THE SOCIOHISTORICAL SETTING FOR THE EMERGENCE OF FBCs

In the presentation of his model for the sociohistorical setting for creolization, Chaudenson applies the metaphor of the three components (trois unités) familiar to all students of French classical literature (p. 73). That notion was a rule formulated in the 17th century by Boileau as a basic rule for the classical tragedy: in a specific place, in 1 day, a single action. The prototypical place for creolization is an island. Although it certainly did not take place in 1 day, it operated relatively rapidly. The action component consists of the economic, demographic, and social factors inherent in the large-scale plantation. Chaudenson also compares the genesis of creoles to a witch's brew whose concoction depends on the presence of a single essential ingredient, be it a toad's spit or earth from a cemetery. In the case of the FBCs, that ingredient consists of the integration of an alloglot population numerically superior to the population speaking the target, socially dominant tongue.⁸

Central to Chaudenson's sociohistorical model is the two-phase development of the plantocratic colonial plantation system: the preliminary societé d'habitation “freehold society” and the second step societé de plantation “plantation society.” He stresses the fact that colonial plantations, particularly those producing sugar, were, in effect, agroindustrial establishments that required a large servile labor force. These plantations could not be established from scratch but required the progressive development of a complex infrastructure: capital, means of communication, administrative organization, and so forth. In the 17th and 18th centuries, the term habitation was polysemous. It referred to a type of freehold, a small-scale farm usually located on a leased track of land, as well as a plantation with White overseers and several hundred slaves. It is the first sense of that term that Chaudenson retains to label the first stage in the establishment of plantocratic societies. On what was to become the pearl of the Antilles, the most profitable insular European plantocratic colony, French Saint-Domingue (at present, Haiti), the initial settlements were primitive affairs that involved paired partnerships (matelotage) set up to hunt wild livestock and cure the meat to sell to seafarers. These buccaneers (from boucan, the fire to cure meat) might be assisted by an indentured servant or a slave. These rudimentary establishments diversified to the cultivation of cash crops that required neither a complex infrastructure nor a large workforce: tobacco, indigo, and coffee. A typical habitation would be comprised of a White family and a few indentured servants and, at a later stage, slaves. In this social context, conducive to reduced physical and psychological distance (Schumann, 1978), target input was relatively available and the alloglots' interlanguage did not diverge widely from the speech of the speakers of French. Although the first alloglots were integrated into the colonial society, the other essential ingredient for creolization, their demographic preponderance, is not present at this stage.

It is present, however, at the subsequent plantation stage that Chaudenson places 30 to 60 years after the initial settlement. The masses of newly arrived slaves (the bossals), most of whom were assigned to work in the fields or the sugar mills, had few opportunities to receive direct target language (TL) input. Their models were mainly the acculturated slaves, some of whom—the creoles—were born in the colonies and had acquired approximative versions of the TL of varying distances from the model. As Chaudenson puts it, this situation represented SLA twice or several times removed, and it resulted in approximations of the original to the nth degree. To support his claim that some of these approximative interlinguistic varieties were not far removed from the target input, he offers the testimony of Jesuit priests who declared that the slaves' linguistic adaptation was rapid, so much so, according to Mongin (1679/1984), that many metropolitan French could learn French from them. It should be borne in mind that in the first linguistic survey conducted in France during the Revolution (the Abbé Grégoire Survey), it was reported that at most one-sixth of the population spoke Standard French (SF) with any degree of fluency.

Chaudenson's two-step sociohistorical context for the genesis of FBCs rules out the possibility of a prior pidgin stage, developed either in situ or imported from Africa. Regarding the possible transplantation of a French-based pidgin learned by slaves in West Africa, in the absence of any concrete evidence, the burden of proof falls on its proponents. At most, during the early years of the establishment of plantocratic societies, newly arriving bossals might use highly deviant approximations of French that Chaudenson labels the beginner's jargon (le jargon des commençants). Later-arriving bossals would be exposed to the local creole that by then would have acquired an autonomous norm. Unless they had subsequent opportunities to have close contacts with speakers of French, they would have no reason to think that the creole was a language distinct from French.

