., no more than one number, no more than
one tense, no more than one case, etc.
j. The Strict Order Condition (SOC).
The order of lexical items in any AR–PR pair must be the same. AR=Abstract
representation; PR=Phonetic Representation.

386
k. The Subject Exclusion Condition (SEC).
The passive morpheme PSP is a barrier to subject assignment.
l. The X Projection Condition (XPC).
The projections of a syntactic feature matrix Ó are Ó0, Ó1, Ó2, Ó3, where Ó = Ó.
m. The X3 Cue Condition (XCC).
Every embedded N3 and V3 with a common head must contain a cue.
n. The X3 Specifier Condition (XSC).
[1] A head can be specified by at most one overt N3 quantifier, one overt N3 subject,
one overt C3 determiner, and one overt C3 TENSE/MODE.
[2] All N3 must be specified by a C3 determiner; all V3 must be specified by a
TENSE/MODE (unless ruled out by independent principles such as the OAC).
(12) The English Noun Phrase Condition (NPC).
a. If the head of a noun phrase is a singular count noun, e.g., a noun like book, then the
noun phrase must have an overt determiner (a word like the or a); if the head is plural
(books), the determiner need not be overt (it can be Ø, the phonologically null DET).
b. Between the determiner and the head noun of a noun phrase, there can be no plural
noun.
c. Between the determiner and the head noun of a noun phrase, no other determiner can
occur.
d. Any item which might occur to the left of the determiner must be separated from it by
the preposition of.
e. A noun head must be separated from posthead elements by some overt marker such as
a preposition or complementizer.

RESIDENTIAL GRAMMAR (RG) SUMMARY
(1) Every word and abstract category (TNS, POS, Ø, [e], [u], etc.) must be dominated by an X0
which must be dominated by an X3.
(2) Every sentence (V3) has a subject, a tense, and a main verb.
(3) A phrase will contain an X2 and/or X1 level only if the phrase has X2 and/or X1 level
elements. Some phrases are head-only, i.e., they never have an X2 or X1 (determiners,
modals).
(4) Each higher level of a phrase must be greater than or equal to the level below, that is, all the
specifiers must be higher than all the modifiers which, in turn, must be higher than all the
complements.
a. OK: N3 above N3 above N2 above N2 above N2 above N1 above N1 above N0
b. *: N3 above N2 above N3 above N2 above N1 above N2 above N1 above N0
(5) There are three major syntactic categories: N (Noun), V (Verb), and C (Characterizer).
a. V verb, modal, auxiliary (be and have), participle, infinitive
b. N noun, pronoun, quantifier, gerund
c. C adjective, (manner and sentence) adverb, preposition, particle, tense, possessive
marker, determiner, degree words, emphatics, conjunctions
(6) There are direct correspondences between RG and TG.
a. V3 = S, MODP, etc. N3 = NP, QP, etc. C3 = PP, ADJP, ADVP, etc.
b. V2 = VP, MOD, etc. N2 = N, Q, etc. C2 = P, ADJ, ADV, etc.
c. V1 = V, MOD, etc. N1 = N, Q, etc. C1 = P, ADJ, ADV, etc.
d. V0 = V, MOD, etc. N0 = N, Q, etc. C0 = P, ADJ, ADV, etc.
(7) modals: can, could, will, would, shall, should, may, might, must, do, does, did
determiners: the, this, these, that, those, a, an, no
quantifiers: all, some, each, every, none, much, many
degree words: so, as, too, that, this, how
emphatics: so, too, not
prepositions: in, out, off, on, up, down, about, from, with, by, at, before, after, etc.
tense (TNS): present (PRS), past (PST), future (FUT), conditional (CND), imperative (IMP)
coordinators: and, but, or, etc.
subordinators: because, since, although, if, unless, until, before, after, etc.
387

388
(8) X3 elements are SPECIFIERS N3 V3
Determiners Subject
Possessive marker Tenses
Quantifiers
X2 elements are MODIFIERS N2 V2
Numerals Time & Place Phrases
Adjectives Adverbs
Predicate nominatives
X1 elements are COMPLEMENTS N1 V1
Complements Complements
Direct objects Direct objects
Compound elements
X0 elements are elements of words N0 V0
Particles Particles
(9) a. nouns:
[1] noun1 singular count nouns show, play
[2] noun2 plural count nouns shows. plays
[3] noun3 mass nouns rice, bread
[4] noun4 proper nouns fred, lucy
b. verbs:
[1] verb1 plural present tense verbs go, show, play
[2] verb2 singular present tense verbs goes, shows, plays
[3] verb3 past tense verbs went, showed, played
[4] verb4 passive and perfective participles gone, shown, played
[5] verb5 progressive participle going, showing, playing
c. adjectives:
[1] adj1 positive adjectives good, happy
[2] adj2 comparative adjectives better, happier
[3] adj3 superlative adjectives best, happiest
[4] adj4 proper adjectives spanish, american
d. adverbs:
[1] adv1 positive manner adverbs well, happily
[2] adv2 comparative manner adverbs better
[3] adv3 superlative manner adverbs best
[4] adv4 sentence adverbs probably

(10) Traditional Analysis:
SYNTACTIC REPRESENTATION
SUBJECT = Person or thing performing the action of the verb.
OBJECT = Person or thing affected by the action of the verb.
Phrase Structure Rules (parentheses indicate optional elements):
SENTENCE SUBJECT + AUXILIARY + PREDICATE
PREDICATE VERB + (OBJECT)
AUXILIARY MODAL
SUBJECT (DET) + NOUN
OBJECT (DET) + NOUN
Problem: Confusion of structure and function
(a) The student will burn the books. He/she will burn them.
(b) The student will read the books. He/she will read them.
(c) The student will have the books. He/she will have them.
Problem: Tense is not a separate element of aux
(a) The student does read the books. Present Tense
(b) The student did read the books. Past Tense
(c) The student will read the books. Future Tense
389

390
(11) TG (Transformational Grammar) Analysis I:
S = SENTENCE
AUX = AUXILIARY TNS = TENSE
PRS = PRESENT TENSE
PST = PAST TENSE
FUT = FUTURE TENSE
VP = VERB PHRASE V = VERB
NP = NOUN PHRASE N = NOUN
Phrase Structure Rules:
S NP + AUX + VP
VP V + (NP)
AUX (MODAL) + TNS
NP (DET) + N
Problem: Phrasal structure is very different for different phrases.

(12) TG (Transformational Grammar) Analysis II:
XP = ANY PHRASE X = HEAD OF THE PHRASE
VP = VERB PHRASE V = VERB
NP = NOUN PHRASE N = NOUN
TP = TENSE PHRASE T = TENSE
DP = DETERMINER PHRASE D = DETERMINER
MP = MODAL PHRASE M = MODAL
Phrase Structure Rules:
S NP + TP + VP
VP V + (NP)
TP (MP) + T
MP M
NP (DP) + N
DP D
Problem: Phrasal structure is very different for different phrases.
XP (XP) + X + (XP)
391

392
(13) RG (Residential Grammar) Analysis:
V3 = VERB PHRASE V = VERBS, AUXILIARIES, MODALS
N3 = NOUN PHRASE N = NOUNS, PRONOUNS
C3 = CHARACTERIZER PHRASE C = DETERMINERS,ADJ E C T IV E S ,
ADVERBS, TENSE
X3 = SPECIFIER LEVEL DETERMINERS, SUBJECT, TENSE
X2 = MODIFIERS LEVEL ADJECTIVES, ADVERBS
X1 = COMPLEMENT LEVEL DIRECT OBJECTS
X0 = HEAD/WORD LEVEL COMPONENTS OF WORDS
Phrase Structure Frame:
n m
[X (N3) + (C3) + X + (N3) + (C3) ]
where m n

SAMPLE PARSE ROUTINE
CATEGORY AND FRAME LOOKUP
word the student will read the book
category DET3 NOUN1 MODAL MVERB DET3 NOUN1
super-category DETS NSUBJ VERBS VPTD DETS NOUNS
frame 138 330 505 614 138 261
LEXICAL INSERTION
138 DETS
[N3[C3=[C0{DETS}C0]=C3][N0{NNEI}N0]N3]
[N3[C3=[C0|the:|C0]=C3][N0{NNEI}N0]N3]
330 NSUBJ
[V3[N3[N0{NSUBJ}N0]N3][C3[C0[V3=[V0{MODAL}V0]=V3][C0({FUT})C0]C0]C3]
[V0V0]V3]
[V3[N3[N0|student:|N0]N3][C3[C0[V3=[V0{MODAL}V0]=V3][C0({FUT})
C0]C0]C3][V0V0]V3]
505 VERBS
[V3[V0{VERBS}V0]V3]
[V3[V0|will:|V0]V3]
614 VPTD
[V3[V1[V0V0][N3[N0{NCOMP}N0]N3]V1]V3]
[V3[V1[V0|read:|V0][N3[N0{NCOMP}N0]N3]V1]V3]
138 DETS
[N3[C3=[C0{DETS}C0]=C3][N0{NNEI}N0]N3]
[N3[C3=[C0|the:|C0]=C3][N0{NNEI}N0]N3]
261 NOUNS
[N3[N0{NOUNS}N0]N3]
[N3[N0|book:|N0]N3]
393

394
MERGING FRAMES
Merge 138 & 330
138 [N3[C3=[C0|the:|C0]=C3][N0{NNEI}N0]N3]
330 [V3[N3[N0|student:|N0]N3][C3[C0[V3=[V0{MODAL}V0]=V3][C0({FUT})
C0]C0]C3][V0V0]V3]
Result:
[V3[N3[C3=[C0|the:|C0]=C3][N0|student:|N0]N3][C3[C0[V3=[V0{
MODAL}V0]=V3][C0({FUT})C0]C0]C3][V0V0]V3]
Above merged with 505
505 [V3[V0|will:|V0]V3]
Result:
[V3[N3[C3=[C0|the:|C0]=C3][N0|student:|N0]N3][C3[C0[V3=[V0|
will:|V0]=V3][C0({FUT})C0]C0]C3][V0{VERBS}V0]V3]
Above merged with 614
614 [V3[V1[V0|read:|V0][N3[N0{NCOMP}N0]N3]V1]V3]
Result:
[V3[N3[C3=[C0|the:|C0]=C3][N0|student:|N0]N3][C3[C0[V3=[V0|
will:|V0]=V3][C0({FUT})C0]C0]C3][V1[V0|read:|V0][N3[N0{
NCOMP}N0]N3]V1]V3]
Above merged with 138
138 [N3[C3=[C0|the:|C0]=C3][N0{NNEI}N0]N3]
Result:
[V3[N3[C3=[C0|the:|C0]=C3][N0|student:|N0]N3][C3[C0[V3=[V0|
will:|V0]=V3][C0({FUT})C0]C0]C3][V1[V0|read:|V0][N3[C3[
C0|the:|C0]C3][N0{NNEI}N0]N3]V1]V3]
Above merged with 261
261 [N3[N0|book:|N0]N3]
Output:
[V3[N3[C3[C0|the:|C0]C3][N0|student:|N0]N3][C3[C0[V3[V0|will:
|V0]V3][C0|FUT:({FUT})|C0]C0]C3][V1[V0|read:|V0][N3[C3[
C0|the:|C0]C3][N0|book:|N0]N3]V1]V3]

Total Time Elapsed in Seconds: 19
Parse Time Elapsed in Seconds: 8
PARSE SUMMARY
Date: Friday Nov 24, 2000 11:05:58 AM EST
Input: the student will read the book
UFO (User Friendly Output):
The clause “the student will read the book” is a statement.
The phrase “the book” is a noun phrase and the direct object of the verb “read.”
The phrase “the student” is a noun phrase and the subject of the verb “read.”
The word “book” is the head of the noun phrase “the book.”
The word “book” is a singular count noun.
The word “student” is the head of the noun phrase “the student.”
The word “student” is a singular count noun.
The word “the” is a determiner specifying the noun “book.”
The word “the” is a determiner specifying the noun “student.”
The word “will” is a modal auxiliary specifying the meaning of the verb “read.”
[V3[N3[C3[C0|the:|C0]C3][N0|student:|N0]N3][C3[C0[V3[V0|will:|V0]V3][C0|TNS:({FUT})|C0]C0]C3][V1[V0|read:|V0][N3[C3[C0|the:|
C0]C3][N0|book:|N0]N3]V1]V3]
138 330 505 614 138 261
395

