Document Sample
Decreolization Powered By Docstoc
					Pidgins & Creoles                                                                      March 2002
Univ. of Essex                                                                        Peter Patrick

                         MODELLING DIACHRONIC VARIATION:

The basic idea underlying the concept of decreolization is simple and surely correct:
       the standard and lexifier language of a creole-speaking society exerts a very powerful
       influence on the development of the creole at all stages.

As with the basic idea underlying the continuum, this is a linguistic commonplace, familiar from
numerous cases of language contact (see e.g. Weinreich 1953, Thomason & Kaufman 1988).
       But the reification of this notion in the field of creole studies under the label
‘decreolization’ suggests some additional content: for example, that
       • linguistic processes at work in creoles are somehow unique in intensity, effect, etc.

The first use of the term ‘decreolization’ was at the 1968 Mona conference. Keith Whinnom
proposed it (Hymes 1971:111) as
        •       an analogue to the biological term ‘despeciation’.
The linguistic term describes a process that can
        “transform [a] creole... into a dialect of the standard”, a process of (re-)convergence.
As with the biological process,
        •       it is a teleological one: the definitive element is the outcome.
In Whinnom's sense, decreolization may apply perfectly well to a creole that shares no historical
relationship with the target standard; he sees it as a sort of ‘primary hybridization’.

Around the same time, Wm Stewart used ‘decreolization’ in a way that’s no longer common,
conceiving it as
        • a process used– perh consciously– by individual speakers on particular surface forms
He and Dillard (1971) resisted characterizing it as a specific linguistic process, and saw it as a
very superficial operation which may even be incapable of causing deep changes in the syntax
or in semantic categories. Later, Washabaugh (1976) also claimed that decreolization is
constrained by surface rather than deep structures.

The more common view today is of
        •      decreolization as a community-wide change occurring over decades,
with far-reaching effects such as the alteration of underlying categories, even the whole mode of
indicating temporality in a language. Scholars like Bickerton (1980:110) criticize a surface

Creolists have tried to elucidate the process by assimilating it to different cases of language
change: language acquisition has been a frequent choice. Many linguists (e.g. Bickerton 1977,
1980, 1981; Ferguson 1971) have aptly compared creole genesis with a child's first-language
acquisition (FLA).
        •       Bickerton also characterizes de-creolization as FLA
(1981: 191ff.), and builds the analogy into his Language Bioprogram Hypothesis.

Rickford (1983) strongly disputes the main evidence Bickerton offered here-- the variation in
sentential negation across the Guyanese continuum— and instead
        •      compares decreolization to second language acquisition (SLA),
following the persuasive arguments of Washabaugh (1976) that it more nearly resembles SLA in
            duration of process

                                            p 1 of 4
Pidgins & Creoles                                                                        March 2002
Univ. of Essex                                                                           Peter Patrick
           access to the target, and
           learner motivation or lack of it
(see also Andersen 1983, Schumann & Stauble 1983).

       But studies of decreolization suffer from a lack of the kind of longitudinal individual
data that characterizes both kinds of acquisition research. One difficulty in comparing
decreolization with language acquisition is that while the latter can be, and often is, observed,
       •       decreolization is almost always merely inferred from apparent-time data.

What we have evidence for is not the change of individuals over a significant period of time, but
only the contrast among speakers of different ages at the same period (Rickford 1983: 303).

It is only a metaphor, and quite possibly a misleading one, to speak of “acquisition” as one
surveys the variation among speakers moving up the continuum. Too freely assigning
diachronic interpretations to the synchronic data in this manner recalls (and occasionally
compounds) the
          •     historicist fallacy, over-interpreting all variation as a residue of change over time
                (C.-J. Bailey 1973, Bickerton 1975).

An acquisition perspective tends to slight a feature of decreolization that’s significant to
        loss of language varieties & structures over time at the basilectal end of the continuum.

This is the model usually assumed by linguists describing languages such as Gullah
(Jones-Jackson 1984), US AAVE (Fasold 1981), and Bajan (Alleyne 1980). All have apparent
or arguable creole origins but have sustained close contact with English and are said to have lost
the basilectal, and perhaps lower mesolectal, segments of the continuum.

       •   This loss of lects is similar to language death, a sort of language change which
              focuses our attention on the community and takes a longer historical view.

But Rickford (1983: 304) points out that while analogy with language death seems appropriate
on the community level,
        at the individual level “decreolization involves extensions of one's linguistic repertoire
        from an earlier stage [whereas] normal SLA involves replacements''
suggesting, perhaps, that in each case the analogy with decreolization is only partial.

