Upload
others
View
9
Download
0
Embed Size (px)
Citation preview
Creole Formation and Second Language Acquisition.
Table of Contents:
1. Introduction:....................................................................................................................22. Developmental stages in SLA.........................................................................................3
2.1. Early stages of SLA and creole formation................................................................42.2. Strategies in the early stages of SLA and creole formation......................................5
3. Elaborative stages of SLA and creole formation...........................................................104. The elaboration of creole grammar................................................................................11
4.1. Restructuring, target language and superstrate input..............................................124.2. Restructuring and substrate input in creole formation............................................144.3. Restructuring and internal developments...............................................................15
5. TMA and the process of restructuring in creole formation...........................................155.1. The emergence of the Haitian Creole TMA system...............................................165.2. The emergence of TMA in Sranan Tongo..............................................................20
6. Conclusion.....................................................................................................................24References..........................................................................................................................26
1
1. Introduction:
The parallels between creole formation and SLA were noted as early as the nineteenth
century by scholars like Hesseling (1897), and later Jespersen (1922). This relationship
has been the subject of continuing controversy in the field of creole studies. In particular,
there has been disagreement about whether creole formation is the outcome of first or
second language acquisition – an issue explored in Andersen (1983) and more recently in
DeGraff (1999a). Disagreement over this has wider implications for competing theories
of creole formation. Contemporary scholars remain divided over the relative contribution
of superstrate and substrate languages as well as the role of language universals in creole
formation.
Some still adhere to Bickerton’s LBH or some version of it that ascribes the primary
role in creole creation to children who appeal to innate universal principles to compensate
for deficient (pidgin) input to the L1 acquisitions process (Bickerton 1999). Most
creolists, however, maintain that creole formation was essentially a process of second
language acquisition in which adults and, quite probably, bilingual children played a
crucial role. Adherents of this view, however, still disagree on the nature of the earliest
forms of creoles. Some argue that creoles began as second language varieties of the
lexifier or “superstrate” languages and gradually diverged more and more from the latter
via a process of “basilectalization” (Mufwene 1996a, b). This is the so-called superstratist
position first proposed by Chaudenson (1992, 2001). Others adopt the traditional view
that creoles began as pidgins that were subsequently elaborated.
Not surprisingly, scholars in the two camps disagree on the extent of superstrate vs
substrate input to this process of elaboration. The superstratists maintain that most of
creole grammar can be traced to the lexifier language. Some of these, while
acknowledging that some creoles draw heavily on superstrate sources, still allow for
significant influence from substrate languages (Mufwene 1990). The “substratists” on the
other hand claim that the major influence on the grammar of “radical” creoles in fact
came from the substrate languages (Lefebvre & Lumsden 1994; Lefebvre 1996; Lumsden
1999, etc.).
2
Despite these differences, there is consensus that creole formation involved varying
degrees of “input” from both superstrate and substrate sources, and was guided by
principles that regulate all cases of language contact (Mufwene 1990; DeGraff 1999b).
More specifically, there is now wide agreement that creole formation was akin in many
respects to a gradual process of group second language acquisition. The issue then is
whether we can maintain this view and still reconcile the conflicting positions outlined
above.
The present paper attempts such a reconciliation by examining more closely the
similarities in the developmental stages, processes and principles that apply to all
instances of creole formation and (other) cases of second language acquisition (SLA). I
will argue that there are basic similarities in the paths of development characteristic of
both. First, both involve an initial or early stage of learning, in which a highly simplified
interlanguage (IL) system is created. This is followed by elaborative stages in which the
basic IL system is expanded, drawing on three major sources of input. These include
input (intake) from native and non-native varieties of the lexifier language, L1 influence,
and internally driven changes that regularize and expand the grammar.
The interaction between L1 knowledge, intake from superstrate sources and creative
adaptation operates within the developing IL system itself – or more accurately, within
the minds of individual learners creating IL systems or I-languages. This is not to claim,
however, that the parallels between the two broad types of SLA are identical. As we shall
see, there are significant differences in such aspects as the nature of the input, the extent
of L1 influence and the degree of internal innovation involved in each case, which help to
explain how creole formation differs from other cases of natural group SLA. Exploring
the precise nature of the similarities and differences between these two kinds of SLA
promises to enrich our understanding of both.
