Nominalization, relativization, and attribution in ... - LINGUIST List
Nominalization, relativization, and attribution in ... - LINGUIST List
Nominalization, relativization, and attribution in ... - LINGUIST List
Create successful ePaper yourself
Turn your PDF publications into a flip-book with our unique Google optimized e-Paper software.
56<br />
fewer than seven dist<strong>in</strong>ct nom<strong>in</strong>aliz<strong>in</strong>g forms. 2 In addition to dist<strong>in</strong>guish<strong>in</strong>g<br />
categorically between the various grammatical types of embedd<strong>in</strong>g (relative<br />
clause vs. sentential complement, etc.), these forms systematically <strong>in</strong>dicate<br />
whether the situation expressed by the embedded clause is aspectually<br />
realized or unrealized. Burmese nom<strong>in</strong>alizers thus grammatically encode<br />
dist<strong>in</strong>ctions which <strong>in</strong> Angami must be derived from context or specified by<br />
means of additional (e.g. aspect) morphology.<br />
In between these two extremes lie most other Tibeto-Burman<br />
languages, <strong>in</strong>clud<strong>in</strong>g Lotha Naga. In Lotha, two formally dist<strong>in</strong>ct<br />
nom<strong>in</strong>alizers cover the same functional territory as that of Angami k\˙- <strong>and</strong><br />
the seven Burmese embedd<strong>in</strong>g nom<strong>in</strong>alizers. However, while the mean<strong>in</strong>g<br />
dist<strong>in</strong>ctions encoded by the Burmese forms are highly systematic, the<br />
functional constrast between the two Lotha forms appears to resist<br />
systematic description. Given English-based sensibilities <strong>and</strong> the<br />
grammatical subord<strong>in</strong>ation types mentioned above, one might predict that<br />
the two markers would be used to dist<strong>in</strong>guish formally between NP-modify<strong>in</strong>g<br />
types (i.e. relative <strong>and</strong> attributive clauses) <strong>and</strong> those that modify VP’s<br />
(sentential complements). However, such is not the Lotha situation.<br />
Rather, the available evidence suggests that the two forms are used, at least<br />
<strong>in</strong> one area of the grammar, to contrast realized <strong>and</strong> unrealized embedd<strong>in</strong>gs,<br />
as <strong>in</strong> Burmese. The evidence for this observation, as well as its implications<br />
for language change, are considered <strong>in</strong> what follows.<br />
1. Angami<br />
We will beg<strong>in</strong> by consider<strong>in</strong>g the Angami system. As mentioned above,<br />
Angami has a general nom<strong>in</strong>alizer k\˙- which is employed <strong>in</strong> a diversity of<br />
functions: to derive gerundives <strong>and</strong> abstract nouns from verbs, to<br />
subord<strong>in</strong>ate relative clauses <strong>and</strong> adjectival verbs to nom<strong>in</strong>al heads, to embed<br />
sentential complements, <strong>and</strong> to derive deverbal adverbs. k\˙- is prefixed<br />
either directly onto the verb, or onto the clause-f<strong>in</strong>al verbal marker (VM) if<br />
one is present. When the nom<strong>in</strong>alizer attaches directly to the verb, the<br />
result<strong>in</strong>g form, taken out of context, may have several possible<br />
<strong>in</strong>terpretations; for example k\˙m\˙s—a, (from m\˙s—a ‘to clean’) could mean either<br />
‘clean<strong>in</strong>g’ (gerundive), ‘cleanl<strong>in</strong>ess’ (abstract noun), ‘clean’ (attributive<br />
adjective), ‘cleanly’ (adverb), or, when followed by a def<strong>in</strong>ite article, ‘the clean<br />
one’ or ‘one who cleans’!<br />
Of course, structural <strong>and</strong> semantic factors help to dist<strong>in</strong>guish between<br />
these functions <strong>in</strong> actual contexts of use. In relative clause constructions,<br />
2 This figure does not represent the total number of nom<strong>in</strong>alizers <strong>in</strong> Burmese; however, for<br />
the purposes of this paper I am only consider<strong>in</strong>g those which correspond to the basic<br />
grammatical subord<strong>in</strong>ation types identified <strong>in</strong> the first paragraph.