LINGUIST List 13.1114

Tue Apr 23 2002

Disc: New: Van Den Berg's Review of Mufwene

Editor for this issue: Karen Milligan <karenlinguistlist.org>

Directory

1. Salikoko S. Mufwene, Van Den Berg's review of The Ecology of Language Evolution

Salikoko S. Mufwene, University of Chicago

Re: Linguist 13.1100

Margot van den Berg (MB) has written quite an interesting assessment
of my ecological approach to the development of creoles. Subscribing
to the uniformitorian principle, I argue that "creoles have developed
[gradually,] by the same restructuring processes that mark the
evolution of non-creole languages" (1), and, throughout the book, that
contact has played an important role in all cases of language
evolution. (As explained in the Preface, p. xi, I use the term "to
cover long-term changes observable in the structures and pragmatics of
a language, as well as the non-to-unusual cases where a language
speciates into daughter varieties identified at times as new dialects
and at others as new languages. It also covers questions of language
endangerment and death.") I submit that individual speakers (more
specifically, their minds) are the loci of contact and the arenas of
competition and selection of features into new varieties (14, 30-33),
that there is no qualitative difference between the contact of
languages (which has preoccupied creolists in particular) and that of
idiolects, except perhaps in the complexity of the processes that
produce communal language varieties.

In communicative settings involving contact of native and xenolectal
speakers, xenolectal varieties are an important contact component of
the mix (6, and Chapter 2). The extent of their impact on the emergent
communal variety was largely determined by social ecological factors,
such as the demographic proportion of nonnative speakers relative to
that of native speakers (who were not necessarily Europeans in the case
of Atlantic and Indian Ocean plantation settlement colonies - the early
creole populations were native speakers of the local European
vernaculars) and by whether or not populations of non-European descent
continued to socialize with those of European descent (Chapter 2). This
is important so that we can determine whether continuous linguistic
input from Europe was spread among populations on non-European descent
and to what extent. Today much (though not all) about the divergence
between African-American vernacular English (AAVE) and White American
English vernaculars has to do with this factor, because for about a
century after the Abolition of slavery there were more and more
European immigrations into the USA, while there were no more
significant importations of Africans (not even as indentured servants.)
Institutionalization of segregation through the passage of the Jim Crow
laws in the late 19th century prevented Africans from participating in
changes that affected White American vernaculars.

As MB observes, the algebraic equation for the linguistic contacts that
led to the development of creoles is a very complex one (Ecology, p.
21). Complex phenomena of course require complex representations,
regardless of whether or not we are capable of explaining everything
now. In the meantime, I propose that, ultimately, it is the contact of
idiolects that matters and that it is individual speakers who are the
unwitting agents of change (18, and Chapter 2). In this respect, new
non-creole varieties of European languages have evolved by the same
restructuring processes that have produced the vernaculars linguists
have disfranchised as Creoles. The emergence of both kinds of languages
is subject to competition and selection from among the features
attested in a pool to which even xenolectal speakers have made
contributions (6, and Chapter 2). Under specific circumstances, more
xenolectal features may be selected into the new system, especially
those that are (partially) congruent with those of the targeted
language, eliminating less congruent alternatives (52, 132-133). In
this light, one can also account for structural peculiarities of
Surinamese Creoles that do not originate in especially English, but, in
addition, also Dutch (in the case of Sranan) or Portuguese (in the case
of Saramaccan). One can also account in more or less the same way for
variation in the structural features of Creoles world-wide, according
to whether they developed in settings where substrate influence could
have been stronger (such as on sugar cane and rice plantations) or
weaker (such as on tobacco and cotton plantations in the North American
southeast, the cradle of AAVE, to the extent that one wants to claim
that it is a Creole). It is surprising that MB missed these facts,
because they are discussed in Chapter 2 and elsewhere in The ecology of
language evolution. Still, it remains true that none of this substrate-
influenced evolution is peculiar to creoles, as highlighted in Chapter
5: "What research on the development of creoles can contribute to
genetic linguistics."