After presenting a broad outline of the stages and processes involved in SLA
(including creole formation), I will turn my attention to the creation of creole TMA
systems, focusing particularly on those of Haitian Creole and Sranan Tongo.
2. Developmental stages in SLA.
3
Studies of second language acquisition have revealed that there are several stages
through which learners go in their attempt to approximate the TL. The first stage involves
the construction of a relatively simple interlanguage system that is highly reduced by
comparison with the TL grammar. For instance, English-speaking children who are
learning German begin with two-word utterances before producing multiword sentences.
The overall pattern of development in this case is from two-word utterances to copular
sentences to sentences containing auxiliaries and finally main verbs (Felix 1977). SLA
research has also shown that learners go through various developmental stages in their
acquisition of specific areas of TL grammar such as question formation, negation,
relativization, etc. For example, Spanish-speaking learners of English first use no as an
all-purpose negator in all constructions, and later acquire invariant don’t, aux-neg forms
such as isn’t, can’t etc, and finally analyzed do + not (doesn’t, didn’t, etc.) (Schumann
1978:13). In general, in acquiring various aspects of TL grammar, learners first apply a
single invariant rule across the board in the first stage, and gradually acquire more
specific rules in later stages of acquisition. Similar stages of development apply to the
acquisition of TL tense/aspect systems, as we will see.
2.1. Early stages of SLA and creole formation.
Studies of the acquisition of various European languages by immigrants with a
variety of L1’s have demonstrated that learners first create a “basic variety” of the TL
that is quite uniform in structure regardless of L1 background (Klein & Perdue 1997).
This basic variety is characterized by a small but expanding lexicon made up mostly of
nouns and verbs, with a small inventory of adjectives and adverbs. It also employed a few
function words such as quantifiers, a few prepositions and determiners and a single
negative marker. In these and several other respects, the “basic variety’ shares many
characteristics with “prototypical” pidgins.
The conventional wisdom has it that creoles are elaborations of pidgins – the so-
called “two-stage” view of creole formation. As we saw earlier, this view has been called
into question of late, e.g., by Chaudenson and others who argue that second language
varieties of French etc, were the starting point of creole formation, at least in the French
4
colonies. We can reconcile these opposing viewpoints by making a distinction between
the kinds of input that are available from the putative TL, and the kinds of intake learners
incorporate into their developing IL. Whether the input to the first stages of creole
formation consisted of close L2 approximations to the superstrate language, or a
simplified or pidginized variety of that language, individual IL construction would still
begin with a basic variety that has pidgin-like characteristics.
It seems reasonable to assume, then, that in the first stage of both SLA and creole
formation, individual learners create a highly reduced, pidgin-like system, which they
then expand, depending on their access to further input from the TL and other sources, as
well as their motivation to create a more complex system. The I-grammars that
individuals create must be the starting point of our analysis of the processes of both SLA
and creole formation.
2.2. Strategies in the early stages of SLA and creole formation.
It is well known that, in the earliest stages of SLA, learners attempt to learn and
produce TL structures by appealing to various learning and communication strategies.
The former may include, for instance, memorizing, guessing, comparing L1 and L2
elements etc. We will not be concerned with these here. Communication strategies
include avoidance (avoiding certain structures, elements or topics) and compensatory
strategies. The latter include appeal to L1 knowledge, creative adaptation of existing IL
resources, and non-linguistic strategies such as gesture and mime (Poulisse 1996).
Studies of SLA (e.g., Poulisse 1996:149) have revealed that such communication
strategies are more common among early (less proficient) than advanced learners, though
of course not restricted only to the former.
Following Meisel (1977, 1983), simplification will be used here to refer to two kinds
of process – reduction of TL structures (reductive simplification) and strategies aimed at
regularization of the grammar (elaborative simplification). The former is a strategy of
avoidance that is particularly common in the earliest stages of IL construction. Its well-
known consequences include the elimination of TL morphology and the reduction of TL
syntactic strategies, among others.
5
Elaborative simplification on the other hand is a compensatory strategy that relies on
the available resources of the IL. For example, learners may compensate for loss of
morphology by employing periphrastic means instead. They may use adverbs to convey
temporal or aspectual meanings, or fixed word order to distinguish grammatical functions
such as subject and object. Another such strategy is rule generalization to eliminate
irregularities, as in the extension of past tense suffix –ed to irregular verbs like steal and
tell in L2 English.