Dorian (1981) claims no single linguistic mechanism or unique set of processes to explain all
the types of loss she observes in East Sutherland Gaelic – the classic study of language death –
discussing them as normal processes of change which also occur in vital languages.
        •       Decreolization may be, as a sociolinguistic event, unique to creoles,
                while yet resulting from perfectly general linguistic processes.

Another move in elucidating decreolization is to see it as
      •      language convergence.

This is of course true by definition, but it creates problems for the study of creole continua.
        • Emphasis on decreolizing changes in the creole literature deflects attention away
            from those divergent changes that do not tend toward the metropolitan variety

                                             p 2 of 4
Pidgins & Creoles                                                                      March 2002
Univ. of Essex                                                                         Peter Patrick

If US AAVE is descended from a creole, the processes must necessarily be called
decreolization. (The mainland U.S. African American community has its roots in the same
African diaspora as Caribbean creole-speaking societies, and has played a central role in the
development of the variationist linguistic methods that are best suited to study decreolization as
language change.)
Current differences between AAVE & white vernaculars are said to result from this creole
        • But AAVE is also said to be diverging rather than converging.
(see Spring 1987 American Speech; papers by Ash, Harris, Labov, & Myhill in Sankoff 1986)

Best proposals so far for specific cases of decreolization are for AAVE features (Fasold 1981,
Bailey & Maynor 1987), or that AAVE shares w/transitional creoles like Gullah (Rickford

    • Yet complete convergence of the AAVE speech community w/other American
       has not occurred, despite conditions more conducive than in Jamaica:
          being a minority among a large majority population promoting their own
       vernaculars, which are often closer to the targeted standard.

Most accounts of decreolization fail to specify precisely what the target is & who speaks it; nor
exactly how its structures can diffuse outwards, in the absence of robust and influential social
networks connecting standard speakers to those lower on the continuum. (Recall the separate-
norms argument of Winford 1988.)
       In light of these problems, there seems little reason for decreolization to be the default
model of language change in creole societies;
       • a neutral framework which does not force one to interpret variation in terms of its
           directionality is to be preferred.

Interdependence of the Continuum and Decreolization Models

        •       The Life-Cycle model of pidgin- and creole-genesis,
explicitly proposed by Hall (1962) but retrospectively traced to Schuchardt (1914 by Holm
        • links all similar varieties observable today as representing historical stages of a
            single process,
            • which begins with language contact and pidginization,
            • proceeds via nativization into a creole, and then
            • through decreolization gives birth to the creole continuum.
Though the first transition has been questioned, not many creolists have questioned the causal
link between decreolization and the continuum (but see Patrick 1991, 1999, Mufwene 1994).

Since the direction of decreolization is by def. towards the superstrate, it’s often assumed that
        •       the basilect existed first, then the mesolect was formed, and then the acrolect
(for those creolists who do consistently distinguish the acrolect from the metropolitan standard),
a progression subscribed to as recently as 1984 by Bickerton.
        However, as far back as 1968 Alleyne maintained that

                                            p 3 of 4
Pidgins & Creoles                                                                      March 2002
Univ. of Essex                                                                        Peter Patrick
        the varieties intermediate between the polar ends of the continuum existed all along:

        “[Even] at the beginning of the process, [the] creole was in fact everywhere
        only a major segment of a continuum of variation” (Alleyne 1971:182, my emph).

If true, this severs the assumed causal link between decreolization and the continuum,
         since the latter would have existed before the former could have gotten underway.

Later, Bickerton (1986: 226) reversed his earlier position, suggesting instead that
        • “the creole continuum... must have been formed 'backwards', so to speak,
        • acrolect first, then mesolect, then basilect,
        • as the pyramid of slave society slowly formed itself... from the top down”

These multiple proposals, after over 20 years, are somewhat embarrassing for a field which
  • hopes to make strong predictive connections b/w social & linguistic behaviors and
as both creole studies and sociolinguistics do!
        For the present, not enough is known about the relation, if any, between the ill-defined
set of processes called decreolization and the creole continuum model of synchronic variation.

The proposal I and others make is:
 •    to discard the concept of decreolization as in any way special to PCs, and
 •    proceed w/normal concepts of language contact, pressure from standard
     etc, as used in non-creole situations, but
 •    taking special account of the unusual typological and historical relations
     between Creoles and their lexifier languages when these remain in contact.

                                            p 4 of 4

Shared By: