To appear in the Journal of Pidgin and Creole Languages:

From French to Creole: The development of new vernaculars in the French colonial world. By Chris Corne. London: University of Westminster Press, 1999. Pp. x, 263. Paper,Ý £20 UK pounds (= approx. $30.00). [To order via e-mail (credit card orders) contact Battlebridge Publications at <battlebridge@talk21.com>]

Reviewed by Salikoko S. Mufwene, University of Chicago

ÝÝÝÝÝ This book was just out from the Press when Chris Corne died, after over two decades of productive and stimulating research on French creoles (FC), especially on Mauritian. It is a precious culmination of his scholarship, as he redirected it toward more breadth and depth, endeavoring to highlight commonalities and differences not only among FCs of the Atlantic, the Indian Ocean, and the Pacific, but also between them and other vernaculars that developed concurrently in the same colonies. The book also provides invaluable information on the following admittedly non-creole vernaculars: (1) Laurentian and Acadian in Canada, the ancestors of todayís Cajun in Louisiana, because they shed light on features of Louisiana Creole; (2) St. Bartsian and ìSt. Thomasian,î as they help us better understand the origins of features of the Lesser Antillean creoles; and (3) the pidgin-like ìfranÁais tirailleurî and ìfranÁais populaire ivoirien,î as well as the ìintertwinedî Michif. The latter group sheds light on the question of substrate influence. Creolists interested in comparisons of creoles lexified by the same language will appreciate the wealth of historical, contact, and structural information provided in this book.

ÝÝÝÝÝ From French to Creole is really up to the spirit of its subtitle, as itÝ is focused more on the contact aspect of the development of the new vernaculars and the origins of their structural features than on the question of whether or not they are creoles. Its conception and contents are somewhat reminiscent of Holm (1989), except that it focuses on language varieties that have evolved from the same lexifier. This approach enables Corne to provide more in-depth sociohistorical and structural information about vernaculars that I will venture to categorize as belonging to the same genetic group. The volume provides extensive sample texts and appendices of data, all of which are useful for examining the structural features that bear on the authorís genetic conclusions.

ÝÝÝÝÝ In more than one respect, the book is also reminiscent of Chaudenson (1992). First, being focused on French vernaculars, it is likewise shy of generalizing strongly about the development of creoles, though it provides useful information for anyone interested in making general conjectures on the subject matter. Like Chaudenson, Corne pays more than lip service to the fact that the diverse varieties that made up the lexifier in the colonies were typically nonstandard folk varieties. They were certainly not the standard varieties with which creoles have been misguidedly compared in the literature. He argues that the lexifier played a more important role in determining the structural features of the creole vernaculars than some creolists have acknowledged. He also agrees with Chaudenson that African substrate influence on creoles accounts for (some) structural differences between these colonial vernaculars and those spoken primarily by descendants of the French, for instance, St. Bartsian.

ÝÝÝÝÝ Some creolists will undoubtedly appreciate Corneís predilection for the term congruence over the more commonly used convergence. Though he presents no arguments for his preference, the term resolves a problem stemming from the ambiguity of convergence. Traditionally it has meant ìdevelopment toward a common pointî, but genetic creolists have generally used it to note that ìlanguages in contact have structural similarities.î Those arguing for decreolization as part of the ìlife-cycleî of some creoles, in which the traditional meaning of convergence is implied, will definitely welcome this disambiguation.

ÝÝÝÝÝ There are, however, three important differences worth highlighting between Corne and Chaudenson: (1) Corne covers his structural data with much more depth; whereas (2) Chaudenson engages the reader into an informative comparison of different forms of ìcreolization,î viz., that of language versus those of other aspects of culture. This is a more global approach, which highlights the fact that the development of creoles was not an isolated evolution, and that the extent of substrate influence seems to have varied from one cultural domain to another. For reasons that will become obvious below, I consider this book a useful complement to Chaudensonís (1992) brief discussions of structural features. 3) Unlike Chaudenson, Corne argues that the development of Mauritian Creole owes little to that of RÈunionnais.

ÝÝÝÝÝ From French to Creole can also be compared profitably to Valdman (1978). Unlike Valdman, Corne treatsÝ every new vernacular individually, and without suggesting that there are ìcreole featuresî shared by most of them. The assumption that there are ìcreole featuresî was widely accepted in the 1970s but has been questioned by some (including myself) since the mid 1980s. Corne does not really address the question of pidgin ancestors of creoles, either.

ÝÝÝÝÝ Corne observes on page 1 that ìthis book is in no way a comprehensive survey of all (Ö) contact-induced vernacularsî (CIV) lexified by French. Indeed, though non-creoles are included among the CIVs he covers, he still says little about indigenized varieties of French in Africa, which unlike ìle franÁais tirailleurî and ìle franÁais ivoirien,î have been lexified by scholastic French. These CIVs are equally informative about substrate influence and continuity in the transmission of the lexifier. They inform us that transmission of the lexifier (a topic to which I return below) need not be associated with the presence of native speakers in a particular community. Corne justifies his position with the following:

Of all the other places where the French left linguistic footprints, New Caledonia, Reunion, Louisiana (and indeed North America generally), the Caribbean, French Guiana, Mauritius, and the Ivory Coast have been chosen as representative cases. This selection is far from being arbitrary. Other places such as Algeria, Tahiti, parts of subSaharan Africa and so on, are just as interesting linguistically, but are subsumed to a great degree by the places and languages dealt with here. For example, Algeria, like New Caledonia, was une colonie de peuplement [ëa settlement colonyí], and at roughly the same period. (...) Tahitian French is comparable with New Caledonian French, except that the indigenous people at the time of French settlement all spoke Tahitian, as opposed to the multiplicity of Melanesian languages in New Caledonia, and French settlement was on a much smaller scale; consequently, where New Caledonian French shows rather little influence from Melanesian languages [presumably because the latter canceled out each otherís influence in relation to French], the influence of Tahitian on L1 Tahitian French is rather marked [i.e., strong]. (p. 4)

ÝÝÝÝÝ From the point of view of restructuring of the lexifier into a creole, Corne advocates the Complementary Hypothesis, though he does not say so explicitly. In the above quotation, he hints at factors that bear on the structures of the emergent CIV, the kinds of things which Mufwene (in press) discusses under the cover term ìthe ecology of language evolution.î In this connection, I will miss Corne because I think we would have a lot of notes to trade about the ways these factors influence the structures of his CIVs. In From French to Creole, he argues that while the substrate populations in New Caledonia were linguistically less homogeneous than their counterparts on Tahiti, they were typologically less heterogeneous than their counterparts in New World French colonies. This variation in the make-up of the substrate languages explains why Tayo, for instance, exhibits more extensive incontrovertible Kanak substrate influence than Lesser Antillean CIVs do of African languages. According to Corne, the more diversity there was in the structures of the substrate languages, the more likely it was for features from the lexifier to prevail in the emergent CIV. This was the case for the major constituent orderÝ in New Caledonian CIVs, whose SVO order mirrors that of French. The substrate languages have both VOS and SVO orders. We must, however, also note here that French itself offered two basic constituent orders: SVO for free constituents versus VOS for clitics. The selection of SVO vs. VOS is thus a more complex matter. Markedness factors of the kind discussed by Mufwene (1991), Chaudenson et al. (1993), and Corne himself play a significant role in this particular case, with allowance made for including congruence among those factors.

ÝÝÝÝÝ One may also argue that structural diversity alone is not the full story. One must also invoke the Founder Principle (Mufwene, 1996), and periodize the presence of the different substrate languages in the contact setting. Corne does not overlook this aspect of language evolution:

The relative lack of postposed adjectives in Tayo may well be a pointer to the importance of [CËmuhÓ] as the language of the first group of Kanaks at St-Louis; alternatively, since for [CËmuhÓ] and [X’r’c˜˜] there is simply no category ìadjectiveî, it may be surmised that the items classed as adjectives in [DrubÈa] in fact are (or derive from) verbs; in either case Tayo again reflects the result of substratal congruence and shows signs of (possibly relatively recent) French influence. (p. 50)

ÝÝÝÝÝ Corne also invokes the ì(new) ecologyî of the colony to advantage in accounting for ways in which CIVs differ from metropolitan folk French, even if only in the statistical frequencies of some construction patterns, such as verb serialization or the final position of QU-/WH- phrases. On the latter construction, he goes as far as to specify the conditions of its occurrence (as opposed to the alternative with QU- in sentence-initial position). In this particular case, Corne is quite successful in highlighting the amplifying role of the ìnew ecologyî which, in my metalanguage (Mufwene, in press), is the new ìfeature poolî to which the New Caledonian languages have added their own features to reinforce, or simply compete with, some alternatives in the lexifier. In Corneí s own words:

It seems reasonable to conclude that the higher incidence of QU-final in NCF than in metropolitan (and other) varieties of French is another reflection of the contact between French and the Kanak languages. (...) The pattern of QU-final already exists in French, a Kanak L2 speaker of French therefore perceives French as being similar to the Kanak language concerned with respect to the placement of QU, and the result of this congruence is the promotion of the Qu-final pattern. (p. 39)

Ý

ÝÝÝÝÝ Corne clearly argues against any straightforward transfers of substrate features into the CIVs, observing that generally the features reveal simplifications and/or regularizations relative to their sources. He then warns the reader against simplistic conjectures of relexification:

if Tayo with a relatively homogeneous collection of substratal languages, reflects the structures of these only in a general way, then one wouldnít expect other varieties of Creole French with more diverse substrates to be the relexification of just one of them, and if it turns out that they are, there has to be an historical reason for this; if Tayo, in spite of its similar substrates, reflects congruence with French, one would expect languages with more diverse substrates to do so as well, although the congruence may be well-disguised enough to render the substrate presence almost invisible (...) ó put another way, the less in common, the greater the impact of French (...) (p. 66)

One shouldÝ not forget that those who developed creoles and the like did intend and make an effort to speak another language. In the process, they acquired most of the structures that they targeted, and this despite the consequences of normal imperfect replication (Mufwene, in press).

ÝÝÝÝÝ Corne is also very informative in making clear that typically more than one variety developed from the contact of French with more or less the same substrate languages. In the case of New Caledonia, he distinguishes New Caledonian French from Tayo. The former developed in a setting where there was more regular interaction with native/fluent French speakers, whereas the latter emerged in a setting where the proportion of fluent speakers of French was very low.

ÝÝÝÝÝ The same approach to the evolution of French into variable CIVs is typical of subsequent chapters in the book. Overall they are more informative on the CIVsí structures than most books on the development of creoles. Experts on specific creoles are better qualified to assess what is inaccurate and what other data bear on the positions defended in the book. Every now and then Corne tackles another theoretically interesting question, such as that of break in the transmission of the lexifier, a factor often invoked to account for structural deviations of pidgins and creoles from their lexifiers. He identifies this as ìrupture in ënormalí language transmission from generation to generationî (p. 72). This is one of the most gratuitous working assumptions in creolistics, made more complicated by Corneís own formulation. First of all, the notion of generation in language is relevant more to gauging evolution than to transmission. Language is transmitted more horizontally than vertically, even in settings in which creoles developed. Restructuring is partly determined more by who a learner interacts with than by generational differences. Commanding the language better than a learner is of course a significant factor, but this is independent of generation. In immigrant groups, children often help their parents in acquiring the target language or improving their command thereof. This leads to another observation: vertical transmission is bidirectional in any normal community, which makes insistence on generation pointless.

ÝÝÝÝÝ Coming back to the putative break in the transmission of the lexifier, if creoles had really developed under such conditions, everything that can be traced back to the lexifiersí systems would be due to accidents, in the same way that all natural languages share some features. For instance, the same segmental phonemic sequence /di:p/ is associated with the same physical denotation in English and in Kiyansi, my ethnic language (Bantu B85). It would be particularly difficult to account for the more obvious and overwhelming lexical similarities established between the creoles and their lexifiers, as opposed to substrate lexical contributions. Indeed, this observation takes us to the question of whether creoles are genetically related to their lexifiers. Since genetic kinship is posited on the assumption that a (proto-)language was transmitted continuously from one group to another, it has been easier to support the position that creoles are not genetically related to their lexifiers by positing breaks in their transmission. However, Corne presents no evidence to support this view. His endorsement of the break in transmission assumption is a distraction. Fortunately, his otherwise sensible approach to the development of creoles is not negatively affected by this shortcoming.

ÝÝÝÝÝ One of the cases in which the assumption is a distraction is when Corne argues that ìno clear break in language transmission occurredî in the case of RÈunionnais, which ìis to be seen as a continuation of French (dialectally leveled)î (p. 73). Interestingly, the CIV is by his own admission influenced by substrate languages. It is not clear whether the difference between those settings in which ìtrue creolesî developed and those where non-creole CIVs have emerged is one of kind or one of degree. Corneís identification of RÈunionnais as a CIV straddling over the creole-versus-non-creole status suggests that the difference must be one of degree. This prompts me to ask whether cases in which ìtrue creolesî such as Mauritian and Haitian developed as additional language varieties ó without replacing the lexifier in the contact community ó really involveÝ a break in its transmission. In genetic creolistics, the putative break is typically also associated with the low proportion of native speakers. Historically, this assumption is disputable, because it is based on a misunderstanding of the nature of the lexifier which was typically a colonial koinÈ (a fact that Corne got right). Children born especially in the homestead communities acquired the koinÈ natively, regardless of race.