Both kinds of simplification seem to be motivated by the need for transparency in the
emerging IL grammar. This in turn is related to general cognitive (processing) principles
that guide the acquisition process. For instance, Van Patten (1996:14-15) suggests that
the following principles (among others) determine which aspects of TL input learners are
likely to process earlier:
Learners process input for meaning before they process it for form;
Learners process content words first;
Learners tend to process lexical items before grammatical items for semantic
information.
Principles like these help explain why early IL systems include mostly lexical rather
than function morphemes and lack bound morphology. They constrain the amount of
input that actually makes its way into learner versions of the TL. This modified input, or
intake, becomes the primary material for restructuring of IL grammar.
Similar cognitive principles lead early learners to regularize IL grammar via
elaborative simplification. Among the principles that have been suggested in this
connection are the following:
The uniqueness principle (one form expresses one meaning);
The principle of canonical word order (Main clause word order constitutes the
basic word order. (Jordens 1996:32).
The principle of continuity (constituents that belong together are placed together).
6
Such principles allow learners to maximize ease of perception and production, and
help explain many of the characteristics of early IL. Andersen (1984, 1990) suggests that
features of early IL such as invariant word order and use of invariant negative markers
etc. can be explained in terms of his “One to One Principle”, which condenses the first
two principles stated above. The same kinds of principle have been proposed for pidgin
and creole formation, which is not surprising, given the strong similarities between these
and the earlier stages of SLA.
The well-known characteristics of pidgins include reduced vocabulary, absence of
bound morphology, a limited range of syntactic structures etc. These are precisely the
characteristics found in early IL systems, for instance the “basic varieties” of a host
community TL that immigrants create (Klein & Purdue 1997). In addition, pidgins share
strategies of L1 transfer and internal innovations found in early IL. It follows that the
creation and expansion of pidgin systems may shed light on the ways in which early IL
and particularly early creoles develop. A particularly relevant aspect of this is the creative
innovation that is a key characteristic of pidgin grammar. Like L2 learners, pidgin
speakers creatively adapt their limited intake from the source language(s) to achieve
communication. For instance, they expand the lexicon through compounding, paraphrase
and other strategies, and exploit various means (adverbials or other function morphemes)
to convey temporal and aspectual meanings.
Innovations in pidgin grammar are often reminiscent of the incipient patterns of
grammaticization that are quite common in language change generally, and in SLA in
particular. For example, Kotsinas (1996:133) compares uses of stannom ‘stay’ as a
locative copula in Russenorsk (RN)and immigrant L2 Swedish (IS). The following
examples illustrate:
(1) a. RN kor yu stannom på gammel ras
where you stay on old time
“Where were you last time?”
b. IS den tjugo år stanna Joannina
it twenty year stay Joaninna.
7
“She lived in Joannina [a town] for twenty years”
This is reminiscent of the use of stay as a copula in Hawaii Creole English. Pidgins
also employ a single preposition in a variety of functions, a strategy also found in early
SLA and creole formation. For example, both Russenorsk and Immigrant Swedish
employ the preposition på to mark various spatial meanings such as location, direction
and origin (Kotsinas (1996:139). På is also used in both contact varieties to mark indirect
objects and possession, as in the following examples (Kotsinas (1996:141-42):
(2) a. RN moja paa ju presentom baanbaan
I P you give candy
“I will give you candy”
b. IS Köpa på barn
buy P child
“[I] bought [clothes] for [my] child”
(3) a. RN mangeli klokka på ju?
how-much clock P you
“What time is it?” [Lit. how much is your clock?]
b. IS Stan på din mamma
town P your mother
“Your mother’s village”
In other uses, på seems to have the potential to function either as a preverbal marker
of some kind of desiderative mood (example 5), or as a complementizer (example 6) in
both RN and IS. (Kotsinas 1996:144-45)
(4) a RN Moja på-slagom på tvoja
I P hit P you
8
“I will hit you”
b. IS den barn sex månar kommer på skriva på kyrka.
it child six month come P write P church
“The child became six months old and we had to register him at church”
(5) a. RN gå på slipom
go P sleep
“Go to sleep”
b. IS och gå vi på simma och åta
and go we P swim and eat
“And we went to swim and eat.”