396

REFERENCES
Abney, Steven. 1987.
The English Noun Phrase in Its Sentential Aspects. Doctoral dissertation. Cambridge, Mass:
MIT Press.
Allan, Robin, Philip Holmes, and Tom Lundskær–Nielsen. 1995.
Danish: A Comprehensive Grammar. London: Routledge.
Allen, J. H., and James B. Greenough. 1931.
New Latin Grammar for Schools and Colleges. Boston: Ginn and Company.
Anderson, John M. 1971.
The Grammar of Case: Towards a Localistic Theory. Cambridge: Cambridge University Press.
Anderson, John M. 1977.
On Case Grammar. London: Croom Helm.
Anderson, John R. 1985.
Cognitive Psychology and Its Implications, 2nd ed. New York: W. H. Freeman and Co.
Babcock, Clinton L. 1901
A Study in Case Rivalry. Ithaca, New York: Andrus & Church.
Beekes, Robert S.P. 1995.
Comparative Indo–European Linguistics. Philadelphia: John Benjamins.
Belletti, Adriana and Ur Shlonsky. 1995.
“The Order of Verbal Complements: A Comparative Study.” Natural Language and Linguistic
Theory, 13, 3, 489–526.
Binkert, Peter J. 1984.
Generative Grammar without Transformations. Berlin: Mouton.
Binkert, Peter J. 1989.
“Syntax, Short Term Memory and Parsing.” Paper presented at the Annual Meeting of the
Linguistic Society of America, Washington, D.C., December, 1989.
Binkert, Peter J. 1996
The Cognitive Foundations of Language: Semantic and Syntactic Categories in Universal
Grammar, ms. Milford: Langtech.
Blake, Barry. 1994.
Case. Cambridge: Cambridge University Press.
Bolinger, Dwight. 1972.
Degree Words. The Hague: Mouton.
Borras, F.M. and R.F. Christian. 1971.
Russian Syntax. London: Oxford University Press.
Bower, G. H, ed. 1970.
Psychology of language and motivation (Vol. 4). New York: Academic Press.
Bowerman, Melissa. 1973.
Early Syntactic Development. Cambridge: Cambridge University Press.
Bowers, John S. 1981.
The Theory of Grammatical Relations. Ithaca: Cornell University Press.

398
Bresnan, Joan. 1973.
“The Syntax of the Comparative Clause Construction in English.” Linguistic Inquiry
4.3.275–344.
Bresnan, Joan. 1982.
The Mental Representation of Grammatical Relations. Cambridge, MA: MIT Press.
Brown, Roger. 1973.
A First Language / The Early Stages. Cambridge, MA: Harvard.
Buck, Carl Darling. 1933.
Comparative Grammar of Greek and Latin. Chicago: University of Chicago Press.
Butt, John and Carmen Benjamin. 1988.
A New Reference Grammar of Modern Spanish, 2nd ed. Lincolnwood, Illinois: NTC.
Caplan, David. 1972.
“Clause boundaries and recognition of latencies for words in sentences. Perception and
Psychophysics, 12, 73-76.
Carnie, Andrew. 2012
Syntax A Generative Introduction, 3rd ed. Malden, MA: Blackwell.
Cattell, Ray. 1984.
Syntax and Semantics, Volume 17, Composite Predicates in English. North Ryde, New South
Wales: Academic Press Australia
Chomsky, Noam. 1957.
Syntactic Structures. The Hague: Mouton.
Chomsky, Noam. 1965.
Aspects of the Theory of Syntax. Cambridge, MA: MIT Press.
Chomsky, Noam. 1995
The Minimalist Program. Cambridge, MA: MIT Press.
Chomsky, Noam and Morris Halle. 1968.
The Sound Pattern of English. New York: Harper & Row.
Chomsky, Noam and Howard Lasnik. 1977.
“Filters and Control.” Linguistic Inquiry, 8, 425–504.
Culicover, Peter W., Thomas Wasow, and Adrian Akmajian, eds. 1977.
Formal Syntax. New York: Academic Press.
Curme, George O. 1964.
A Grammar of the German Language. New York: Frederick Ungar.
Deepadung, Sujaritlak. 1989.
The Noun Phrase in Thai in the Residential Grammar Framework. Doctoral dissertation.
Lawrence: University of Kansas.
Dempster, F. N. 1978.
“Memory span and short–term memory capacity: A developmental study.” Journal of
Experimental Child Psychology, 26, 419–431.
Dempster, F. N. 1981.
“Memory span sources of individual and developmental differences.” Psychological Bulletin,
89(1), 63–100.

Dempster, F. N. 1985.
“Short-term memory development.” Basic processes in memory development, ed. by C. J.
Brainerd and M. Pressley, 209–248. New York: Springer.
Deny, Jean. 1971.
Grammaire de la Langue Turque. Wiesbaden: Dr. Martin Sändig oHG.
Durrell, Martin. 1983.
Hammer’s German Grammar an Usage. Lincolnwood, Illinois: NTC Publishing Group.
Eliot, Charles Norton Edgecumbe. 1890.
A Finnish Grammar. Oxford: Clarendon Press.
Entwistle, W. J. and W. A. Morrison. 1949.
Russian and the Slavonic Languages. London: Faber and Faber.
Fillmore, Charles. 1968.
“The Case for Case.” Universals in Linguistic Theory, ed. by Emmon Bach and Robert T. Harms,
1–88. New York: Holt, Rinehart and Winston.
Fillmore, Charles. 1969.
“Toward a Modern Theory of Case.” Modern Studies in English, ed by David Reibel and Sanford
Schane, 361–375. Englewood Cliffs, N. J.: Prentice–Hall.
Fletcher, Paul and Michael Garman. 1986.
Language Acquisition: Studies in First Language Development, 2nd Ed. Cambridge: Cambridge
University Press.
Gildersleeve, B. L. and G. Lodge. 1960.
Latin Grammar. New York: Macmillan & Co.
Givón, T. 1984.
Syntax. Volume One. Amsterdam: John Benjamins Publishing Company.
Goodwin, William W. 1963.
A Greek Grammar. London: Macmillan & Co.
Greenberg, Joseph H. 1963.
Universals of Language, 2nd ed. Cambridge, MA: MIT Press.
Greenberg, Joseph H., ed. 1978.
Universals of Human Language, 4 Vols. Stanford: Stanford University Press.
Grimes, Barbara F., ed. 1992.
Ethnologue, Languages of the World. Dallas, TX: Summer Institute of Linguistics.
Gruber, Jeffrey. 1965.
Studies in Lexical Relations. Doctoral dissertation. Cambridge, MA: MIT Press.
Gruber, Jeffrey. 1976.
Lexical Structures in Syntax and Semantics. New York: North–Holland.
Haegeman, Liliana. 1994.
Introduction to Government & Binding Theory, 2nd ed. Oxford: Blackwell.
Hakes, D.T. 1980.
The development of metalinguistic abilities in children. New York: Springer.
Hale, Kenneth and Samuel J. Keyser, eds. 1993.
The View from Building 20. Cambridge, Mass.: MIT Press.
399

400
Halle, Morris, Joan Bresnan, and George Miller, eds. 1978.
Linguistic Theory and Psychological Reality. Cambridge, MA: MIT Press.
Hawkins, John A. 1983
Word Order Universals. New York: Academic Press.
Hess, T. M., and Radke, R. C. 1981.
“Processing and memory factors in children’s reading comprehension skill.” Child Development,
52, 478–488.
Hills, Robert A. 1990.
“A Syntactic Sketch of Ayutla Mixtec.” Studies in the Syntax of Mixtecan Languages, Vol. 2, ed.
by C. Henry Bradley and Barbars E. Hollenbach, 1–260. Dallas: Summer Institute of Linguistics.
Holmes, Philip and Ian Hinchliffe. 1994.
Swedish: A Comprehensive Grammar. London: Routledge.
Jackendoff, Ray. 1972.
Semantic Interpretation in Generative Grammar. Cambridge, Mass.: MIT.
Jackendoff, Ray. 1983.
Semantics and Cognition. Cambridge, Mass: MIT Press.
Jespersen, Otto. 1961.
A Modern English Grammar on Historical Principles. London: George Allen & Unwin.
Karlsson, Fred. 1999.
Finnish: An Essential Grammar. London: Routledge.
King, Gareth. 1993.
Modern welsh, A Comprehensive Grammar. London: Routledge.
Kuno, Susumu. 1983.
The Structure of the Japanese Language. Cambridge, Mass.: MIT Press.
Lasnik, Howard. 2000.
Syntactic Structures Revisited, Contemporary Lectures in Classic Transformational Theory.
Cambridge, MA: MIT Press.
Lenneberg, Eric H. 1967.
Biological Foundations of Language. New York: John Wiley & Sons.
Levin, Beth. 1993.
English Verbs Classes and Alternations. Chicago: University of Chicago Press.
Lewis, G.L. 1967.
Turkish Grammar. Oxford: Oxford University Press.
McGregor, R.S. 1995.
Outline of Hindi Grammar. Oxford: Oxford University Press.
Miller, George A. and Philip N. Johnson–Laird. 1976.
Language and Perception. Cambridge, Mass.: Harvard.
Norman, Jerry. 1988
Chinese. New York: Cambridge University Press.
Oinas, Felix J. 1966.
Basic Course in Estonian. Berlin: Mouton.
Olli, John B. 1958.
Fundamentals of Finnish Grammar. New York: Northland Press.

Osumi, Midori. 1995.
Tinrin Grammar. Honolulu: University of Hawaii Press
Parker, Gary John. 1969
Ayacucho Quechua Grammar and Dictionary. Berlin: Mouton.
Pollock, Jean–Yves. 1989.
“Verb Movement, Universal Grammar, and the Structure of IP.” Linguistic Inquiry, 20, 3,
365–424.
Quirk, R., Sidney Greenbaum, Geoffrey Leech, and Jan Svartvik. 1985.
A Comprehensive Grammar of the English Language. New York: Longman.
Rounds, Carol H. 2001
Hungarian: An Essential Grammar, 2nd Ed. London: Routledge
Ruhlen, Merritt. 1987
A Guide to the World’s Languages. Stanford, CA : Stanford University Press,
Schachter, Paul and Fe T. Otones. 1972.
Tagalog Reference Grammar. Berkeley: University of California Press.
Schank, Roger and Robert Abelson. 1977.
Scripts, Plans, Goals, and Understanding. Hillsdale, N.J.: Lawrence Erlbaum.
Skinner, B. F. 1957.
Verbal Behavior. New York: Appleton-Century-Crofts.
Smyth, Herbert Weir. 1956.
Greek Grammar. Cambridge: Harvard.
Thorne, David A. 1993.
A Comprehensive Welsh Grammar. Oxford: Blackwell.
Tuladhar, Jyoti. 1985.
Constituency and Negation in Newari. Doctoral dissertation. Georgetown: Georgetown
University.
Whitney, William Dwight. 1955.
Sanskrit Grammar. Cambridge: Harvard.
Williams, Edwin. 1994.
Thematic Structure in Syntax. Cambridge, Mass.: MIT Press.
Woodcock, E.C. 1959.
A New Latin Syntax. Cambridge, Mass.: Harvard University Press.
Yasutake, Nobuko. 1995.
An Analysis of Polysemous Meanings for the Japanese Particle NI. Master’s Thesis. Rochester,
MI: Oakland University.
401

