4 years ago

Structural analogy in language and its limits [Contrast and analogy v.2]

Structural analogy in language and its limits [Contrast and analogy v.2]

John Anderson 6 As is

John Anderson 6 As is again familiar, some pre-centre s-initial clusters cannot be reconciled with sonority sequencing requirements, as illustrated in (11.a): (11) a. sport, strip, squeeze b. slow, snow c. cusp, disk, pest d. chasm, prism The clusters in (11.b) conform with sonority sequencing, as do the post-centre clusters in (11.c); whereas that in (11.d), /-zm/, doesn’t. The initial two elements in (11.a) are both obstruents, therefore remain unsequenced by hierarchy 1 in (2). If we extend (2) as in (12), sonority-based sequencing can be seen to be in accord with what we find in the clusters in (11.b/c): (12) sonority hierarchy 2 a. vowel < sonorant consonant < fricative < plosive b. {|V|} {V;C} {V.C} {C;} (12.b) introduces the specification for plosives, with dominant C, which includes voiceless, {|C|}, and voiced, {C;V}, and for fricatives, which involves contrastive non-dependency, represented by ‘.’: the V and C in the representation of fricatives does not involve dependency, thus contrasting in this respect with sonorant consonants, {V;C} and voiced plosives, {C;V}, whose representations involve contrary dependencies. Fricatives are characterised by being in opposition with segment-types whose representations involve the same elements in a dependency relation: the representation for fricatives involves contrastive nondependency. 3 All the clusters we have considered so far are compatible with (12) except for those in (11.a/d). (11.d) may simply constitute an isolated exception (and there is a tendency to syllabify the sonorant – though I return to them later, in §1.3. But (11.a) represent a well-integrated part of the phonology of English – and other languages. A second familiar area of discrepancy with respect to the requirements of sonority sequencing is illustrated in (13): (13) a. sex, apse, bits b. seeks, creeps, beats c. logs, jobs, adze d. leagues, vibes, weeds e. act, apt f. creaked, steeped g. dragged, bobbed h. plagued, daubed i. width, depth In the first four sets in (13) the final fricative is preceded by a plosive, contrary to application of hierarchy (12). In the next four – (e) to (h) – the final cluster consists of two plosives, which cannot be sequenced by (12). The final set again involve a (voiceless) fricative further from the centre than a plosive, which latter can be either voiced or voiceless. Every second set in (13) involves an intransitive vowel; but there is no such set corresponding to the transitive (13.i). These forms with an intransitive vowel are almost always morphologically complex. This also applies to voiced final clusters: (13.c) and (d), and

7 Structural analogy in language, and its limits (13.g) and (h). So too the examples with mixed voicing in (13.i). The possibilities involving items with transitive vowel and final voiceless cluster – i.e. those exemplified by (13.a) and (13.e) – seem to be the most fully integrated into the phonology of English. Let’s start by looking at the structures involved with the kinds of items illustrated in (13). The final segments in (13), which are all coronal obstruents, are often described as ‘appendixes’ or ‘extra-metrical’. They extend our usual expectations concerning the dimensions of the syllable. And they do this not merely by allowing a final segment which violates the requirements of sonority sequencing but also in some cases by allowing more segments than are normally associated with the post-centre cluster. Normally, an intransitive centre in English is followed by only one segment, and so is a transitive centre and its complement. 4 Thus we do not find the possibilities in (14.a): (14) a. *helmp, *creamp b. helm, help c. cream, creep These forms all conform to the requirements of sonority sequencing – though the first example in (14.b) illustrates an extension to the hierarchy as given so far, an extension which is embodied in (15): (15) sonority hierarchy 3 a. vowel < liquid < nasal consonant < fricative < plosive b. {|V|} {V;C} {(V;C),C} {V.C} {C;} (Sonority can no longer be read off readily from the representations in (15): I take this up below, and propose a simple metric based on these representations in §1.3.) Nasal sonorant consonants are represented as more complex than other sonorants by virtue of an additional C, related to the V;C in non-contrastive non-dependency, a simple combination. That is, there are no contrastive segment types in English in whose representation C and V;C are in a dependency relation; this is the significance of the ‘,’ notation (for noncontrastive non-dependency), compared with the ‘.’ (for contrastive non-dependency) in the representation of fricatives in (12) (a significance again to be pursued below). It is only the co-presence of the two elements joined by ‘,’ that is contrastive, not the relation between them. Before returning to the coronal obstruents appendix, I want to consider how we are to capture the restriction on the dimensions of post-centre clusters which do not include such an appendix, i.e. the restriction that the latter apparently violates. The final consonants in (14.b) and (c) are extensions to the core of the syllable represented by the centre (± its complement). We can formulate this structural relationship as in (16): (16) consonant adjunction {C }⇒ { \{|V|}} (16) states that there is available to consonants the possibility of modifying the maximal obligatory construction headed by a vowel (so both transitive and intransitive projections in this case). To the left of the backward slash in (16) is specified the category that the consonant seeks to modify. Maximality is guaranteed by the head convention, a notion famil-

Contrast in phonology, structural analogy, and the interfaces
Structural analogy and universal grammar
Structural Realism: Continuity and Its Limits - Ioannis Votsis
1 Finegan, Edward, Language. Its Structure and Use. 3rd edition ...
The Structure and Content of the Body of an OLIF v.2 File
Correlation Between Structure of Bcl-2 and Its Inhibitory Function of ...
Correlation Between Structure of Bcl-2 and Its Inhibitory Function of ...