part of speech correlation

[D. N. S. Bhat]

	The suggestion of establishing a typological distinction between
noun-prominent and verb-prominent languages is very interesting. I believe,
however, that a typological distinction would be more interesting and more
profitable if it can be correlated with an underlying functional distinction.

	In the present case, the question that I would like to raise is whether we
can think of two distinct language types that make use of two entirely
different strategies of sentence-formation such that one leads to
predominently verbal languages and the other one to predominently nominal
languages.

	There is a basis for the former claim. Several grammarians (Hoffmann 1903
on Mundari, Boas on Chinook, Kinkade on Salishan languages, Van Valin on
Lakhota, etc.) have argued that there are languages that have only a single
part of speech, called predicates or contentives, with the function of
reference being carried out by pronominal markers occurring with the
predicate, and not by a distinct category of nouns.

	These languages do not need a distinct category of nouns because the
strategy of sentence structure used by them does not require it. They use
non-finite forms of predicates in the position in which familiar languages
place noun phrases, and these modify either the personal markers that occur
in the predicate or the predicate itself. The languages can easily be
regarded as verb-prominent as they manifest several distingishing
characteristics that can be associated with such a classification.

	Can we have any basis for the latter claim, namely that there can be
languages are highly nominal in the sense that they do away with the notion
of verbs? Sanskrit appears to provide some basis for such a possibility.
There is a tendency, in classical Sanskrit, of frequently replacing finite
verbal forms by verbal participles, which are nominal in character. That
is, it is possible to have a whole text containing sentences that possess
only nouns and nominals but no verbs.

	We can also think of an alternative strategy of sentence structure that
can form the basis of such a "verb-less" language.  Sanskrit uses
juxtaposition rather than predication in its nominal sentences. That is,
its nominal sentences contain two juxtaposed noun phrases, neither of which
can be specified as the predicate. (See Gren-Eklund, G., 1978, "A study of
nominal sentences in the oldest Upanishads" Almqvist and Wiksell
International, Sweden, for an interesting discussion of this topic). Thus
the strategy of juxtaposition can replace the strategy of predication, in
which case we would only need nouns but not verbs. Sanskrit also uses
juxtaposition as the strategy for the structuring of its noun phrases, and
hence it also does not require a distinct category of adjectives.

	The case of Sanskrit, however, is rather peculiar. In the classical period
it has developed several characteristics that may not obtain in any natural
language. It has developed these characteristics as result of the fact
that, at that period, it was primarily used as a tool for writing rather
than as one for speaking. For example, the process of compounding has
practically taken over the function of noun phrases. In some texts, we can
even find a compound that can cover a whole page!

	Further, this process of replacing finite verbs by participles hase not
led to a noun-prominent language at a later stage. The participles have
later on become verbals, as for example, in some of the modern Indo-Aryan
languages.

	The possibility of the  second alternative, therefore, is rather doubtful.
We might probably generalize that languages could be noun-less, but not
verb-less. This, however, is a very tentative generalization. I would not
be surprised if someone discovers a language that makes it necessary for us
to discard it.

	Regarding the occurrence of adjectives (either as parts of the verbal
category or of the nominal category), I believe that the underlying
functional distinctions provide a basis for a three-fold or even a
four-fold typological classification and not a two-fold classification.
This is because there do occur languages like English that have a distinct
class of adjectives, and hence we need to postulate a third language type
that is neither nominal nor verbal.

	Another point is that even languages that combine together adjectives and
verbs into a single category need not necessarily be "verb-prominent".
Manipuri, for example, does not have a distinct category of adjectives. It
combines them with verbs. But still, I do not think we can regard it as
verb-prominent. It has no agreement markers on its verbs, and has several
case markers that are added to its nouns. In fact, its inflectional markers
fall into two distinct, almost exclusive sets, a nominal set and a verbal set.

D.N.S.Bhat,
533, 15th Main, 5th Cross,
Saraswatipuram,
Mysore 570 009
India