402

SAMPLE TEST ONE
LIN 180: LINGUISTIC ANALYSIS
PRINT NAME: _________________________________________
STUDENT NUMBER: ____________________
POINT DISTRIBUTION:
MULTIPLE CHOICE: 50 (2 points each)
TOTAL 100
DIRECTIONS (10 POINTS WILL BE SUBTRACTED FROM YOUR GRADE IF YOU DO
NOT FILL OUT THE SCANTRON CORRECTLY):
1. Please mark all answers to the questions on your scantron answer sheet with a #2 pencil.
2. Write your ID Number in the blank boxes starting on the lefthand side and blacken the
appropriate circles directly under the numbers. Your ID is your Grizzly ID without the “G”;
thus, everyone should have only 8 (EIGHT) digits.
3. Put your name on the answer sheet in the blank boxes and blacken the appropriate circles
directly under the letters, again starting on the lefthand side with your last name, followed by
a space, and then your first name. Omit your middle initial.
4. Indicate which version of the test you have. Blacken Test Form A, for Version One; Blacken
Test Form B, for Version Two
5. Hand in your answer sheet and this test.
Professor Binkert

404
1. Evidence for cerebral lateralization comes from studies of
a. laryngectomy.
b. hemispherectomy.
c. acquired deafness.
d. congenital deafness.
2. Which of the following is INCORRECT?
a. Deaf children exposed to sign language acquire sign language with a strategy that
parallels the way hearing adults learn a second language.
b. Children who are exposed to more than one language during the critical period will
have psychological problems when they grow up.
c. Chimpanzees who have been taught ASL (American Sign Language) generally can
attain the linguistic competence of a three-year-old child.
d. All of the above are incorrect.
3. Which of the following is typical of word processors?
a. They are controlled by arbitrary commands.
b. They perform textual substitutions in terms of random units.
c. They allow deletion of any arbitrary sequence of text.
d. All of the above are typical of word processors.
4. Which of the following contains typical examples of open words?
a. [dgi], [dædi], [hrsi]
b. [dgi], [ðæt], [kray]
c. [sm], [bli], [lgan]
d. [ðæt], [ðiz], [ðoz]
5. The idea that humans are born with very little in the way of innate mechanisms for acquiring
language is typical of
a. rationalism.
b. empiricism.
c. behaviorism.
d. both b and c.

6. Which of the following is of greatest relevance in children’s differentiation of syntactic
categories (parts of speech like noun, verb, etc.)?
a. The order of the words in the sentences the children hear.
b. The size of the vocabulary of the children’s parents (or caregivers).
c. The level of cognitive maturation the children have attained.
d. The amount of time the children’s parents spend talking to them.
7. A major feature of all animal communication systems is
a. extensive paraphrase.
b. imprecision in meaning.
c. strong limbic associations.
d. all of the above.
8. An example of an evolutionary change in the structure of the human language apparatus which
had potentially very disadvantageous side effects is
a. the development of the angular gyrus.
b. the lowering of the larynx.
c. the development of the alveolar ridge.
d. the separation of the nasal and oral cavities.
9. Which of the following is an example of overgeneralization?
a. [bri] for [brt].
b. [rod] for [ru]
c. [wnt] for [kem]
d. [mn] for [mænz]
10. The theory that human language is based, at least in part, on GENERAL neurological abilities
is supported by the fact that
a. removal of the left hemisphere produces immediate aphasia in over 95% of adult
humans today.
b. the acquisition of language parallels the development of the vocal and the auditory
apparatus.
c. the processing of visual data is under many of the same restrictions as the processing of
acoustic data.
d. sustained thought without language is impossible.
405

406
11. A major defect of quantitative continuity theories is that
a. they ignore phylogenetic relationships.
b. they ignore ontogenetic relationships.
c. they link linguistic competence with intelligence.
d. they link linguistic performance with cognitive skills.
12. The innateness hypothesis claims that
a. children are born knowing the regular rules of a language.
b. people of different ethnic backgrounds will grow up speaking different languages.
c. the ability to acquire a human language is species specific.
d. people are usually smarter than monkeys.
13. Which of the following sentences would be of particular interest to prescriptive grammarians?
a. You are never angry for the reason you think.
b. The way to get everything in life that you want is to see to it that everyone else gets
everything in life that they want.
c. If you are constantly being mistreated, then you are cooperating with the treatment.
d. Teach only love for that is what you are.
14. The babbling stage of language acquisition
a. occurs even in congenitally blind children.
b. occurs only in congenitally deaf children.
c. never occurs in hearing children born to non-speaking deaf parents.
d. occurs in all children.
15. Which of the following is NOT an example of a substantive universal?
a. [±ANTERIOR]
b. [±AURAL]
c. [±PREHEAD]
d. [±DISJUNCTIVE]

16. The manner in which speech and motor milestones emerge in both normal and intellectually
disabled children can be used to support the hypothesis that
a. there is a critical period for native language acquisition.
b. native language development is linked to general development.
c. both a and b.
d. neither a nor b.
17. The claim that human language evolved out of animal communication systems is falsified by
the fact that
a. some animals that communicate are deaf.
b. it is possible to chart a progressive development of linguistic abilities from the most
primitive forms of life to humans.
c. most animal communication systems lack a one-to-one correspondence between sound
and meaning.
d. the essential characteristics of human language are absent in animal communication
systems.
18. In grammatical characterization, linguists attempt to account
a. for the linguistic competence of native speakers.
b. for the linguistic performance of native speakers.
c. for native language acquisition in children.
d. for the use of language by normal adults.
19. The critical period for language acquisition is
a. from about two years old to about the time the child begins first grade.
b. from about the time the child begins first grade to the onset of puberty.
c. from birth to about two years old.
d. from about two years old to the onset of puberty.
20. Which of the following constructional types is the easiest to process?
a. Multiple branching.
b. Nested.
c. Self-embedded.
d. Left branching.
407

408
21. Which of the following is a sentence that is UNGRAMMATICAL but ACCEPTABLE?
a. They carried all the books which were shelved in alphabetical order on successive
shelves of six different bookcases made of Indian mahogany into their bedroom.
b. They carried into their bedroom all the books which were shelved in alphabetical order
on successive shelves of six different bookcases made of Indian mahogany.
c. They carried into their bedroom all the books.
d. They carried all the books into their bedroom.
22. The parts of the brain devoted to language are located in
a. the left hemisphere for most people.
b. the left hemisphere for most people except those who speak languages like Japanese
where the writing system is not phonetically based.
c. the left hemisphere for most right-handed people and the right hemisphere for most lefthanded
people.
d. the right hemisphere for most people.
23. Which of the following statements about UG is CORRECT?
a. UG is activated when cerebral lateralization is complete.
b. UG is the set of all possible regular rules existing in human languages.
c. UG is a reflection of the human language apparatus.
d. All of the above.
24. Which of the following contains a construction that is multiple branching?
a. John said that Bill said that Harry said that Frank said nothing.
b. Unfortunately, it upsets some men if their sons cry when they get hurt.
c. John’s mother’s sister’s husband disappeared.
d. None of the above.
25. Which of the following is the correct phonetic transcription of the word thumbed?
a. [md]
b. [ðmd]
c. [md]
d. h
[t md]

26. Which of the following is the correct phonetic transcription of the word ice cubes?
a. [askyubs]
b. [asbz]
c. [ayskybz]
d. [ayskyubz]
27. Which of the following is the correct phonetic transcription of the word claws?
a. h [k lz]
b. h [k læs]
c. h [k lwz]
d. [clawz]
28. Which of the following is the correct phonetic transcription of the word charges?
a. [arz]
b. [argz]
c. [sarjs]
d. [carjz]
29. Which of the following is the correct phonetic transcription of the word sketched?
a.
h
[sk št]
b. [skd]
c. [skesd]
d. [skt]
30. Which of the following is used by linguists to justify syntactic constituents (units).
a. Induction.
b. Reference.
c. Citation.
d. Deduction.
31. The study of the meaning is called
a. ballistics.
b. semantics.
c. pragmatics.
d. phonetics.
409

410
32. When linguists attempt to reach explanatory adequacy,
a. they attempt to determine what makes sentences ungrammatical.
b. they attempt to characterize the structure of grammatical sentences.
c. they attempt to base their descriptions on human capacities.
d. they do all of the above.
33. Which of the following pairs of sounds are distinguished by the feature opposition
[+VOICED]/[–VOICED]?
a. [g] and [k].
b. [g] and [].
c. [g] and [b].
d. [b] and [d].
34. The fact that native speakers are able to produce the appropriate negative for made-up English
adjectives like [prms] demonstrates that
a. the rules for forming negatives are innate.
b. native speakers acquire rules.
c. the rule for prefixing [m-] to adjectives is innate.
d. the rule for prefixing [m-] to adjectives is a substantive universal.
35. One objective of linguistic theory is
a. to describe the characteristics of human languages.
b. to irritate undergraduates.
c. to reform English spelling rules.
d. to do all of the above.
36. Which of the following is CORRECT?
a. Latin is a more perfect language than English because all of the words in Latin are of
Indo-European origin.
b. There are about 100 trillion possible sentences in English.
c. One of the major reasons it is so difficult to become a native speaker of English is
because English spelling has so many exceptions to rules.
d. English is an SVO language.

37. Which of the following is most likely to act like a limbic stimulus.
a. A bell.
b. A flashlight.
c. An electric screw driver.
d. A poisonous snake.
38. Which of the following is ALWAYS TRUE of empty categories.
a. They block contraction.
b. They occur initially in sentences.
c. They occur finally in sentences.
d. They occur anywhere in sentences.
39. Which of the following is a sentence that is GRAMMATICAL but UNACCEPTABLE?
a. They carried all the books which were shelved in alphabetical order on successive
shelves of six different bookcases made of Indian mahogany into their bedroom.
b. They carried into their bedroom all the books which were shelved in alphabetical order
on successive shelves of six different bookcases made of Indian mahogany.
c. They carried into their bedroom all the books.
d. They carried all the books into their bedroom.
40. In phrase structure rules, parentheses indicate
a. alternative elements.
b. optional elements.
c. required elements.
d. random elements.
41. Which of the following is an example of a formal linguistic universal?
a. If a language contains verbs, then it must contain adverbs.
b. All languages must have a distinction between voiced and voiceless sounds.
c. The rules of language are structure dependent.
d. All of the above.
42. If a three-year-old boy acquiring English says [ay hæft mek wiwi], then
a. he has never heard anyone say [ay hæv tu mek wiwi].
b. he has mastered the rules of contraction of [hæv] and [tu] to [hæft].
c. he has not learned that [hæv] and [tu] are separate words.
d. none of the above is necessarily true.
411

412
43. Pivot words are
a. high in number and low in frequency.
b. low in number and high in frequency.
c. low in number and low in frequency.
d. high in number and high in frequency.
44. Which of the following is a type of error that is likely to occur in the speech of a native
speaker of English?
a. The speaker says [nu ædn yuzd bks] instead of [nu ænd yuzd bks].
b. The speaker says [yu ænd nuzd bks] instead of [nu ænd yuzd bks].
c. The speaker says [nu yuzd ænd bks] instead of [nu ænd yuzd bks].
d. The speaker says [nu ænd bks yuzd] instead of [nu ænd yuzd bks].
45. If a three-year-old boy says [ay dont gat nn] instead of [ay dont hæv ni], then
a. his parents or caregivers are probably lower class.
b. the people he spends the most time with probably say things like [ay dont gat nn].
c. he probably has below normal intelligence.
d. both a and b.
46. If a three-year-old girl says [p yrz] after being continuously annoyed and frustrated by a
three-year-old boy, then
a. her parents or caregivers probably do not know the difference between acceptable and
unacceptable sentences.
b. her parents have not yet taught her the difference between acceptable and unacceptable
sentences.
c. she has probably heard someone say [p yrz] in a similar context.
d. all of the above.
47. Which of the following is an example of inductive thinking?
a. A travel guide for Germany stipulates that Germans never touch the people they are
speaking to; therefore, during your visit to Germany you never touch any of the Germans
you speak to.
b. Most Italians one has observed use their hands a great deal when they talk; therefore,
Giuseppe, an Italian one has not yet met will probably use his hands when speaking.
c. Both of the above.
d. Neither of the above.

