 |
 |
 |
 |
 |
At the global level, we have several sociolinguistic
models available. I have already briefly talked about the one
developed by Labov (1966) and Trudgill (1974); I have also indicated
the limitations of this variability paradigm. Though work in
this framework does go a long way in trying to capture diachronic
change in synchronic, across the generations, linguistic analysis,
it does not illuminate the power structures obtaining in a speech
community. In fact, most of these studies assume that norms
of social and linguistic behaviour which acquire a particularly
oppressive and exploitative character over time emerge as normal
workings of society and that there is a general consensus about
their relative hierarchy (see e.g. Labov 1966:64). The work
done by Fishman (1972, 1978) and his followers offers an alternative
framework in terms of language maintenance and language shift
and domain analysis. Instead of language structure, it focuses
on language use in different domains but once again in terms
of a theory of peaceful and collective consensus (cf. William
1992). The question of how language is used in sustaining and
manipulating power relations is not discussed. In Agnihotri
(2000) (in the St. Petersburg Paper to appear in the Sage Yearbook
2001), I tried to show how some of the sociolinguistic studies
(e.g. Agnihotri 1979, Mukherjee 1980, Satyanath 1982) constitute
a counterpoint to these models and argued that the sociolinguist's
faith in the theory of consensus is almost unethical. Another
proposal comes from Bernstein (1971-75) who did raise the questions
of differential language socialization. He made a distinction
between the elaborated and restricted code, the former being
associated with the middle class and the latter with the lower
class. He argued that social class determines linguistic structure,
which in turn reproduces social structure. Middle class is the
ideal and therefore its language must be elaborated and the
lower classes, through remedial teaching, must imitate the behaviour
of the middle class. This may also help them to overcome their
cognitive deficits. Bernstein's deficit hypothesis is an insult
to human intelligence and linguistic competence. This framework
forces you to characterize the middle class speech positively
and then makes you look for the absence of those features in
the lower class speech. His work reminds one of the far more
important work of Sapir and Whorf and several other scholars
including Hymes, Gumperz, Levinson, Lucy, and Brown who have
since worked on the theory of relativity. Sapir (1949:162) felt
that people were at the mercy of their language: |
No
two languages are ever sufficiently similar to be considered
as representing the same social reality. The worlds in which
different societies live are distinct worlds, not merely the
same world with different labels attached. |
Later Whorf (1956:221) pointed out that users
of markedly different grammars are pointed by their grammars
toward different types of observations and different evaluation
of extremely similar acts of observation. In other words, meanings
you construct depend on your frames of reference and these frames
it seems are given a priori by your language. Once again we
end up with a hopelessly fatalistic position - one which first
ignores the power relations in society and which does not even
hint at a way out of this determinism. It is only in some recent
work (see e.g. Gumperz and Levinson 1996; studies cited in Foley
1997) that scholars of the relativity model are beginning to
realize that meanings may actually be negotiated in the communicative
practices of a society and interpretations of those negotiations
may not be possible without ideological considerations. |
As I (Agnihotri 1998, 2000) have argued elsewhere,
most of these approaches to the relationship between language
and society take the existence of a language and 'a society'
or a culture as given - well organized systems in place, conceived
and functioning in an ideologically neutral space. The conflicts
that are an essential part of these systems and regulate their
maintenance and change are often ignored. That's why when such
scholars are confronted with systems that are fluid and volatile,
their efforts to fit them into prefabricated slots give way.
Neat formulations of linguistic systems get into trouble when
they are confronted with pidgins, creoles or mixed codes. When
he was confronted with mind-boggling alternation of languages,
Labov (1971:457) said: |
So
far no one has been able to show that such rapid alternation
is governed by any systematic rules or constraints and we
must therefore describe it as the irregular mixture of two
distinct systems. |
First of all, I don't think it is a question
of two systems; in a stretch of speech several so-called language-systems
may be involved. Secondly, I don't think any system, however,
codified and crystallized, is really distinct. There are always
grey areas. In spite of Labov's warning to the contrary the
search for constraints on mixing languages has continued. I
am not suggesting that there are no constraints; I am simply
saying that all individuals in all contexts function under the
same universal constraints. If we, for example, maintain that
'CVCV' is a universal structural pattern, we will find all linguistic
behaviour, be it in highly codified standardized languages or
in pidgins, creoles or mixed codes, obeying that pattern of
an alternation of consonantal and vocalic sounds. Again, if
we maintain that it is a part of the Universal Grammar that
languages will divide their lexical stock into lexical and grammatical
categories and that, lexical categories such as nouns, verbs
and adjectives can have comparable symmetrical expansions; we
should expect to find this phenomenon in all human speech. When
you begin to examine mixed codes from the point of view of two
independent discrete systems, you get into a process of almost
arbitrarily assigning segments to a 'matrix' and an 'embedded'
language, the hypothesis being that every constituent must belong
to an a priori given language (see e.g. Kachru 1975; Disciullo,
Muysken and Singh 1986; Pfaff 1979, Sankoff and Poplack 1981;
Sankoff, Poplack and Vanniarajan 1991 among others). |
Recent research on mixed codes has centred around
'free morpheme constraint' and the 'equivalence constraint'.
The former forbids the mixed code users to inflect the words
of one language with the morphology of another, something most
fluent mixed code users do all the time. It is possible, as
Singh (2000) has argued, that before an inflectional morpheme
is really isolated, a set of words with inflections glued on
must, first be borrowed. Yet we must accept, that for a multilingual,
whose language proficiency travels along a continuum varying
from native like control to minimal incipient competence in
some varieties, inflectional morphemes float across the linguistic
spectrum. The essence of the equivalence constraint is that
mixing can only occur in those area of grammar that are shared
by both the languages, i.e., mixing should not violate the syntactic
rules of either language (Kak and Agnihotri 1996). |
|
 |
 |
 |
 |
|
 |
