| Lehmann
noted that the correlation of these four parameter can
be better formulated if we omit subject from the clause
order type and recognize only two types of languages i.e.
VO type which will include VSO, SVO and VOS, and OV type,
which will include SOV, OVS and OSV. Given these two basic
orders, it is suggested that VO languages in general will
tend to have the following cluster of properties - VO,
Pr, NG, NA. That is, such languages would be prepositional
and will place the genitive and adjective after the head
noun. As against this, OV language would tend be postpositonal
and will place the genitive and the adjective before the
head noun. Not only this it was also suggested that VO
languages will be shown certain other correlations, namely,
in such a language the relative clause may precede the
head noun and the auxiliary verb may follow the main verb.
Further such language may show the opposite of these tendencies. |
|
| Before we
pass on to the next section in which we will be dealing
with the generative approach to linguistic universals
it would not be out of place to point out here that Greenberg
was extremely cautious in naming his universals as tendencies
rather than claiming them as absolute truths. But then
we also have some absolute universals i.e. universals
which are exceptionless. For instance the statement "all
languages have vowels" is an absolute universal. You have
thus seen that a language universal can be implicational
or non-Implicational, it can be absolute or non absolute. |
|
| 421.1.3.1.2 : GENERATIVE APPROACH AND
UNIVERSAL GRAMMAR |
|
| The 'generative
approach' to language universals originated in the works
of Noam Chomsky. According to Chomsky and serious study
of language should be concerned with the following two
important questions. First, what is the type of knowledge
or 'competence' that the native speakers of a language
have, which helps them to use the language in questions
to suit different situations? Secondly, how is this knowledge
obtained by them? To take up the first question first,
Chomsky says that the knowledge that we have of our language
consists of besides other things, an abstract system of
linguistic rules and principles. It is this hearer of
a language to produce and understand the grammatical sentences
of his language which are infinite in number. These principles
and rules constitute the formal properties of a language.
These are basically syntactic in nature but also belong
to other domains of language structure like phonology,
morphology and certain areas of semantics. Further the
knowledge or competence that we have of our language is
at an unconscious level. For instance, the following sentence
in Hindi is ambiguous in that it has the meaning (1a)
or (1b). |
|
| 1. |
sudha ne nalini
ko apne |
liye caay banaane
ko kahaa |
|
sudha Nalini to oneself |
for tea make to told |
|
|
(1a) Sudha asked Nalini to make tea for
herself (i.e. Sudha)
(1b) Sudha asked Nalini to make tea for herself (i.e.
Nalini) |
|
| Now, a native
speaker of Hindi, who is not trained in linguistics, may
not be able to explain precisely what makes this sentence
ambiguous. But all Hindi speakers, who know their language
well will have the tacit knowledge that he sentence has
two distinct meanings. |
|
| A grammar,
according to Chomsky, is the characterization of this
tacit knowledge which is rule governed. Let us consider
a couple of examples to see what we mean by our competence
being rule-governed. Let us first consider an example
from Tamil morphology. Tamil, like other Indian languages
has borrowed a number of nouns including items such as
time and carrom pronounced as tayam and kaaram. Now like
any Tamil noun, these borrowed items can take care post
positions such as accusative -ai, or dative -kku etc but
when a tamil speaker has to add a post position to these
borrowed items he is likely to render these as Taya-tt-ai
'time (acc.). Taya-tt (u) -kku 'to the time', or kaara-tt-at
'carrom (acc.) and kaara-ttu (u)-kku 'to the carrom'.
Note that the Tamil speaker is not adding the post position
directly to the stem form namely kaaram and Tayam, which
would yield forms like karam-ai or Tayam-ai or kaaram-kku
which are not very common. What this indicates is that
when Tamil speakers use these English nouns with postpositions,
they follow certain subtle rules of Tamil morphology.
In Tamil if a noun ends in the phoneme /m/ and if such
a noun belongs to the class of neuter nouns, i.e., nonhuman
class of nouns, then a few processes have to be carried
out in order to obtain the right output. These processes
involve the deletion of the final /m/ and addition of
a stem formative -tt, to which are affixed the postpositions.
The Tamil speaker's creation of such forms as kaara-tt-ai
or Taya-tt (u) - kku is a clear indication that he is
using here the rules of word formation of Tamil which
he has internalized as speaker of that language. |
|
|