48. Which of the following is typical of children learning their native language.
a. Their production of language is more advanced then their comprehension of language.
b. They internalize rules on the basis of what their parents have told them is correct.
c. They overgeneralize rules.
d. All of the above are typical.
49. An important aspect of human language is its creative aspect. Which of the following
statements best reflects that aspect?
a. Speakers are able to produce and understand sentences they have never heard before.
b. Speakers are able to distinguish grammatical sentences from ungrammatical sentences.
c. Speakers are able to distinguish acceptable sentences from unacceptable sentences.
d. Speakers are able to discuss events they have not personally witnessed, e.g., the
invention of the toilet bowl.
50. There is evidence that the acquisition of songs by some birds like chaffinches also must occur
during a critical period; therefore,
a. if chaffinches are not exposed to the songs of their species during that period, they will
not acquire songs.
b. the acquisition of language in humans evolved out of the acquisition of songs in birds.
c. there is a good chance that cerebral lateralization also occurs in some birds.
d. all of the above.
413

SAMPLE TEST TWO
LIN 180: LINGUISTIC ANALYSIS
PRINT NAME: _________________________________________
STUDENT NUMBER: ____________________
POINT DISTRIBUTION:
MULTIPLE CHOICE: 50 (2 points each)
TOTAL 100
DIRECTIONS (10 POINTS WILL BE SUBTRACTED FROM YOUR GRADE IF YOU DO
NOT FILL OUT THE SCANTRON CORRECTLY):
1. Please mark all answers to the questions on your scantron answer sheet with a #2 pencil.
2. Write your ID Number in the blank boxes starting on the lefthand side and blacken the
appropriate circles directly under the numbers. Your ID is your Grizzly ID without the “G”;
thus, everyone should have only 8 (EIGHT) digits.
3. Put your name on the answer sheet in the blank boxes and blacken the appropriate circles
directly under the letters, again starting on the lefthand side with your last name, followed by
a space, and then your first name. Omit your middle initial.
4. Indicate which version of the test you have. Blacken Test Form A, for Version One; Blacken
Test Form B, for Version Two
5. Hand in your answer sheet and this test.
Professor Binkert

1. Which of the following is the correct phonetic transcription of the noun phonetics?
a. [fnD´ks]
h b. [p óntks]
c. [fóntks]
d. [fn´Dks]
2. Which of the following is the correct phonetic transcription of the word psychic?
h
a. [sáyk k°]
h
b. [séyk k°]
h
c. [sák k°]
d. [šáyk°k ]
h
3. Which of the following is the correct phonetic transcription of the word Ypsilanti?
h
a. [p°slæ´ nt i]
h
b. [p°salæ´ nt y]
h
c. [´p°slnt i]
h h
d. [p slæ´ nt i]
4. Which of the following is the correct phonetic transcription of the word chimpanzees?
h
a. [èmp nzíz]
h
b. [šímp ænzz]
h
c. [èmp ænzíz]
h
d. [èmp ánziz]
415

416
5. Which of the following is the correct phonetic transcription of the word Caesar’s?
a. [sízrz]
b. [sízrs]
c. [síyzars]
d. [s´zrz]
6. Which of the following pairs of words is a minimal pair?
a. polite and police
b. they're and there
c. journal and colonel
d. wring and ring
7. Which of the following pairs of words is a minimal pair?
a. racer and razor
b. daughter and slaughter
c. bomb and tomb
d. cough and rough
8. Which of the following is the correct phonological representation of the word scripts?
a. /skrb3t3z/
b. /skrpt3s/
c. /skrbt3z/
d. /skrp3t3s/

9. Which of the following is the correct phonological representation of [lvmrgan]?
a. /all3of3them3are3gone/
b. /l3v3ðm3ar3gan/
c. /l3v3ðm3hæv3gan/
d. /l3v3m3ar3gan/
10. In the articulation of [m],
a. the velum is lowered.
b. the lips are rounded.
c. the velum is raised.
d. the tongue is lowered.
11. Which of the following sounds is represented by the feature cluster [–CONSONANT,
+VOCALIC, –HIGH, +LOW, +BACK, +TENSE]?
a. [a]
b. []
c. [æ]
d. []
12. Which of the following sounds are represented by the feature cluster [+CONSONANT,
–VOCALIC, +ANTERIOR, +CORONAL]?
a. [š] and [ž]
b. [t] and [d]
c. [è] and []
d. [k] and [g]
417

418
13. Which of the following is a correct feature representation of the sound []?
a. [-CONSONANT, +VOCALIC, +HIGH, –LOW, +BACK, –ROUND, +TENSE]
b. [–CONSONANT, +VOCALIC, +HIGH, –LOW, +BACK, +ROUND, –TENSE]
c. [–CONSONANT, +VOCALIC, –HIGH, –LOW, +BACK, +ROUND, –TENSE]
d. [–CONSONANT, +VOCALIC, –HIGH, +LOW, +BACK, +ROUND, +TENSE]
14. Which of the following is a correct feature representation of the sound [p]?
a. [+CONSONANT, –VOCALIC, +ANTERIOR, +CORONAL, –VOICED, –NASAL]
b. [+CONSONANT, –VOCALIC, +ANTERIOR, –CORONAL, –VOICED, –NASAL]
c. [+CONSONANT, –VOCALIC, +ANTERIOR, –CORONAL, +VOICED, –NASAL]
d. [+CONSONANT, –VOCALIC, –ANTERIOR, +CORONAL, +VOICED, –NASAL]
15. Which of the following feature clusters is common to all sounds that are [+ROUND]?
a. [–CONSONANT, +VOCALIC, +BACK, +LOW]
b. [–CONSONANT, +VOCALIC, +BACK, –HIGH]
c. [–CONSONANT, +VOCALIC, +BACK, –LOW]
d. [–CONSONANT, +VOCALIC, –BACK, +LOW]
16. Which of the following groups of sounds forms an exhaustive natural class?
a. [e], [], [o], []
b. [i], [], [u], []
c. [u], [], [o], []
d. [e], [], [u], []

17. Which of the following groups of sounds forms an exhaustive natural class?
a. [p], [t], [k], []
b. [f], [], [s], [ž]
c. [m], [n], [], []
d. [m], [w], [y], [h]
18. Which of the following groups of sounds forms an exhaustive natural class?
a. [w], [y], [l], [³]
b. [w], [y], [l], [r]
c. [w], [y], [], [h]
d. [w], [y], [h], [r]
19. Which of the following sets of features correctly represents the natural class of sounds [i], [],
[u], []?
a. [–CONSONANT, +VOCALIC, +HIGH]
b. [–CONSONANT, +VOCALIC, –HIGH]
c. [–CONSONANT, +VOCALIC, +LOW]
d. [–CONSONANT, +VOCALIC, –LOW]
20. Which of the following sets of features correctly represents the natural class of sounds [a], [e],
[i], [o], [u]?
a. [–CONSONANT, +VOCALIC, +HIGH]
b. [–CONSONANT, +VOCALIC, –TENSE]
c. [–CONSONANT, +VOCALIC, –LOW]
d. none of the above
419

420
21. Consider the following rule:
n
[–TENSE] [+STRESS] / _____ C 2 #
This rule will change
a. /absin/ to [absín].
b. /rivolt/ to [rivólt].
c. /nsst/ to [ns´st].
d. all of the above.
22. Consider the following rule:
This rule will change
a. the nature of the vowel of /sem/.
b. the nature of the vowel of /mes/.
c. /me/ to [men].
d. all of the above.
23. Which of the following rules will produce voicing assimilation?
a. [áVOICED] [–âVOICED] / ___ [–áVOICED]
b. [áVOICED] [–áVOICED] / ___ [áVOICED]
c. [áVOICED] [–áVOICED] / ___ [–áVOICED]
d. [áVOICED] [–áVOICED] / ___ [âVOICED]

24. Which of the following rules is the standard way of saying palatalize initial alveolar stops
before [y]?
25. Which of the following rules is the standard way of saying unrounded vowels in the next-tothe-last
syllable of a word are rounded, if the last syllable contains a rounded vowel?
n n
a. [–ROUND] [áROUND] / ___ C 0 [áROUND] C 0 #
n n
b. [–ROUND] [+ROUND] / ___ C 0 [+ROUND] C 0 #
n n
c. [–ROUND] [–áROUND] / ___ C 0 [–áROUND] C 0 #
n n
d. [–ROUND] [+ROUND] / ___ C 0 [–ROUND] C 0 #
421

422
26. Which of the following rules is the standard way of saying unrounded, tense vowels in the
next-to-the-last syllable of a word are rounded, if the last syllable contains a rounded vowel?
THE NEXT THREE QUESTIONS (27-29) ARE BASED ON THE FOLLOWING DATA FROM
A DIALECT OF ENGLISH:
[byt] bite [bayd] bide [bay] by
[wyf] wife [wayvz] wives [way] why
[dyk] dike [daym] dime [day] die
[tryp] tripe [trayb] tribe [tray] try
On the basis of these data alone, answer the next three questions.
27. Which of the following is correct?
a. [] and [a] are separate phonemes.
b. [b] and [d] are separate phonemes.
c. [t] and [d] are separate phonemes.
d. [w] and [y] are separate phonemes.

28. Which of the following rules probably occurs in the dialect?
29. Which of the following is correct?
a. The word slight is probably pronounced [slayt].
b. The word sly is probably pronounced [sly].
c. The word slide is probably pronounced [slayd].
d. The word slime is probably pronounced [slym].
423

424
30. The word indescribable has
a. 2 bound morphemes and 2 free morphemes.
b. 3 bound morphemes and 1 free morpheme.
c. 3 affixes and 1 root.
d. 2 affixes and 3 roots.
31. In terms of number and types of morphemes, murderers has a morphological structure that is
similar to
a. bachelors.
b. cylinders.
c. governors.
d. interiors.
32. In terms of number and types of morphemes, irreplaceable has a morphological structure that
is similar to
a. incomprehensible.
b. indescribable.
c. unreasonable.
d. irreproducible.
33. The words phoneme and morpheme are examples of
a. jargon.
b. epithets.
c. folk etymologies.
d. compound nouns.

34. The fact that some speakers of English regularly produce sentences like He don’t know nothing
suggests
a. that they can't understand the rules for forming negative sentences.
b. that they can't understand the rules for number agreement between subjects and verbs.
c. neither a nor b.
d. both a and b.
THE NEXT THREE QUESTIONS (35-37) ARE BASED ON THE FOLLOWING DATA
FROM CLASSICAL LATIN:
SINGULAR PLURAL MEANING
[lîs] [lîtes] strife
[fraws] [frawdes] fraud
[frôns] [frôntes] brow
[frôns] [frôndes] leaf
[dens] [dentes] tooth
[sors] [sortes] lot
35. Which of the following rules captures the morphological changes?
425

426
36. Which of the following is correct?
a. The phonemic form of [dens] is /den3s/.
b. The phonemic form of [dens] is /den3es/.
c. The phonemic form of [dens] is /dent3s/.
d. The phonemic form of [dens] is /dent3es/.
37. Which of the following is correct?
a. The morpheme for 'brow' is /frônts/.
b. The morpheme for 'brow' is /frônd/.
c. The morpheme for 'leaf' is /frônd/.
d. The morpheme for 'leaf'' is /frônt/.
THE NEXT SIX QUESTIONS (38-43) ARE BASED ON THE FOLLOWING DATA FROM A
HYPOTHETICAL LANGUAGE:
[grub] 'cardinal' [grupz] 'cardinals'
[krog] 'oriole' [krokz] 'orioles'
[grud] 'robin' [grutz] 'robins'
[prog] 'sparrow' [prokz] 'sparrows'
[krop] 'starling' [kropz] 'starlings'
[grod] 'swallow' [grotz] 'swallows'
[grut] 'thrush' [grutz] 'thrushes'
[brob] 'wren' [grukz] 'finches'
38. On the basis of these data, one can conclusively say that
a. [g] and [r] are separate phonemes.
b. [u] and [] are separate phonemes.
c. [u] and [o] are separate phonemes.
d. [d] and [z] are separate phonemes.