Further evidence of the complexity is provided by structural variation
even among creoles that have evolved from the same European language.
One of the reasons is that the nature of the feature pool (both in
terms of the features coexisting in the relation characterized as
"competition" and the statistical strength of the variants) varied from
one ecology to another. This observation applies to all cases involving
contact of separate languages or just dialects of the same language.
Any exception? None that can conceivably be invoked, unless differences
among idiolects, however minor (but they can also be large!), are
dismissed. But the restructuring equation certainly also allows the
selection of non-congruent features into the emergent variety (Section
5.2.4). This has been true of the evolution of Latin into the Romance
languages, the diversification of Proto-Bantu into several new
languages (one cannot easily dismiss the influence of Pygmy and Xhoisan
languages), and I doubt very much that it was not the equally true of
the speciation of Indo-European languages. The evidence is obvious from
more recent evolution such as Irish English or the early emergence of
English from the contact of its Germanic ancestors both among
themselves and with Celtic languages of England. There is no particular
reason why a book that professes that all languages are mixed to some
extent (19) and argues that creoles have developed by the same
restructuring processes as noncreole languages would preclude such a
regular phenomenon from the development of creoles!

I did not invoke the Founder Principle to totally discard substrate
influence. I invoked it because structures of creoles have typically
been compared with those of the standard varieties of the European
languages from which they evolved. (In The ecology of language
evolution, I have focused on those varieties that evolved from contacts
of nonstandard varieties of European languages with non-European
languages. I find the extension of the term creole to other cases
rather problematic.)

European settlement colonies were founded by proletarian European
populations that spoke nonstandard vernaculars. Non-European slaves and
indentured servants worked side by side with European indentured
servants who also spoke either the same nonstandard vernaculars or
xenolectal varieties thereof. (The picture is ironically reminiscent of
the spread of scholastic varieties of, for instance, English across the
world since the late 19th century largely through nonnative speakers
and its subsequent indigenization.)

Focusing on those nonstandard varieties gives us a more accurate window
into the development of Creoles than the traditional practice has been
able to. There is no particular reason why the progressive construction
in Haitian Creole with preverbal ap(e) should be attributed
(exclusively) to African substrate influence when several of the
relevant African languages do not even have a similar construction and
when there is evidence of such a construction in le fran�ais populaire
(nonstandard French). As matter of fact, from the approach I have
adopted arises an interesting question: When and how did the local
acrolects with which the basilects have been compared emerge?
Unfortunately I do not address this question in my book. Nonetheless,
it is important because standard varieties played a marginal role, if
any at all, in the development of Creoles, particularly when these are
equated with basilectal varieties. Comparisons of creole structures
with their standard counterparts have been a waste of time and energy
with respect to highlighting differences and similarities between the
ways Creoles and their noncreole counterparts have developed (e.g.,
Gullah, on the one hand, and Old Amish English or Appalachian English,
on the other).

Whatever the situation may be, my ecological approach (which is an
elaboration of the complementary hypothesis I proposed in Mufwene
1986), does certainly not rule out the interesting cases adduced by MB.
I never argue against substrate influence (a normal and common
phenomenon in language evolution, as acknowledged above). I argue
against substratist accounts which claim either that Creoles' grammars
(with a possessive indeed) have their origins (largely) in the grammars
of substrate languages or that they are relexifications of particular
substrate languages. I do this also against universalist accounts that
invoke children as dei ex machina to invent new languages almost ex
nihilo. As much as I have been misidentified as a superstratist, I have
just intended to remind fellow researchers that one cannot ignore the
important role of the specific language varieties that were targeted by
those who were shifting from their respective native languages, knew
that the target language differed structurally from theirs, and they
had to learn them to the best of their abilities. They wound up
producing divergent vernaculars despite this effort. Substrate
influence consists of unwitting transfers into the target language of
features from previously spoken languages. (The reader might identify a
few in this Response!)