All theories of creolization based on an essential prior pidgin stage assume a preponderant linguistically diverse servile population early in the establishment of a plantation colony. However, Chaudenson offers convincing demographic data demonstrating that in most French plantocratic colonies, the European segment of the population of plantation islands remained statistically dominant for at least the first 30 years of the settlement. Although he remains a staunch defender of the polygenetic origin of FBCs and of the lack of mutual intelligibility between those of the IO and the Atlantic-Caribbean zone (ACZ), he identifies a small core lexicon shared by the two groups, the vocabulaire des isles “vocabulary of the islands.” He adds that this points to administrative and commercial links between the two zones of the French colonial empire. Many sailors, traders, public officials, and some colonists who landed at Bourbon Island (present-day Reunion) and Isle de France (present-day Mauritius) had gained their first colonial experience in the Antilles. Additionally, one has to reckon with a certain degree of diffusion, at least within each zone. In a long-standing disagreement with Baker (1987), who claimed an independent genesis of Isle de France Creole, Chaudenson presents both demographic and linguistic evidence in support of his notion of at least partial diffusion: In terms of demographics, he cites the sending of experienced slaves from earlier developed Bourbon to jumpstart a fledgling plantation economy and, in terms of linguistics, he notes the uniquely shared structural features—for example, the incipient pluralizer bane (from French une bande de “a lot”). The transfer of technical expertise from older to newer colonies served to shorten the habitation stage and, because this transfer was accompanied by the importation of expert personnel who spoke the creole formed in the older colony, that variety influenced at least in part the endogenous variety that would develop in the newer territory. Thus, Mauritian Creole, the heir of Isle de France Creole, constitutes a second-generation creole. The varieties spoken in the Seychelles and Rodrigues Island based on Mauritian Creole are third-generation creoles. In the case of the islet of Saint-Barthélemy (St.-Bart), the local creole constitutes a second-generation one imported from Guadeloupe and Martinique or a third-generation one if it can be demonstrated that Guadeloupean and Martinican Creoles owe a considerable part of their structure to a variety imported from Saint-Kitts, the first French plantocratic colony in the Antilles.⁹

The crucial factor in Chaudenson's sociohistorical model is demographic: the relative proportion of speakers of the TL and alloglots. More specifically, the higher the proportion of the former, the less likely is the emergence of a creole. For Chaudenson, this explains the absence of Spanish-based creoles in Cuba and the Dominican Republic. Although plantation economies flourished there, they were instituted much later than in the British and French insular New World colonies, after several centuries of the existence of fully integrated Spanish-speaking communities. In his preface, Mufwene supports this view with an account for the absence of a Portuguese-based creole in Brazil explained by differences in the nature of sugar plantations, despite the large demographic African influx. Unlike the large-scale agroindustrial establishment of the West Indies, smaller Brazilian plantations remained at the habitation stage during the formative period of the colony.

Chaudenson's sociohistorical model suggests that the plantocratic colonies were monolithic entities. However, the history of the establishment of a French colony in Louisiana suggests a more complex picture. As Klingler (2003) showed, that territory must be viewed as consisting of several subunits showing different social structures and, as a corollary, different mixes of alloglots and speakers of the TL. Although Louisiana remained essentially a societé d'habitation throughout French rule, in the first decades of the colony, a few large-scale plantations coexisted with small freeholds. On the other hand, Chaudenson's model gains support from the relative freedom of movement of plantation slaves, many of whom engaged in petty trade in New Orleans and thus interacted with TL speakers. This fact also suggests that the model needs to take into account the effect of the intermingling of slaves, freedmen, and Europeans (especially the transient sailors and soldiers) in the urban centers essential to the infrastructure of the societé de plantation.

THE NATURE OF TARGET INPUT: VERNACULAR FRENCH

A major weakness shared by creolistics and SLA studies consists of the lack of empirical research on the exact nature of the input available to the acquirers; as a result of this lack, researchers often arbitrarily select the standard variety of the TL as the target. As Chaudenson's exhaustive review of the morphosyntax of FBCs demonstrates, the structural distance between SF and vernacular and dialectal varieties of the language is considerable—in some instances, greater than between FBCs and vernacular varieties. In the case of FBCs, Chaudenson accounts for the reductionist choice of SF by the lack of knowledge of vernacular, everyday varieties of the language on the part of foreign creolists. That is undoubtedly partially true, but the problem is rooted in large part on the relative paucity of studies on nonstandard varieties as well as on the shortcut taken by creolists in the identification of both the superstrate and the substrate languages in creole genesis. Indeed, native speakers (NSs) of French are not exempt from this reductionist shortcut. On several occasions Chaudenson expresses his profound surprise at the choice of SF as the input in the development of Haitian Creole on the part of Lefebvre (1998), who is not only a NS of Quebec French (a variety containing more nonstandard features than metropolitan vernacular varieties) but who, more importantly, has conducted pioneering empirical studies on the vernacular form of her own dialect.

Rigorous research on the termimus a quo of FBCs represents a signal contribution of Chaudenson's to creolistics and permeates La créolisation. His contention—substantiated by empirical studies (Calvet & Chaudenson, 1998; Chaudenson, Mougeon, & Beniak, 1993)—is that it can be reconstructed from the numerous available descriptions of diaspora varieties of French in the Americas and of French regional speech and local dialects. Of particular interest are those in surviving isolates that serve as reservoirs of features lost in current varieties exposed to the pressures from SF (Valdman, 1979, 2005a). The comparison of Mauritian Creole mo pa pou dansé “I will not dance” to SF je ne danserai pas leads one to the conclusion that creolization led to drastic restructuring involving a shift from the use of inflectional morphology to the use of function words and changes in the position of the semantically salient negator pas. The difference becomes less striking if the basis of comparison chosen is Quebec French moi, ch'us pas pour danser. Reference to the classic work on the periphrastic expression of tense and aspect in vernacular and dialectal varieties of French (Gougenheim, 1929) indicates that the verbal system of FBCs, which consists of a bare stem and preposed verb markers, derives from a wide variety of periphrastic constructions all of which occur in various overseas French dialects. In fact, some of these constructions are still attested in regional varieties in France, as is illustrated by a comparison of semantically matching forms of Haitian Creole and vernacular French that express progressive and future events, respectively: m ap (pe) danse and je suis après danser “I'm dancing;” m a (ava, va) danse and je vas (vais) danser “I'll dance.” Chaudenson points out that this periphrastic future construction occurs even in a literary 17th-century text: Rome s'en va renaître “Rome will be reborn” by Corneille. Its French origin is reflected by its presence in all FBCs. In the nominal system, substratists claim an African origin for the postposition of the definite determiner la because it matches semantically equivalent forms found in Kwa languages of West Africa (Lefebvre, 1998; Sylvain, 1936/1979). All vernacular varieties of French—especially those of North America—exhibit the use of the deictic adverb -là “there” postposed to noun phrases containing the preposed definite determiner for emphasis: l'oiseau là “the bird.” However, a structure identical to that of FBCs has been observed in the French of the Old Mines (Missouri) dialect (Carrière, 1937; Valdman, 2005a) as a variant of the SF construction dans tas d'pierre-là “in the pile of stones:” m'as casser ta tête avec baton-là “I'll break your head with the stick” versus l'p'tsit sifflet “the little whistle.”

In addition to his broad knowledge of vernacular and dialectal varieties of French, Chaudenson brings to the identification of the terminus a quo of FBCs his fieldwork experience in St.-Bart (Calvet & Chaudenson, 1998). The linguistic situation of that tiny island has long remained an enigma. It holds the distinction of being one of the rare islands of the West Indies to be inhabited by a nearly exclusively White local population whose roots go back to the middle of the 17th century. The enigmatic aspect of the linguistic situation of St. Bart is that part of the population speaks a creole and the other a French dialect. Chaudenson explains this surprising linguistic division with reference to conditions on St.-Bart: Because the island did not provide the suitable ecological conditions for the establishment of a plantation economy, it remained at the habitation stage except for an attempt to introduce cultivation of cash crops on the windward part of the island in the late 18th and early 19th centuries. Slaves were imported in that part of the island but left after emancipation. The White population of the windward area shifted to creole and, because of a high level of endogamy, it has maintained its linguistic distinctiveness.

The linguistic situation of St.-Bart informs research on FBCs in two important regards. First, of all the American vernacular varieties of French, the leeward St.-Bart dialect best reflects the colonial koinè that served as the TL for imported slaves; however, Chaudenson cautions that this dialect should not be viewed as a relic form of that koinè. Second, the total shift to the imported creole on the part of the windward population suggests that this language had become its exclusive means of communication and the symbol of its identity. There is considerable evidence of that role of FBCs in other former French colonies. In Louisiana, a significant proportion of Whites uses Louisiana Creole rather than Cajun French (Klingler, 2003), particularly in regions where poor Whites and Black slaves worked side by side. Additionally, as Chaudenson points out, White children on a plantation were placed in the care of slave nannies and interacted with Black playmates (for details on the Saint-Domingue and Louisiana situations, see Valdman, 2005b). In a sense, after their stabilization, FBCs became the lingua franca of French colonies. This situation reduced the servile population's access to French: On the one hand, it allowed for the internal evolution of the particular creole, but, on the other, it introduced a wide range of variation reflected today by the distinction between mesolectal and basilectal forms and the presence of relic forms reflecting earlier stages of the creole (Fattier, 2000; Valdman, 2004).

One aspect of the koneized colonial French that is not clearly delineated in La créolisation is whether the koneization process took place in the colonies or in France. An issue that divides specialists in North American French, especially Quebec French, and one on which Chaudenson touches somewhat (p. 187), is whether the relatively uniform variety of French that emerged in New France (present-day Quebec Province) derives from the contact between various dialects of northern France (Barbaud, 1984) or from a koinè not very divergent from the vernacular varieties used in urban centers and the harbors from which settlers embarked. In a recent article, Canac-Marquis and Poirier (2005) updated a notion first put forward by Faine (1939). After obtaining a description of Mauritian Creole, this self-learned Haitian linguist abandoned his initial theory of a Picard origin for Haitian Creole and posited the existence of a nautical dialect transported by settlers of the Indian Ocean and of the American plantocratic colonies. His main evidence was a large set of what he took to be nautical terms whose meaning was extended—for example, amarrer “to tie up a boat” > “to tie;” haler “to move a boat forward by pulling” > “to pull.” This theory was recast more narrowly by Hull (1979), who invoked a “maritime” French he located in the Atlantic harbors of Nantes and La Rochelle. Canac-Marquis and Poirier retraced the influence of French seafarers in the denomination of American fauna and flora beginning in the late 15th century and in the constitution of what Chaudenson, inspired by Arveiller (1963), calls the vocabulaire des isles, a lexicon containing many loanwords gleaned along the coasts of the Americas from Brazil to Newfoundland (e.g., the naming of mosquitoes maringouin from Tupi and moutique or moustique from Spanish). However, although they detailed the contribution of fishermen and sailors to the formation of the initial lexicon of overseas French varieties, they stressed that the phonological, morphosyntactic, and lexical features of their everyday speech were not restricted to seafarers but, instead, reflected the usage of ordinary speakers and constituted a precolonial koinè that they label français populaire véhiculaire (i.e., the everyday speech of ordinary folks).

SUBSTRATAL TRANSFER

Rightly, Chaudenson decries the inappropriate meaning assigned to the antonymous pair substrate/superstrate that creolistics adapted from the field of historical linguistics. In its original sense, the term substratum referred to the language of local populations undergoing language shift toward a language imported by a socially and politically dominant group. The superstratum was a newly imported language whose speakers, although politically dominant, were acquiring the previously imported language modified by speakers of the substratum. Starting with the base of Gallo-Roman—an approximation to Vulgar Latin initially acquired by the urbanized Gallic elite and then by all Gauls who interacted with them and Roman tradesmen and soldiers—Romanists refer to transfers from the local Gallic tongues that perdure as substratum items but changes induced from the languages of later arriving Germanic tribes that gained political hegemony as superstratum influence. From that perspective, in the formation of FBCs, there existed no substrate languages because, in the IO, the French colonists found uninhabited islands and, in the ACZ, the aboriginal populations had been previously eliminated by and large. If anything, it was French, approximative versions of which were being acquired by later-arriving alloglots, that constituted the substratum, and the African languages introduced massively after the formative period of the colonies that constituted the superstratum.¹⁰

Like most of his so-called superstratist colleagues, Chaudenson does not deny the influence of substrate language, but he poses severe constraints on explicatory resort to them. First, in castigating the appeal by many substratists to the “cafeteria” principle (which consists of attempting to derive arbitrarily particular features of creoles distinct from the putative TL input from any substrate language that shows them), he insists on a clear identification of a specific substrate language on the basis of sociohistorical data.¹¹

In the case of heterogeneous servile populations, it must be further shown that speakers of the particular substrate language selected were either demographically or culturally dominant among their peers. For the FBCs of the ACZ, that is a challenging task given the ambiguity of the term nation used in accounts from the colonial period to describe slave groups. Chaudenson contends that such labels as Arada, assigned to speakers of Kwa languages, referred neither to ethnic nor to linguistic groups but instead to the region of embarkation in Africa. Second, transfer should be invoked only after attempts to relate apparently innovative features to the TL have failed. For instance, Taylor and Keller (1963) derived the Haitian Creole word moun [mun] “person” from mun-tu “human being,” a word in Kikongo, a Bantu language. A more plausible origin is the French word monde “people,” pronounced [mõn], appearing in isolation or in compounds (petit monde “child,” grand monde “adult”) in a wide variety of French dialects, particularly those in North America. Third, at least for the FBCs (because they are found in two geographically distant zones and have developed with different substrate inputs) comparison must be performed across the two groups of creole (for a wide variety of West African languages in the ACZ and chiefly Malagasy in the IO) in order to discount the influence of the shared TL or common restructuring processes.

For Chaudenson, transfer from the substrates is most likely at the phonological level and at areas of the morphosyntax where the TL and the substrates converge. In a phrasing reminiscent of Andersen's (1990, p. 62) transfer to somewhere principle, he states that morphosyntactic transfers cannot force their way into interlinguistic systems; the TL must contain some homologous feature. At the phonological level, the application of the third constraint mentioned previously—namely comparison between ACZ and IO creoles—identifies the only uncontroversial case of substrate transfer: the replacement of palatals by corresponding sibilants (manger “to eat” > manzé; changer “to change” > sanzé) in the IO attributable to transfer from Malagasy. However, Chaudenson points out that this opposition is one of the unstable areas of the French phonological system, as attested by alternations between members of the two sets of phonemes in vernacular and dialectal varieties (e.g., chaze-femme “midwife” alternating with sage-femme). However, he fails to consider prosodics and intonation that constitute levels of phonological structure where extensive transfer is likely. Casual observation of the speech of English- or French-speaking West Indians shows striking similarities in these two central areas of phonology that remain inadequately explored in all creoles.

Substrate influence is relatively minor at the lexical level. Chaudenson indicates that substrate transfer accounts for about only 7% of the lexicon of Reunion Creole. A sampling of Haitian Creole shows an even lower proportion, about 4%. Calquing is one lexicogenetic process where substrate influence is likely, but Chaudenson provides several examples of the failure of substratists to look at the TL carefully enough before claiming transfer. For example, Parkvall (2000) attributed compounds such as caca nez “snot” (nose dirt) and caca yeux “rheum” (eye dirt) to African languages, yet these also occur in the IO. The choice of doigt de pied “toe” (finger of the foot) and trou de nez “nostril” (hole of the nose) is even less felicitous given that these are widely used in vernacular French.

By applying his constraints at the morphosyntactic level, Chaudenson identifies several cases of transfer in the pronominal system. Haitian Creole is the only FBC that has merged the 1st- and 2nd-person plural pronouns in the direction of the former: nou [nũ] from French nous [nu]. The presence of the 2nd-person unu in English-based Jamaican Creole points to an African transfer. He cautiously suggests that the merger of the 2nd- and 3rd-person pronouns to zot (from French vous-autres “y'all” [2nd person plural] and eux-autres “they”) in the IO is facilitated by phonological similarities between corresponding forms in Malagasy: 2nd-person ianareo, anareo versus 3rd-person izi ireo, azy ireo.

There are numerous cases at the morphosyntactic level of both broad and specific convergences between the TL and substrate languages. For the former type, in the analytical expression of verbal time and aspect by combinations of preposed independent forms and invariables, the TL (the colonial vernacular koinè, not present-day SF) and African languages concord, as do autoregulatory and SLA processes. An interesting case of convergence is provided by the interrogative adverb kofè “why.” Harris (1973, p. 100) traced the semantic model for that form to the combinations “why” + “make” of Ibo or Yoruba, which is also reflected in other Atlantic English-based creoles: wa mek in Jamaican Creole and mek in Gullah. However, he also found the combination quoi “why” + faire “make” in the western French dialect Saintongeais and in Cajun French: Quoi faire vous dit ça? “Why do you say that?,” but he decides on a transfer from these African languages. Whereas substratists are prompt to stop looking as soon as they find a substrate influence in such cases, Chaudenson and other superstratists would dig further into sources related to the TL. In the final analysis, Chaudenson considers the structure of substrate languages as serving as a filtering device that focused the attention of alloglots on certain features of the TL. For example, the fact that their L1s express verbal tense and aspect with preposed free forms led African alloglots to select French auxiliaries and modals rather than highly diverse inflected forms, which also occurred in the TL.