39. The correct phonetic form for the word wrens is
a. [bropz].
b. [brobz].
c. either a or b given the data.
d. neither a nor b given the data.
40. The correct phonemic form for the word wrens is
a. /brop3z/.
b. /brob3z/.
c. either a or b given the data.
d. neither a nor b given the data.
41. The correct phonetic form for the word finch is
a. [gruk].
b. [grug].
c. either a or b given the data.
d. neither a nor b given the data.
42. The correct phonemic form for the word finches is
a. /gruk3z/.
b. /grug3z/.
c. either a or b given the data.
d. neither a nor b given the data.
427

428
43. Which of the following rules correctly relates the forms in the first column to those in the
second?
44. Which of the following is correct?
a. An alternation in sound that is predictable is not redundant.
b. Allophones are found in minimal pairs.
c. The phonetic transcription represents what a person actually hears.
d. The position of phonemes is predictable.
45. Which of the following is correct about English phonology?
a. The number of morphemes in a word is never equal to the number of syllables.
b. The number of morphemes in a word is never more than the number of syllables.
c. The number of morphemes in a word always equals the number of syllables.
d. The number of morphemes in a word is independent of the number of syllables.

46. Consider the following English data:
[rilet] [rilešn]
[divot] [divošn]
[inved] [invežn]
[krod] [krožn]
In terms of the overall phonological structure of English, the rule that correctly accounts for
these contrasts is
429

430
47. Which of the following feature oppositions are distinctive in Standard American English?
a. [±ANTERIOR] and [±ASPIRATED]
b. [±ASPIRATED] and [±RELEASED]
c. [±VOICED] and [±ANTERIOR]
d. [±VOICED] and [±RELEASED]
48. Which of the following feature oppositions are not distinctive in Standard American English?
a. [±ANTERIOR] and [±ASPIRATED]
b. [±ASPIRATED] and [±RELEASED]
c. [±VOICED] and [±ANTERIOR]
d. [±VOICED] and [±RELEASED]
49. Which of the following are accidental gaps in the English lexicon?
a. [sprak] and [stregn]
b. [vwu] and [flomn]
c. [gno] and [biki]
d. [strayl] and [boèz]
50. Which of the following are nonaccidental gaps in the English lexicon?
a. [sprak] and [stregn]
b. [vwu] and [flomn]
c. [gno] and [biki]
d. [strayl] and [boèz]

SCRATCH PAGE NAME:__________________________________________
431

SAMPLE TEST THREE
LIN 180: LINGUISTIC ANALYSIS
PRINT NAME: _________________________________________
STUDENT NUMBER: ____________________
POINT DISTRIBUTION:
MULTIPLE CHOICE: 50 (2 points each)
TOTAL 100
DIRECTIONS (10 POINTS WILL BE SUBTRACTED FROM YOUR GRADE IF YOU DO
NOT FILL OUT THE SCANTRON CORRECTLY):
1. Please mark all answers to the questions on your scantron answer sheet with a #2 pencil.
2. Write your ID Number in the blank boxes starting on the lefthand side and blacken the
appropriate circles directly under the numbers. Your ID is your Grizzly ID without the “G”;
thus, everyone should have only 8 (EIGHT) digits.
3. Put your name on the answer sheet in the blank boxes and blacken the appropriate circles
directly under the letters, again starting on the lefthand side with your last name, followed by
a space, and then your first name. Omit your middle initial.
4. Indicate which version of the test you have. Blacken Test Form A, for Version One; Blacken
Test Form B, for Version Two
5. Hand in your answer sheet and this test.
Professor Binkert

1. Consider the following sentences:
Sue put on the new hat.
Sue sat on the new hat.
Which of the following can be used to show that these sentences must be assigned different
structural descriptions?
a. Sue put the new hat on.
*Sue sat the new hat on.
b. *The horse which Sue put on was wild.
The horse which Sue sat on was wild.
c. Both a and b can be used.
d. Neither a nor b can be used.
2. Consider the following sentences:
They warned the president of the danger.
They warned the president of the country.
Which of the following can be used to show that these sentences must be assigned different
structural descriptions?
a. *The president of the danger was warned.
The president of the country was warned.
b. The president was warned of the danger.
*The president was warned of the country.
c. The president is the one they warned of the danger.
*The president is the one they warned of the country.
d. All of the above can be used.
433

3. Consider the following sentences:
They gave her dog biscuits.
They have her dog biscuits.
Which of the following can be used to show that these sentences do not have the same range
of structural descriptions?
a. They gave dog biscuits to her.
*They have dog biscuits to her.
b. They gave biscuits to her dog.
*They have biscuits to her dog.
c. Both a and b can be used.
d. Neither a nor b can be used.
4. Which of the choices below correctly abbreviates all and only the following rules:
VP V + NP
VP V + PP
VP V + ADJP
a. VP V + (NP) + (PP) + (ADJP)
b. VP V +
c. VP V +
d. VP V +
434

5. Which of the choices below correctly abbreviates all and only the following rules:
NP DET + ADJ + N
NP DET + N
NP N
a. NP ((DET) + ADJ) + N
b. NP (DET) + (ADJ) + N
c. NP (DET + ADJ) + N
d. NP (DET + (ADJ)) + N
6. Which of the choices below correctly describes the structure of the following adjective
phrases: dependent; totally dependent; dependent on Bill; dependent on Bill for help;
dependent on the people he knows; totally dependent on the people he knows; completely
dependent on his wife for help.
a. AP (ADV) + ADJ +
b. AP (ADV) + ADJ + (PP) + (S)
c. AP (ADV) + ADJ + (PP) + (PP)
d. None of the above.
7. In terms of all that we have discussed about the structure of English, which of the following
categorial descriptions is a possible description of the phrase hit the boy with the bat?
a. V + NP
b. V + NP + PP
c. V + NP + P + NP
d. All of the above are possible.
435

8. In terms of all that we have discussed about the structure of English, which of the following
categorial descriptions is a possible description of the phrase throw away the leftover turkey?
a. V + PP
b. V + PP + NP
c. V + NP
d. both b and c are possible.
9. Which of the following words does not belong with the other three in terms of
subcategorization frames?
a. embrace
b. kiss
c. meet
d. shave
10. Which of the following sentences indicates that the word back can be a verb?
a. He says we should get her back.
b. He says we should get back at her.
c. He says we should back her up.
d. Both band c.
436

11. Consider the following sentences:
The elderly man can’t handle going up and down the stairs.
The elderly man can’t handle the ups and downs of the economy.
On the basis of these data, it seems that
a. both up and down can be nouns.
b both up and down can be verbs.
c. neither a nor b is correct.
d. both a and b are correct.
THE NEXT FIVE QUESTIONS (12 - 16) ARE BASED ON THE FOLLOWING SET OF
PHRASE STRUCTURE RULES:
I. S NP + AUX + VP
II. VP V + (NP) +
III. PP P + (NP)
IV. NP
V. AUX (MOD) + TNS + (NEG)
VI. TNS
VII. AP (ADV) + ADJ
437

12. Which of the following noun phrases indicates that Rule IV is inadequate?
a. the shirt in the closet in the hall
b. the very expensive shirt from Italy
c. the expensive shirt from a small town in the south of Italy
d. the shirt in the hall closet
13. Which of the following sentences indicates that the internal structure of noun phrases has more
levels than the rules allow?
a. If you iron those shirts, I’ll iron these.
b. I’ll iron that shirt if you iron this one.
c. I’ll iron those silk shirts after you iron these.
d. All of the above.
14. Which of the following sentences has a structure that CANNOT be generated with these rules?
a. Mary took the very stained shirt to the laundry.
b. Mary took the shirt stained with ink to the laundry.
c. Mary took the shirt off the hanger.
d. None of the above can be generated.
15. Given the above rules, the best way to generate a sentence like That laundry irons shirts
cheaply would be to add the element ADV to some part of
a. Rule I.
b. Rule II.
c. Rule III.
d. Rule IV.
438

16. Consider the following data:
STATEMENTS QUESTIONS
The housekeeper will iron the shirts. Will the housekeeper iron the shirts?
The housekeeper doesn’t do windows. Doesn’t the housekeeper do windows?
Given the PS-rules on page six, statements and questions such as the above can be related by
which of the following transformational rules?
a. NP + TNS TNS i + NP + [e] i
b. NP + MOD MOD i + NP + [e] i
c. NP + AUX AUX i + NP + [e] i
d. NP + VP VP i + NP + [e] i
THE NEXT SEVEN QUESTIONS (17 - 23) ARE BASED ON THE FOLLOWING DATA
FROM LATIN:
I. Clara regina pulchram fabulam laudat.
The famous queen praises the beautiful story.
II. Reginae claras fabulas laudabant.
The queens praised the famous stories.
III. Clara regina fabulas laudabit.
The famous queen will praise the stories.
IV. Amica regina aegras agricolas curabit.
The friendly queen will care for the sick farmers.
V. Inimica agricola aegram lupam curabat.
The unfriendly farmer cared for the sick wolf.
VI. Inimicae agricolae aegram lupam declinabant.
The unfriendly farmers avoided the sick wolf.
VII. Inimica regina declinabat.
The unfriendly queen turned away.
439

17. Which of the following sets of PS-Rules best describes the above data (in each set, assume that
there is a rule rewriting TNS as PRS, PST, or FUT)?
a. S NP + TNS + VP
VP (NP) + V
NP (ADJ) + N
b. S NP + TNS + VP
VP V + NP
NP ADJ + N
c. S NP + VP
VP (NP) + VERB
VERB V + TNS
NP (ADJ) + N
d. S NP + VP
VP NP + VERB
VERB V + TNS
NP (ADJ) + N
18. According to the data given, which of the following is correct regarding the verb declina-?
a. It is either transitive or intransitive.
b. It is either transitive or copulative.
c. It is either copulative or intransitive.
d. It is always intransitive.
19. Which of the following is correct?
a. Adjectives agree with nouns in case and number.
b. Verbs agree with subjects in case and number.
c. Both a and b are correct.
d. Neither a nor b is correct.
440

20. Which of the following is correct?
a. The morpheme signifying PST is -bi-.
b. The morpheme signifying a plural verb is -nt.
c. The morpheme signifying a singular direct object is -s.
d. Both a and c are correct.
21. Which of the following is a grammatical Latin sentence?
a. Reginae agricolas curabant.
b. Regina agricolas curant.
c. Reginae agricolam curabat.
d. Regina agricolas curabant.
22. Which of the following is a grammatical Latin sentence?
a. Amica agricola fabulas laudabant.
b. Inimica regina lupas declinant.
c. Lupae inimicas reginas curat.
d. Aegra regina aegras lupas curat.
23. If the morpheme meaning ‘assist’ is juva-, and if the morpheme meaning ‘bad’ is mala-, then
which of the following Latin sentences is grammatical?
a. Malae reginae malam lupam juvant.
b. Mala lupa malas agricolas juvant.
c. Malae agricolae inimicam reginam juvat.
d. Lupa mala reginas malas juvat.
441

THE NEXT FOUR QUESTIONS (24 - 27) REFER TO THE FOLLOWING STRUCTURE:
24. If a sentence has the above structure, then
a. its main verb might have a direct object.
b. its subject noun phrase must have an overt determiner.
c. its main verb must be transitive.
d. all of the above are correct.
25. If a sentence has the above structure, then
a. its auxiliary could be PRS.
b. its auxiliary could be FUT.
c. its auxiliary could be IMP.
d. its auxiliary could be CND.
26. If a sentence has the above structure, then
a. its main verb could be give.
b. its main verb could be become.
c. its main verb could be fall.
d. all of the above are possible.
442

27. Which of the following sentences has the above structure?
a. The butler put the coats in the closet.
b. The maid washed up the linens.
c. Good butlers keep dirty clothes in the hamper.
d. Butlers should wash windows with a mop.
THE NEXT EIGHT QUESTIONS (28 - 35) REFER TO THE FOLLOWING STRUCTURE:
28. If a sentence has the above structure, then
a. its subject phrase might contain a modifier.
b. its verb phrase might contain a sentence adverb.
c. its main verb must be intransitive.
d. its direct object phrase must contain an adjectival modifier.
29. If a sentence has the above structure, then
a. housekeepers could be the object of its main verb.
b. good butlers could be the object of its main verb.
c. housekeepers could be its subject.
d. those butlers could be its subject.
443

30. If a sentence has the above structure, then
a. threw could be its main verb.
b. turned could be its main verb.
c. throw could be its main verb.
d. all of the above are possible.
31. If a sentence has the above structure, then
a. its tense might be FUT.
b. its tense might be PST.
c. its tense might be CND.
d. all of the above are possible.
32. If a sentence has the above structure, then which of the following words can be in the
sentence?
a. out
b. not
c. my
d. quickly
33. If a sentence has the above structure, then which of the following morphemes cannot be in the
sentence?
a. PST
b. Ø
c. PRS
d. IMP
444

34. If a sentence has the above structure, then the sentence cannot contain any
a. intransitive prepositions.
b. possessives.
c. adjectives.
d. compound nouns.
35. Which of the following sentences has the above structure?
a. The auto mechanic couldn’t pick the tiny screw up.
b. The disobedient child will put the toy soldiers away.
c. The figure skater can perform the required moves perfectly.
d. The alarm system won’t scare the jewel thief away.
36. Consider the following possible diagram for the structure of English sentences:
One problem with this kind of representation is that it
a. it cannot be reduced to a set of rules.
b. it makes the AUXILIARY a constituent of SENTENCE.
c. it makes the DIRECT OBJECT part of the VERB PHRASE.
d. it fails to account for the structural similarities of SUBJECT and DIRECT OBJECT.
445

37. Which of the following is correct in the TG model of syntax?
a. Every sentence has two structural levels, an abstract underlying structure roughly
equivalent to what is understood, and a real superficial structure roughly equivalent to
what is heard.
b. The order of phrases in the abstract underlying structure can be different from the order
of phrases in the real superficial structure.
c. Superficial structures can contain empty elements like [e].
d. All of the above are correct for TG.
38. Consider the following sentence:
The Spanish teacher resigned.
This sentence exhibits
a. structural ambiguity only.
b. lexical ambiguity only.
c. both structural and lexical ambiguity.
d. neither structural nor lexical ambiguity.
39. Which of the following words does not belong with the other three in terms of syntactic
categorization?
a. alligator
b. dentist
c. calendar
d. pester
446

40. Which of the following words does not belong with the other three in terms of syntactic
categorization?
a. author
b. actor
c. anchor
d. doctor
41. Which of the following sentences has a different syntactic structure from the other three?
a. She approached the bus stop.
b. They removed the bus stop.
c. He hit the bus stop.
d. He heard the bus stop.
42. Which of the following sentences has a different syntactic structure from the other three?
a. She played the piano in the lounge.
b. They gave an award to the actress.
c. She bought a present for the baby.
d. They asked a question of the teacher.
43. Which of the following sentences has a different syntactic structure from the other three?
a. John works to live.
b. John wants to live.
c. John stopped to eat.
d. John walked to exercise.
447

44. Which of the following sentences has a different syntactic structure from the other three?
a. They attributed that painting to Picasso.
b. I despise that painting of Picasso
c. He cherishes the letter from Picasso.
d. She has a book about Picasso.
45. Which of the following labeled brackets correctly represents the structure of the sentence The
jurors considered her as guilty as sin?
a. [ VP V – NP – NP ]
b. [ VP V – NP – NP – PP ]
c. [ VP V – NP – AP ]
d. [ VP V – NP – P – NP ]
46. Which of the following labeled brackets correctly represents the structure of the sentence They
told me she lied?
a. [ VP V – NP – S ]
b. [ VP V – NP ]
c. Both a and b.
d. Neither a nor b.
47. Which of the following sentences can be cited to support the argument that sentence adverbs
are constituents of S whereas manner adverbs are constituents of VP?
a. The passengers will probably listen to the stewardess attentively.
b. The passengers probably will listen to the stewardess attentively.
c. Both a and b.
d. Neither a nor b.
448

48. Which of the following transformations correctly describes the relationship between the
sentences in [1] and, at the same time, accounts for the data in [2]?
[1] The students passed in the tests.
The students passed the tests in.
[2] The cars passed in the tunnel.
*The cars passed the tunnel in.
a. X 1 + V + P + NP + X 2 => 1 + 2 + Ø + 4 + 3 + 5
where VP does not immediately dominate 3 and 4
b. X 1 + V + P + NP + X 2 => 1 + 2 + Ø + 4 + 3 + 5
where PP immediately dominates 3 and 4
c. X 1 + V + P + NP + X 2 => 1 + 2 + Ø + 4 + 3 + 5
where NP immediately dominates 3 and 4
d. X 1 + V + P + NP + X 2 => 1 + 2 + Ø + 4 + 3 + 5
where PP does not immediately dominate 3 and 4
49. Which of the following sentences can be cited to support the argument that there must be an
intermediate level V´ between VP and V?
a. The deer disappeared into the woods when the gun went off.
b. John said he would win, but I didn’t believe him.
c. Both a and b.
d. Neither a nor b.
50. Which of the following sentences can be cited to support the argument that there must be an
intermediate level N´ between NP and N?
a. Give me liberty or give me a salami sandwich.
b. An experienced candidate is preferable to an inexperienced one.
c. Both a and b.
d. Neither a nor b.
449

SAMPLE TEST FOUR
LIN 180: LINGUISTIC ANALYSIS
PRINT NAME: _________________________________________
STUDENT NUMBER: ____________________
POINT DISTRIBUTION:
MULTIPLE CHOICE: 50 (2 points each)
TOTAL 100
DIRECTIONS (10 POINTS WILL BE SUBTRACTED FROM YOUR GRADE IF YOU DO
NOT FILL OUT THE SCANTRON CORRECTLY):
1. Please mark all answers to the questions on your scantron answer sheet with a #2 pencil.
2. Write your ID Number in the blank boxes starting on the lefthand side and blacken the
appropriate circles directly under the numbers. Your ID is your Grizzly ID without the “G”;
thus, everyone should have only 8 (EIGHT) digits.
3. Put your name on the answer sheet in the blank boxes and blacken the appropriate circles
directly under the letters, again starting on the lefthand side with your last name, followed by
a space, and then your first name. Omit your middle initial.
4. Indicate which version of the test you have. Blacken Test Form A, for Version One; Blacken
Test Form B, for Version Two
5. Hand in your answer sheet and this test.
Professor Binkert
450

1. Consider the following sentences:
Marilyn slipped on her new shoes.
Marilyn tripped on her new shoes.
Which of the following can be used to show that these sentences do not have the same range
of structural descriptions?
a. Marilyn slipped her new shoes on.
*Marilyn tripped her new shoes on.
b. Marilyn will let you slip on her new shoes if you ask her nicely.
*?Marilyn will let you trip on her new shoes if you ask her nicely.
c. both a and b can be used.
d. neither a nor b can be used.
2. Consider the following sentences:
Roxanne climbed down the rope to Cyranno.
Roxanne threw down the rope to Cyranno.
Which of the following can be used to show that these sentences must be assigned different
structural descriptions?
a. It was down that rope that Roxanne climbed to Cyranno.
*It was down that rope that Roxanne threw to Cyranno.
b. *Roxanne climbed that rope down to Cyranno.
Roxanne threw that rope down to Cyranno.
c. Both a and b can be used.
d. Neither a nor b can be used.
451

3. Consider the following sentences:
The president is difficult to understand.
The president is unlikely to understand.
Which of the following can be used to show that these sentences must be assigned different
structural descriptions?
a. *The president is difficult to understand the problem.
The president is unlikely to understand the problem.
b. *The president is difficult to appear in public in pajamas.
The president is unlikely to appear in public in pajamas.
c. both a and b can be used.
d. neither a nor b can be used.
4. Which of the choices below correctly abbreviates all and only the following rules:
VP V
VP V + NP
VP V + NP + NP
VP V + NP + AP
VP V + NP + PP
a. VP V + (NP) + ((NP) + (AP) + (PP))
b. VP V + NP +
c. VP V + (NP) +
d. None of the above.
452

5. Which of the choices below correctly abbreviates all and only the following rules:
N3 N3 + C3 + N2
N3 C3 + N2
N3 N2
a. N3 ((N3) + C3) + N2
b. N3 (N3) + (C3) + N2
c. N3 (N3 + C3) + N2
d. N3 (N3 + (C3)) + N2
6. Which of the choices below correctly describes the structure of the following adjective
phrases: dependent; totally dependent; dependent on Richard; dependent on Richard for
support; dependent on the people he knows; totally dependent on the people he knows;
completely dependent on his friends for support.
a. AP (ADV) + ADJ +
b. AP (ADV) + ADJ +
(PP) + (S)
c. AP (ADV) + ADJ + (PP) + (PP)
d. None of the above.
7. In terms of all that we have discussed about the structure of English, which of the following
categorial descriptions is a possible description of the phrase examine the patient with German
measles?
a. V + NP
b. V + NP + PP
c. V + NP + P + NP
d. all of the above are possible.
453

8. In terms of all that we have discussed about the structure of English, which of the following
categorial descriptions is a possible description of the phrase fall down a whole flight of stairs?
a. V + PP
b. V + PP + NP
c. V + NP
d. both b and c are possible.
9. In terms of all that we have discussed about the structure of English, which of the following
categorial descriptions is the correct description of the phrase the reporter’s relentless pursuit
of the truth?
a. N3 + POS + C3 + N0 + of + N3
b. N3 + POS + C3 + N1 + of + N3
c. N3 + POS + C3 + N2
d. N3 + POS + C3 + N3
10. Consider the following sentences:
The does do do doo-doo all over the lawn.
No doe does do more doo-doo than a dodo does.
On the basis of these data, it seems that
a. does can be a noun.
b does can be a modal.
c. does can be an adjective.
d. both a and b are correct.
454

11. Which of the following sentences indicates that the word back can be a verb?
a. Please take these clothes back to the laundry.
b. The backers of the bill voted for it.
c. The doctor examined the bruise on the patient’s back.
d. None of the above.
THE NEXT FIVE QUESTIONS (12 - 16) ARE BASED ON THE FOLLOWING SET OF
PHRASE STRUCTURE RULES:
I. S NP + TP + VP
II. VP V + (NP) +
III. PP P + (NP)
IV. NP
V. TP (MP) + T + (NEGP)
VI. T
VII. AP (ADVP) + A
VIII. XP X
455

12. Which of the following sentences has a structure that cannot be generated with these rules?
a. Sue sells slews of shiny shells.
b. Sue sold Sean shiny shells.
c. Sue sells slews of sea shells.
d. Sue sells shiny shells to Shawn.
13. Which of the following noun phrases indicates that Rule IV is inadequate?
a. sacks stuffed with stinky shells
b. shiny shells from sunny shores
c. stinky sacks of shiny shells
d. shinier shells from Sanibel
14. Which of the following sentences indicates that the internal structure of noun phrases has more
levels than the rules allow?
a. If Sue sells shells, Shawn should sue her.
b. If Sue sells shells, Sean should see them.
c. Sue sells sea shells, and Sean sells them too.
d. None of the above.
15. Given the above rules, the best way to generate a sentence like Sue surely sells shells would
be to add the element ADVP to some part of
a. Rule I.
b. Rule II.
c. Rule III.
d. Rule IV.
456