Like definitions of creoles that suggest that creoles inherited their
lexica from one language (hence the misnomer lexifier) but selected
their grammars from outside the European language (either partially or
almost totally), such approaches suggest that individuals can learn a
language simply by appropriating its lexical forms without the
morphosyntactic (and semantic) constraints associated with their usage,
thus they can literally associate the forms with grammars external to
their sources. Chaudenson (2001) argues convincingly that there is no
reported evidence of such a strategy in the literature of second
language appropriation. I doubt that this mistaken assumption is
supported by anything in the literature on first language development-
in any remote case that Creoles could have been made by children. One
can see even from Lefebvre's (1998) integration of her position that
even if relexification were plausible, there would be so many
exceptions. As DeGraff (2001, to appear) demonstrates, there are
several inaccurate matches between Haitian Creole morphosyntax and that
of Ewe-Fon so often invoked in support of the relexification
hypothesis.

One must also remember that where African languages are involved, these
languages are typologically heterogeneous, and the "competition and
selection" which recurs throughout my book obtained not only between
the European target language and the African languages, but even within
each of these groups. MB has adduced no counter-evidence to my position
so far, although it is true that I have not worked out all the
ecological factors that bear on the evolution of creoles, let alone
figure out how to formulate the language restructuring equation that I
suggest. Still, these shortcomings have nothing to do with recognizing
substrate influence, regardless of whether it works by congruence or
introduces features completely extraneous to the European language.

MB should remember that the most powerful constraint against a flood of
substrate influence on creole vernaculars is that those targeting the
European language wanted to speak it in the best way they could. Some
were more successful than others (just as there is variation among
native speakers - who are not equally competent, if they are measured
by some biased communal standard), though there is the important factor
of what particular varieties they were exposed to in the first place.
Substrate influence occurred despite the learners' determination to
develop the best competence they could in the target language, and
there were rewards for doing this, such as not being condemned to the
field-hands lot for the rest of one's life. We should definitely do
more research on how substrate influence works and why it cannot be
avoided. Where it is extensive, such as in Melanesian expanded pidgins
(Keesing 1988) and Berbice Dutch (Kouwenberg 1994), a good
understanding of idiosyncracies of the ecology is a must.

To conclude, there is no doubt on my mind that there is non-European
substrate influence in the development of creoles, just as there is
Celtic substrate influence in the development of Romance languages.
This does not, however, mean that the origins of all the features
usually associated with substrate influence lie primarily or
exclusively in the non-European languages in the case of creoles or in
the Celtic languages in the case of Romance languages, not when we can
find partial models in the target language itself (see also Corne
1999). In all cases, we are dealing with appropriation with
modification, which I identify as "imperfect replication," and this is
another area that requires a more adequate account than I have seen in
the relevant literature. My goal was quite modest: to develop more
awareness of the complexity of the subject matter of the development of
creoles and propose a research avenue worth pursuing.

REFERENCES
Chaudenson, Robert. 2001. Creolization of language and culture. London:
Routledge.

Corne, Chris. 1999. From French to Creole: The development of new
vernaculars in the French colonial world. London: University of
Westminster Press.

DeGraff, Michel. 2001. Morphology in Creole genesis: Linguistics and
Ideology. In Ken Hale: A Life in Language, ed. by Michael Kenstowicz,
53-121. Cambridge, MA: MIT Press.

DeGraff, Michel. to appear. Morphology & word order in `creolization'
and beyond. In Handbook of Comparative Syntax, ed. by Richard Kayne &
Guglielmo Cinque. New York: Oxford University Press.

Keesing, Roger. 1988. Melanesian Pidgin and the Oceanic substrate.
Stanford: Stanford University Press.

Kouwenberg, Silvia. 1994. A grammar of Berbice Dutch. Berlin: Mouton de
Greuter.

Lefebvre, Claire. 1998. Creole genesis and the acquisition of grammar:
The case of Haitian Creole. Cambridge: Cambridge University Press.

Mufwene, Salikoko S. 1986. The universalist and substrate hypotheses
complement one another. In Universals versus substrata in creole
genesis, ed. by Pieter Muysken and Norval Smith, 129-162. Amsterdam:
John Benjamins.