CREOLIZATON PROCESSES

Autoregulation

For Chaudenson, the term autoregulation subsumes a set of processes that lead to linguistic variants that are the precursors of diachronic change. Variation does not affect all features of French but only certain sensitive zones whose aggregate he labels français zéro. As it were, variation is swept under the rug by such institutions as the school system and the French Academy, whose role is to institute a uniform norm, particularly in the written mode. He acknowledges a debt to Frei (1929/1971), whose classic description of variable tendencies in vernacular, lower-class French (aptly entitled La grammaire des fautes) identified five processes triggering variation in spoken usage: assimilation, differentiation leading to greater clarity, economy, invariability or regularization, and expressivity.¹³

Frei's data were, in a sense, indirect. They were abstracted from letters written during World War I by French prisoners of war and their family members that transited via the Swiss Red Cross.

In the case of French, autoregulation is most clearly observable in varieties exported to the 17th- and 18th-century overseas colonies. Separated from metropolitan institutions that reinforced exposure to SF and not sufficiently evolved to develop a cultural infrastructure that would relay the standard norm, overseas colonies provided an optimal context for the unfettered operation of autoregulation.

As discussed in an earlier section entitled The Nature of Target Input: Vernacular French, settlers in the overseas colonies used vernacular varieties that showed a certain degree of uniformity. However, these contained variants emanating from regional dialects of northern (oïl) French. For example, dialects in the northern and western regions of France from which most settlers originated show two different present tense systems. In the first and most widely distributed, verbal person is expressed by subject pronouns, except for the 2nd-person plural. In the second system, characteristic of the regions bordering the Atlantic Ocean and best reflected in the Acadian dialects of Nova Scotia and New Brunswick, the opposition of singular versus plural in the 3rd person tends to be expressed by inflection. Compare the present indicative paradigm for chanter “to sing,” shown in (1).

Alternations between these two systems still exist in Acadia and Louisiana.¹⁴

Some of those French-speaking regions were settled by Acadians, but it is the more regularized System I that is more extensively distributed. In fact, it is not uncommon to find extensive free variation between the two systems in certain areas (Rottet, 2001).

Differences of that sort tended to be leveled out and the vernacular colonial French that served as input in the process of creolization constituted a koinè.

In his survey of the morphosyntax of FBCs, Chaudenson provides a variety of examples of autoregulatory processes that operate in the français zéro zones that he uses to counter proposals of substrate transfer. The evolution of the demonstrative determiners illustrates the operation of processes leading to invariability and salience. First, invariability led to the elimination of the distinction between near and remote deixis, expressed in SF by the contrast between -ci and -là: ce livre-ci “this book” versus ce livre-là “that book.” Chaudenson notes a universal tendency to eliminate the distinction in favor of the form expressing remoteness. Indeed, in American varieties of French, -là tends to eliminate -ci, and in dialects such as Acadian that preserve the distinction, nearness is marked by a more salient form, isit. Second, the preposed demonstrative determiner inflected for gender and number in SF is replaced in vernacular varieties by an invariable phonologically reduced form: ce garçon versus c'te garçon “this boy,” and cette fille versus c'te fille “this girl” (SF and vernacular variants, respectively). The postposed -là, optional in SF, becomes obligatory and carries the deictic force; in some varieties—for example Cajun French—it alternates with or is even reinforced by a second deictic form derived from the indefinite pronoun ça “this, that:” cet argent-ça “this money,” dans ces temps-ça-là “in those days.” Interestingly, the same pattern occurs in child L1 acquisition: ce so(l)dat cassé là “that broken soldier,” le ga(r)son-ça “this boy” (Grégoire, 1937). The fact that postposed demonstratives occur in Kwa languages (e.g., Ewe's -sia singular and -si-a-wo plural cited by Sylvain, 1936/1979) attests at best to a convergence between the input and the substrate, particularly given that the form -sa-la occurs in both the ACZ and the IO; for the latter, any direct influence from West African languages is extremely doubtful.

Chaudenson states that autoregulation operated on French chiefly at the habitation period (p. 182).¹⁵

La phase essentielle de l'autorégulation est celle qui s'exerce dans la période d'habitation et elle concerne le français lui-même.

This is an assertion whose validation by empirical evidence would be problematic, for it is most unlikely that chroniclers who incorporated samples of the nascent and established FBCs would deem it interesting to note down samples of French speech they considered deviant relative to SF. Vernacular variant features were so abundant that they were beyond their perceptive threshold. Because support for autoregulation is provided mainly from French dialects and vernacular American varieties, most of which did not become acquisitional targets for servile alloglots, it would appear that that complex of processes also operated earlier both in France and in the individual colonies before the importation of alloglots.