16. Consider the following data:
STATEMENTS QUESTIONS
Sue will sell the shells. Will Sue sell the shells?
Sean shouldn’t sue Sue. Shouldn’t Sean sue Sue?
Given the PS-rules on page six, statements and questions such as the above can be related by
which of the following transformational rules?
a. NP + T T i + NP + [e] i
b. NP + M M i + NP + [e] i
c. NP + TP TP i + NP + [e] i
d. NP + VP VP i + NP + [e] i
THE NEXT SEVEN QUESTIONS (17 - 23) ARE BASED ON THE FOLLOWING DATA
FROM LATIN:
I. Clara regina pulchram fabulam laudat.
The famous queen praises the beautiful story.
II. Reginae claras fabulas laudabant.
The queens praised the famous stories.
III. Clarae reginae casas aedificant.
The famous queens build the cottages.
IV. Amica regina aegras agricolas curabit.
The friendly queen will care for the sick farmers.
V. Inimicae agricolae aegram lupam declinant.
The unfriendly farmers avoid the sick wolf.
VI. Inimicae agricolae declinabant.
The unfriendly farmers turned away.
457

17. Which of the following sets of PS-Rules best describes the above data (in each set, assume that
there is a rule rewriting T as PRS, PST, or FUT)?
a. S NP + TP + VP
VP (NP) + V
NP (AP) + N
b. S NP + TP + VP
VP V + NP
NP AP + N
c. S NP + VP
VP (NP) + VERB
VERB V + T
NP (AP) + N
d. S NP + VP
VP NP + VERB
VERB V + T
NP (AP) + N
18. According to the data given, which of the following is correct regarding the verb declina-?
a. It is either transitive or intransitive.
b. It is either transitive or copulative.
c. It is either copulative or intransitive.
d. It is always transitive.
19. Which of the following is correct?
a. Adjectives agree with nouns in case and number.
b. Verbs agree with subjects in number.
c. Both a and b are correct.
d. Neither a nor b is correct.
458

20. Which of the following is correct?
a. The morpheme meaning PST is -ba-.
b. The morpheme meaning FUT is -bi-.
c. The morpheme meaning PRS is -Ø-.
d. All of the above are correct.
21. Which of the following is a grammatical Latin sentence?
a. Reginas agricolae curabit.
b. Reginae agricolas curabat.
c. Regina agricolas curant.
d. Regina agricolas curabat.
22. Which of the following is a grammatical Latin sentence?
a. Amicas agricolas casae aedificant.
b. Inimica regina lupas declinat.
c. Lupa inimicas reginas curant.
d. Aegrae reginae aegras lupas curat.
23. If the verb meaning ‘free’ is libera-, and if the adjective meaning ‘evil’ is mala-, then which
of the following sentences means ‘The farmer frees the evil queen?
a. Agricola malam reginam liberat.
b. Agricola malas reginas liberabat.
c. Agricolae malam reginam liberant.
d. Agricola malam reginam liberabat.
459

THE NEXT FOUR QUESTIONS (24 - 27) REFER TO THE FOLLOWING STRUCTURE:
24. If a sentence has the above structure, then
a. its main verb might have a direct object.
b. its subject noun phrase must have an overt determiner.
c. its main verb must be transitive.
d. all of the above are correct.
25. If a sentence has the above structure, then
a. its auxiliary could be PRS.
b. its auxiliary could be FUT.
c. its auxiliary could be IMP.
d. its auxiliary could be CND.
26. If a sentence has the above structure, then
a. its main verb could be give.
b. its main verb could be become.
c. its main verb could be fall.
d. all of the above are possible.
460

27. Which of the following sentences has the above structure?
a. That man washes the windows for the neighbors.
b. That man washes his car with rags.
c. People wash cars with rags.
d. The lioness dragged the cubs into the cave.
THE NEXT EIGHT QUESTIONS (28 - 35) REFER TO THE FOLLOWING STRUCTURE:
28. If a sentences has the above structure, then
a. its subject might contain a modifier.
b. its object might contain a compound noun.
c. its auxiliary must contain a negative.
d. its auxiliary might contain a negative.
29. If a sentence has the above structure, then
a. sea shells could be the object of its main verb.
b. some sea shells could be the object of its main verb.
c. sea shells could be its subject.
d. shinny shells could be its subject.
461

30. If a sentence has the above structure, then
a. sell could be its main verb.
b. sells could be its main verb.
c. sold could be its main verb.
d. all of the above are possible.
31. If a sentence has the above structure, then
a. its tense might be PRS.
b. its tense might be PST.
c. its tense might be CND.
d. all of the above are possible.
32. If a sentence has the above structure, then which of the following words cannot be in the
sentence?
a. surely
b. too
c. will
d. sharply
33. If a sentence has the above structure, then which of the following morphemes cannot be in the
sentence?
a. FUT
b. Ø
c. IMP
d. PST
462

34. If a sentence has the above structure, then the sentence cannot contain any
a. adjectives.
b. adverbs.
c. prepositions.
d. quantifiers.
35. Which of the following sentences has the above structure?
a. The sea storm should slosh some shiny shells ashore.
b. The soapy surf shall surely slosh slimy shells ashore.
c. The sea storms shall surely slosh shiny shells ashore.
d. The soapy surf should surely slosh slimy shells ashore.
THE NEXT THREE QUESTIONS (36 - 38) REFER TO THE FOLLOWING STRUCTURE:
36. The object of a sentence having this structure is in the position of
a. N3a.
b. N3b.
c. N3c.
d. Both N3a and N3c.
463

37. Which of the following nodes C-commands the main verb?
a. N3a.
b. N3b.
c. The C3 containing the tense.
d. All of the above.
38. Which of the following nodes does not C-command N3b?
a. N3a.
b. N3c.
c. V1.
d. The C3 containing the TNS.
39. Consider the following possible diagram for the structure of English sentences:
One problem with this kind of representation is that it
a. confuses structural and functional categories.
b. it cannot be reduced to a set of rules.
c. it makes the AUXILIARY a constituent of SENTENCE.
d. it makes the DIRECT OBJECT part of the VERB PHRASE.
464

40. Which of the following is correct in the TG model of syntax?
a. Every sentence has two structural levels, an abstract underlying structure roughly
equivalent to what is understood, and a real superficial structure roughly equivalent to
what is heard.
b. The order of phrases in the abstract underlying structure must be the same as the order
of phrases in the real superficial structure.
c. Abstract underlying structures cannot contain empty elements like [e].
d. All of the above are correct for TG.
41. Which of the following is correct in the RG model of syntax?
a. Every sentence has two structural levels, an abstract underlying structure roughly
equivalent to what is understood, and a real superficial structure roughly equivalent to
what is heard.
b. The order of phrases in the abstract underlying structure must be the same as the order
of phrases in the real superficial structure.
c. Abstract underlying structures can contain empty elements like [e].
d. All of the above are correct for RG.
42. Which of the following is part of the feature set for a word like no.
a. [+X3L, +ENH, –PRH]
b. [+X3L, –ENH, +PRH]
c. [+X3L, –ENH, –PRH]
d. [+X3L, +ENH, +PRH]
43. Which of the following is part of the feature set for the morpheme POS?
a. [+X3L, +ENH, –PRH]
b. [+X3L, –ENH, +PRH]
c. [+X3L, –ENH, –PRH]
d. [+X3L, +ENH, +PRH]
465

44. Which of the following is part of the feature set for the TNS morphemes?
a. [+X3L, +EVH, –PRH]
b. [+X3L, –EVH, +PRH]
c. [+X3L, –EVH, –PRH]
d. [+X3L, +EVH, +PRH]
45. Consider the following sentence:
The British left waffles on the Falkland Islands.
This sentence exhibits
a. structural ambiguity only.
b. lexical ambiguity only.
c. both structural and lexical ambiguity.
d. neither structural nor lexical ambiguity.
46. Which of the following labeled brackets correctly represents the structure of the sentence
Priscilla knows about the letter to Fred?
a. [ VP V – PP – PP ]
b. [ VP V – PP ]
c. Both a and b.
d. Neither a nor b.
47. Which of the following sentences can be cited to support the argument that there must be at
least one intermediate level between VP and V?
a. John wants to stop to rest, and so does Bill.
b. John wants to stop to rest, and Bill does too.
c. Both a and b.
d. Neither a nor b.
466

48. Which of the following transformations correctly describes the relationship between the
sentences in [1] and, at the same time, accounts for the data in [2]?
[1] The farmer looked over the fence. (=The farmer inspected the fence.)
The farmer looked the fence over.
[2] The farmer looked over the fence. (=The farmer gazed above the fence.)
*The farmer looked the fence over.
a. X 1 + V + P + NP + X 2 => X 1 + V + [e] i + NP + P i + X2
where NP immediately dominates 3 and 4
b. X 1 + V + P + NP + X 2 => X 1 + V + [e] i + NP + P i + X2
where VP immediately dominates 3 and 4
c. X 1 + V + P + NP + X 2 => X 1 + V + [e] i + NP + P i + X2
where PP does not immediately dominate 3 and 4
d. X 1 + V + P + NP + X 2 => X 1 + V + [e] i + NP + P i + X2
where PP immediately dominates 3 and 4
49. Which of the following sentences can be cited to support the argument that there must be an
two intermediate levels between NP and N?
a. John met this young history student, and Mary met that one.
b. *John met a history student, and Mary met a biology one.
c. *John met the teacher from Spain of Spanish.
d. All of the above.
50. Which of the following sentences can be cited to support the argument that sentence adverbs
are constituents of S whereas manner adverbs are constituents of VP?
a. Sue will get the mail on Monday, and Mary will too.
b. Sue will get the mail on Monday, but Mary won’t until Tuesday.
c. Both a and b.
d. Neither a nor b.
467

INDEX
Abstract Noun. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 309
Abstract Representation (AR). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 382
Acceptable. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 65
Accidental Gap.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 114
Accusative .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 313
Acoustic Phonetics.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 93
Active.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 313
Adjective. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 92, 310
Adjunct.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 189
Adverb. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 310
Affix.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 312
Affricate. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 99
Agglutinating Language.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 314
Allomorph. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 116
Allophone.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 108
Alphabet.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 325
Alternative Element. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 176
Alveolar. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 97, 98
Ambiguity. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 76
Analytic Language. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 314
Anaphor. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 240
Angular Gyrus. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 80
Antepenult. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 107, 127
Anterior. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 98, 101
Aphasia.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 37
Argumentation. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 40
Articulatory Phonetics. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 93
Aspect. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 313
Aspiration.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 107
Assimilation. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 119
Attributive Position. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 310
Auditory Phonetics.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 93
Back. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 95-97
Behaviorism. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 35
Bilabial.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 97, 98
Binding Relation. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 384
Biolinguistics. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 31
Bound Morpheme. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 312
Brace. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 176
Broca’s Area. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 80
C-adjunct. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 226
C-command. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 223, 239
C-specifier. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 226
Case. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 313
Central. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 95
Cerebral Lateralization.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 36
Chain. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 384

Characterizer. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 212
Chunk.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 167
Clause.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 57, 311
Cleft Sentence Test. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 166
Co-occurrence Restriction. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 188
Collective Noun. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 309
Command.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 239, 259, 383
Common Noun.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 309
Common Nouns. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 178
Comparative Degree. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 310
Complement.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 197
Concrete Noun. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 309
Conditional.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 177, 313
Conjugation. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 312
Conjunction. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 311
Consonant. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 97
Consonantal. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 101
Continuant. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 101
Continuity Theory . .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 79
Contraction Block Condition (CBC). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 385
Convergent Evolution. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 77
Copulative Verb.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 310
Core Grammar. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 47
Coronal.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 98, 101
Count Noun. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 309
Creativity. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 38
Curly Bracket. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 176
Declarative. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 311
Declension. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 312
Deep Structure. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 237
Demonstrative Pronoun. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 310
Descriptive Adequacy. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 51, 173
Descriptive Grammar.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 39
Diphthong. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 97
Direct. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 168
Direct Object. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 91, 168
Discontinuity Theory. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 79
Distinct Reference Condition (DRC).. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 385
Distinctive Feature. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 96, 109
Dominance. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 382
Double Trigger Condition (DTC). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 385
[e]–binding Condition (EBC). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 385
Empiricism.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 35
Empty Categories.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 236, 242
Empty Category Condition (ECC).. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 384
Endocentricity. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 186, 198
English Noun Phrase Condition (NPC). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 386
469

Exclamatory. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 311
Explanatory Adequacy. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 51, 173
Extended Neighborhood. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 383
Extended Neighborhood Condition (ENC). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 384
Feminine. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 313
First Person. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 156, 313
Flap. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 100
Formal Universal. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 48
Free Morpheme. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 312
Free Variation. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 116
Fricative. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 99
Front.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 95
Functional Category.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 174
Fusional Language. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 315
Future. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 177
Gender. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 313
Genitive. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 313
Glide.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 97, 100
Glottal. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 98
Grammar. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 38
Grammatical.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23, 65
Grammatical Characterization Problem.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25
Grammatical Realization Problem.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25
Head. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 193, 311, 382
Head–only. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 187
Hierarchical Order. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 162
Hierarchy. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 174
High. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 96
Historical Classification.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 316
Homorganic. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 119
Immediate C–command. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 383
Immediate Dominance. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 382
Immediate Neighborhood. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 383
Imperative. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 177, 311, 313
Indefinite Pronoun. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 309
Indicative. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 313
Indo–European.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 316
Inflection. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 312
Inflectional Ending.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 312
Inflectional Language.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 315
Intensive Pronoun. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 309
Interdental. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 99
Interrogative. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 311
Interrogative Pronoun. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 310
Intransitive Verb. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 236, 310
Isolating Language. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 314
L-command. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 383
470

Labeled Brackets. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 195, 228
Labiodental.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 98, 99
Language Variation. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 70
Lateral. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 100
Lax. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 97
Left Branching Construction. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 68
Levels of Adequacy. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 51
Lexical Ambiguities . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 202
Lexical Component. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 124
Lexicon. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 124
Limbic. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 82
Linear Order.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 162
Linguistic Competence. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24, 25
Linguistic Performance. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25
Linguistic Structure. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 40
Linguistic Universal.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 47
Linked Chain. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 384
Liquid.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 100
Long Term Memory.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 167
Low. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 96
LTM. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 167
Manner Adverb. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 172
Manner of Articulation. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 98
Marked. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 58, 119, 123
Masculine.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 313
Mass Noun. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 309
Maximal Projection. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 186
Mentalism. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 35
Mid.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 95
Minimal Pair. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 110
Modal. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 177
Mode. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 176, 313
Mood. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 176, 313
Morpheme. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 28, 115, 312
Morphology. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 115, 312
Morphology .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 28
Morphophonemic.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 116
Morphophonological Rule. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 116
Multiple Branching Construction. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 67
N-adjunct. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 226
N-specifier. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 226
Nasal. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 100, 101
Native Language. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23
Native Language Acquisition.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 31
Native Speaker.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23
Native Speaker Intuition. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 43
Nativism.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 35
471

Natural Class. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 53, 109
Necker Cube.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 61
Nested Construction.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 68
Neurolinguistics. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 31
Neuter.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 313
Nominative . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 313
Nonaccidental Gap.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 114
Nonanterior. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 98, 101
Nonback. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 97
Nonconsonantal. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 101
Noncontinuant. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 101
Noncoronal.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 98, 101
Nondistinctive Feature. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 109
Nonhigh. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 96
Nonlow.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 96
Nonnasal. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 101
Nonround. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 97
Nonsibilant.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 101
Nonsonorant.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 101
Nontense. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 97
Nonvocalic. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 101
Nonvoiced. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 101
Noun.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 309
Number. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 313
Nurturism. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 35
Object.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 212
Objective .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 313
Observational Adequacy. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 51, 173
Obstruent. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 101
Occipital Lobe. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 80
Omission. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 41, 164, 172
One Affix Condition (OAC). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 385
Open Class Categories. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 213
Open Phrase Categories. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 213
Open Words. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 33
Optional Element. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 176
Palatal.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 98, 99
Paraphrase. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 76
Paraphrase Test. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 166
Parentheses.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 176
Parietal Lobe. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 80
Parts of Speech.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 309
Passive. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 313
Past.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 177
Penult. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 107, 127
Perfective. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 313
Person.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 313
472

Personal Pronoun. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 309
Phoneme.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 108
Phonetic Alphabet. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 28
Phonetic Representation (PR). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 382
Phonetic Symbol. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 94
Phonetician.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 93
Phonetics. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 28, 93
Phonological Component.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 123
Phonological Rule. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 111, 146
Phonologist. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 93
Phonology. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 28, 93, 107
Phonotactic Rule. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 114
Phonotactics. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 114
Phrase.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 162, 163, 193, 311, 382
Phrase Structure Filter. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 226, 252
Phrase Structure Grammar.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 184
Pitch. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 105
Pivot Words. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 33
Place of Articulation. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 98
Placement.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 41, 172
Polysynthetic Language. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 315
Positive Degree. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 310
Possessive . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 313
Possessive Pronoun. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 157, 309
Precedence. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 382
Predicate.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 168, 311
Predicate Position.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 310
Prefix. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 312
Preposition. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 311
Prescriptive Grammar. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 39
Present. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 177
Progressive.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 313
Projection. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 382
Pronoun. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 309
Proper Noun.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 178, 309
Pseudo-cleft Sentence Test. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 166
Psycholinguistics. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 31
R-command. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 383
Rationalism. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 35
Reciprocal Pronoun. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 310
Recoverability of Deletion Condition (RDC). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 62, 246, 384
Recursion. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 382
Recursive. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 38, 215
Redundancy Rule. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 123
Reference. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 41, 163
Reflexive Pronoun. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 156, 309
Relative Pronoun. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 310
473

Release. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 109
Residence. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 213, 383
Residential Domain. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 383, 384
Residential Grammar. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 211
Right Branching Construction. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 68
Root. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 312
Round.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 96, 97
Rule Formalism. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 127
Sample Test One. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 403
Sample Test Four.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 450
Sample Test Three. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .432
Sample Test Two. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 414
Second Language Acquisition. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 31
Second Person. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 156, 313
Semivowel. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 100
Sentence. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 168, 311
Sentence Adverb. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 40, 172
Short Term Memory.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 167
Sibilant. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 101
Slash–dash Notation.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 111, 149
Sonorant. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 101
Specifier. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 190
Specifier Heaviness Condition (SHC). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 385
Spelling. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 125
Stem. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 312
STM. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 167
Stop. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 97, 101
Stress. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 105
Strict Order Condition (SOC). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 385
Structural Ambiguities.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 203
Structural Category. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 174
Structural Grammar. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 173, 184
Subject. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 91, 168, 212, 311
Subject Exclusion Condition (SEC). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 386
Substantive.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 312
Substantive Universal. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 48
Suffix. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 312
Superlative Degree.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 310
Supramarginal Gyrus. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 80
Suprasegmental. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 105
Surface Structure. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 237
Syntactic Categories.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 309
Syntactic Construction. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 311
Syntactic Feature. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 188
Syntactic Unit.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 193
Syntax. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 155, 312
Tense. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 96, 97, 176, 313
474

Third Person.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 313
Time. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 176
Topicalization. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 235
Trace. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 236
Traditional Grammar. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 168
Transformational Grammar. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 235, 237
Transformational Rule. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 236
Transitive Verb. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 236, 310
Tree Diagram. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 168
Typological Classification. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 314
Ultima. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 107, 127
Unacceptable. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 65
Ungrammatical.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23, 65
Unit. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 382
Universal Grammar. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 38, 46, 47
Unmarked.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 58, 119, 123
Velar.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 97, 98
Verb. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 168, 310
Verb Phrase. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 168, 174
Vocalic. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 101
Voice. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 313
Voiced. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 93, 99, 101
Voiceless. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 99, 101
Voicing.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 93
Vowel.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 93
Vowel Harmony.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 103
Wernike’s Area. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 80
Writing. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 39
X Projection Condition (XPC).. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 386
X-bar Syntax. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 193, 194
X3 Cue Condition (XCC). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 386
X3 Specifier Condition (XSC).. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 386
475

THE LANGTECH PARSER
Enter a sentence without punctuation and then click the Parse! button.
that experienced flight attendant probably should slowly describe all the
safety precautions to the nervous passengers in their native language
User Friendly Output (UFO):
Parse!
The clause “that experienced flight attendant probably should slowly describe all the safety
precautions to the nervous passengers in their native language” is a statement.
The phrase “that experienced flight attendant” is a noun phrase and the subject of the verb
“describe.”
The phrase “experienced” is an adjective phrase modifying the noun “attendant.”
The phrase “probably” is an adverb phrase expressing the speakers attitude.
The phrase “slowly” is an adverb phrase modifying the verb “describe.”
The phrase “all the safety precautions” is a noun phrase and the direct object of the verb “describe.”
The phrase “to the nervous passengers” is a prepositional phrase and the indirect object of the verb
“describe.”
The phrase “the nervous passengers” is a noun phrase and the object of the preposition “to.”
The phrase “nervous” is an adjective phrase modifying the noun “passengers.”
The phrase “in their native language” is a prepositional phrase and a modifier of the verb “describe.”
The phrase “their native language” is a noun phrase and the object of the preposition “in.”
The word “that” is a determiner specifying the noun “attendant.”
The word “experienced” is a common adjective in the positive degree.
The word “flight” is a singular common noun.
The word “flight” is a noun forming a compound noun with “attendant.”
The word “attendant” is a singular common noun.
The word “probably” is a sentence adverb in the positive degree.
The word “should” is a modal auxiliary specifying the meaning of the verb “describe.”
The word “CND” is an abstract morpheme indicating the conditional mode of “should.”
The word “slowly” is a manner adverb in the positive degree.
The word “describe” is the main verb of the statement.
The word “all” is a quantifier specifying the noun “precautions.”
The word “the” is a determiner specifying the noun “precautions.”
The word “safety” is a singular common noun.
The word “safety” is a noun forming a compound noun with “precautions.”
The word “precautions” is a plural common noun.
The word “to” is a preposition introducing the prepositional phrase “to the nervous passengers.”
The word “the” is a determiner specifying the noun “passengers.”
The word “nervous” is a common adjective in the positive degree.
The word “passengers” is a plural common noun.
476

The word “in” is a preposition introducing the prepositional phrase “in their native language.”
The word “their” is a third person plural possessive pronoun.
The word “POS” is an abstract morpheme indicating the possessive case.
The word “native” is a common adjective in the positive degree.
The word “native” is an adjective forming a compound noun with noun “language.”
The word “language” is a singular common noun.
RG Diagram Output:
In the above diagram, the following relationships are directly encoded into the phrasal architecture:
Specifiers (X3 level constituents): that (determiner), probably (sentence adverb), all (quantifier),
the (determiner), their and POS (possessive). In addition, the subject (that experienced flight
attendant) and the tense (should and CND) are specifiers of the verb describe.
Modifiers (X2 level constituents): experienced (adjective), slowly (manner adverb), nervous
(adjective), and in their native language (prepositional phrase)
Compounds (X1 level prehead constituents): flight (noun), safety (noun), and native (adjective).
Complements (X1 level posthead constituents): all the safety precautions (direct object) and to the
nervous passengers (indirect object). In addition, the nervous passengers is the object of the
preposition to, and their native language is the object of the preposition in.
477

478