Though på has not been grammaticalized as either an auxiliary or a purpose
complementizer in Russenorsk or Immigrant Swedish, its similarity to other all-purpose
prepositions such as long (< along(a)) in Tok Pisin and fu (< for) in Atlantic creoles is
striking. It is precisely these kinds of potential for restructuring based on the internal
resources of the developing IL system that we find in creole formation, and to some
extent in natural SLA.
So far I have argued that the starting points of pidgin formation, early IL and creole
creation are quite similar, if we view them from the perspective of individual I-language
construction. Each has the potential to develop into a more elaborate system, but the
nature of that elaboration depends crucially on the nature and availability of continuing
input from the TL and other sources. Most pidgins never develop further, because of lack
of motivation or lack of access to more input. But some pidgins such as Hawaii Pidgin
English and Melanesian Pidgin have developed into elaborate systems, via processes
quite similar to those involved in creole formation. The restructuring process in these
cases involves far more appeal to L1 knowledge and internally-driven innovation than is
found in the more usual cases of SLA. Let us now examine the nature of this
restructuring process with particular attention to the origins of creole TMA systems.
9
3. Elaborative stages of SLA and creole formation.
In SLA, learners progress beyond the basic variety by adding more morphological
apparatus, grammatical rules, vocabulary etc. The major source of these additions is the
TL, but learners also continue to appeal to their L1 knowledge and use the resources of
the IL system itself in expanding their grammar. These contributory factors remain active
throughout the acquisition process, though with different effects at each stage. As Brown
(1980:163) notes:
“By a gradual process of trial and error and by hypothesis testing, the learner slowly
and tediously succeeds in establishing closer and closer approximations to the system
used by native speakers of the language.”
However, the degree of access to native models of the TL is a crucial factor in
determining how successful learners are. In fact, what really distinguishes most cases of
creole formation from other types of SLA is the nature and accessibility of continuing
input from the superstrate source. The degree of access learners have to such sources is in
inverse proportion to the extent to which they appeal to their L1 and to creative
innovation in the expansion of their IL system.
Differences in the degree and nature of superstrate input correspond to differences in
the ecology of the contact situations, including community settings, patterns of
interaction among groups, demographic ratios among groups, etc. We will not pursue
these further here. But such factors make for significant differences among the outcomes
that have traditionally been referred to as creoles. Some of these, like Bajan or Reunion
Creole, are quite close approximations to their respective lexifier languages. Others
diverge from their lexifiers to varying degrees, some quite radically, because they were
created under conditions of decreasing input from, and accessibility to, native varieties of
the lexifier language. I discuss this further below.
10
4. The elaboration of creole grammar.
Scholars have traditionally referred to the elaborative stages of creole formation as
“creolization”, which Hymes (1971:84) described as “that complex process of
sociolinguistic change comprising expansion in inner form, with convergence, in the
context of extension in use.” Unfortunately, researchers have used the term “creolization”
in so many different senses that its usefulness is now in question. Moreover, the term
implies some unitary process of change that is quite at odds with the diversity and
complexities of change involved in creole formation. I will therefore refer to the
elaboration of creole grammar as “restructuring,” in the sense intended by researchers in
first and second language acquisition.
With respect to first language acquisition, van Buren (1996:190) defines restructuring
as “discarding old grammars for new ones.” He adds, “As soon as new relevant data are
encountered, the current grammar is restructured to accommodate the new input”
Referring to SLA, Lalleman (1996:31) defines it as “the process of imposing organization
and structure upon the information that has been acquired” [as new input is encountered –
DW]. Note that this is very different from the sense in which creolists sometimes use the
term, viz, to refer to restructuring of the lexifier language. This implies that creole
creators began with the lexifier, modifying it over time. For the same reason,
restructuring should not be equated with terms like “basilectalization” (Mufwene ???),
which may be more appropriately used to refer to the gradual changes we observe in the
community language over time, than to the processes that lead to such changes.
It is important once more to emphasize that the process of creole formation is both an
individual and a community phenomenon. The restructuring process goes on primarily in
individual learners’ attempts to construct and expand their IL system. The innovations
introduced by these learners then become available for selection as part of the
community’s language. Consequently, creole formation must be seen as a product both of
individual grammar construction (I-language), and of the spread of features across
individual grammars, yielding a shared community vernacular (E-language). Our focus
here is primarily on the first of these, the ways in which individuals create I-languages.