ÝÝÝÝÝ Another weakness of the book lies in Corneís attribution of the following assumption to the substrate and superstrate hypotheses: the common thread of both hypotheses is allegedly failure on the part of those who developed creoles to acquire the lexifier perfectly (p. 225). Here he goes along with Baker (1997) in claiming that there was no target for the developers of creoles. This position presupposes that languages are acquired perfectly in communities with native-speaker majorities. If this were the case, there would be no language change in such communities. Cases of Siamese twins set aside, there are no two children in a family that have been exposed to exactly the same samples of their language and developed identical grammars. Otherwise the notion of ìidiolectî would have no significance at all, especially since children develop their communicative systems based on their individual experiences but are not told what to include in, or leave out of, those idiolects. Eventually, as explained in Mufwene (in press), it is all more a matter of degree than of kind in how an individual develops competence in a language, making allowance for restructuring, which in some cases is influenced by language varieties he/she already speaks. However, just because the subordinate populations who developed creoles were exposedÝ to different models of the lexifier does not make their case so different that the substrate and superstrate hypotheses must be mischaracterized to suggest failure in language acquisition. Among Corneís conclusions is the following:

Ý

in some cases people were already speaking varieties of French, in other cases people were making little attempt to speak ìFrenchî but rather dealing with a communication problem (p. 232).

ÝÝÝÝÝ This observation does not help us address the question of how CIVs have generally developed. In none of the cases were speakers not ìdealing with a communication problemî to solve. His own arguments for the development of some of the varieties by koinÈization underscores my counter-observation. If the substrate populations did not target the lexifier, there would be far fewer words from it in any relevant CIV, and there would be no point in identifying a particular language as a lexifier. Corneís observation also suggests that those who developed creoles did so deliberately, and perhaps somewhat by conscious design. Such a view stems from a misunderstanding of the fact that language changes are generally accidental consequences of how speakers ìexaptî (i.e., unwittingly modify) forms and structures while they communicate with each other. This is as true of creoles as of any other language. It is high time we re-examine some of the prejudices in our working assumptions about the development of pidgins and creoles. That standard varieties of languages around the world are hardly ever used perfectly ó which justifies the institution of academies in some societies ó is evidence that speakers seldom deliberate on how to modify their language.

ÝÝÝÝÝ I have criticized some of Corneís shortcomings in detail simply because I fear that the reader ó rightly seduced by Corneís impressive book ó may overlook some important conceptual issues. If we do not address them, genetic creolistics can hardly make a contribution to genetic linguistics or historical dialectology. Creolistics can help linguistics at large to understand better how languages evolve into new forms. From French to Creoles offers so much toward such a contribution. For this and other reasons, I highly recommend the book for a class on the development of CIVs.

References

Baker, P. (1997). Directionality in pidginization and creolization. In A. Spears & D. Winford (Eds.), The structure and status of pidgins and Creoles (pp. 91-109). Amsterdam: John Benjamins.

Chaudenson, R. (1992).Ý Des Óles, des hommes, des langues: essais sur la crÈolisation linguistique et culturelle. Paris: L'Harmattan.

Chaudenson, R., Mougeon, R., & Beniak, E. (1993). Vers une approche panlectale de la variation du franÁais. Aix-en-Provence: Institut díEtudes CrÈoles et Francophones.

Holm, J. (1989).Ý Pidgins and creoles. Vol. 2: Reference survey. Cambridge: Cam‚bridge University Press.

Mufwene, S. S. (1991).Ý Pidgins, creoles, typology, and markedness. In F. Byrne & T. Huebner (Eds.), Development and structures of creole languages (pp. 123-143). Amsterdam: John Benjamins.

Mufwene, S. S. (1996). The Founder Principle in creole genesis. Diachronica, 13, 83-134.

Mufwene, S. S. (in press). The ecology of language evolution. Cambridge: Cambridge University Press.

Valdman, A. (1978). Le crÈole: structure, statut et origine. Paris: Klincksieck.