11
Let us first deal briefly with the role of superstrate input in the expansion of creole
grammar, before we examine the role played by substrate influence and internal
innovation in that process. We will also consider the similarities between SLA and creole
formation in the way these factors come into play in the elaboration of learners’ IL
systems.
4.1. Restructuring, target language and superstrate input.
The claim made by Chaudenson and others to the effect that creoles “began” as
second language varieties (i.e., close replicas) of the superstrate and gradually diverged
from them appears to be accurate, if we compare the starting point of the contact situation
with its eventual outcome. But, from the perspective of the act of creole creation itself,
this claim appears to be somewhat misleading. If, as we argued earlier, every individual’s
acquisition of a TL begins with a highly reduced system – a basic variety, then it makes
more sense to say that such reduced systems were the true starting point of creole
formation, just as they are for SLA. The nature of the continuing input in each case would
then determine the nature of the outcome. Moreover, one has to distinguish the available
input from the kinds of intake individual learners incorporate into their developing IL
systems.
Distinctions such as these may help resolve some of the controversy regarding the
role of the superstrate in creole formation. For instance, the long standing controversy
concerning the true “target” in creole formation stems in part from the tendency to equate
“target” only with the superstrate, in a way analogous to more usual SLA. Such a position
may be a quite reasonable one to take for some cases of creole formation. As noted
earlier, in some cases, first and/or second language varieties of the superstrate became
consolidated among a significant portion of the population, and continued to be available
as targets of acquisition. Hence the resulting creoles were closely akin to dialects of the
superstrates, as noted earlier for Barbados and Reunion.
But in other cases, such targets were either not available, or changed drastically over
time. For example, in the case of Hawai‘i Creole English, there appears to be agreement
that the primary input came from Hawai‘i Pidgin English, many of the characteristics of
12
which persist in the creole. In this case, expansion of pidgin into creole involved use of
L1 strategies by learners of Chinese, Portuguese and other languages. Hence there is
ample evidence of substrate influence from these languages on HCE (Siegel 2000). On
the other hand, the evidence from Haitian Creole suggests that many of its features are
modeled on regional French dialects, though various kinds of simplification and
reanalysis have occurred. At the same time, the gradual loss of access to such regional
dialects, and the continuing process of SLA by succeeding generations of Africans in
Haiti, created the conditions for significant substratum influence to affect the evolution of
HC.
The circumstances in which Surinamese creoles like Sranan Tongo arose are
somewhat different from both of the scenarios just discussed, though they resemble the
HCE case more. The very early withdrawal of the vast majority of English-speaking
planters and their slaves within roughly thirty years of the colony’s inception in 1651
meant that the major input to new arrivals from Africa after 1680 came from pidginized
or highly changed second language varieties of English (Migge 1998). Hence, in
elaborating early creole grammar, individual learners had to draw heavily on their L1
knowledge as well as the internal resources of the IL system itself. This in part explains
why the Surinamese creoles diverge so radically from their original English sources. In
short, what distinguishes Haiti from Suriname seems to be the continued availability of
lexifier language models (including close approximations acquired by many Africans) in
the former colony, by contrast with the early withdrawal of such models in Suriname.
For reasons such as these, we must be cautious about the notion that creoles like
Sranan Tongo, or for that matter Haitian Creole, are instances of targeted SLA in the
usual sense of that term. Such a designation implies that the creators of creoles were not
only targeting (native varieties of) French, English, etc., but had adequate access to them.
As Arends (1995), Baker (1990), Singler (1990) and others have argued, such
assumptions are questionable. It would seem instead that most slaves who were
transported to these colonies, especially at the height of the plantation system, were
attempting to learn an already established contact variety quite distinct from the lexifier
languages.
13
In many if not most cases of creole formation, the nature and types of superstrate
input changed over time, as successive waves of new learners created their own L2
versions of existing targets. In such cases, if we were to freeze the contact situation at
different points in time, we would find quite different scenarios, with different targets,
and hence differences in the superstrate-derived input. This presumably is what led Baker
(1990) to argue that, in the formation of many creoles and “expanded” pidgins, the true
target was not the superstrate, but the emergent contact variety itself. From the
perspective taken here, this position is not incompatible with the view that creoles were
always the result of second language acquisition.
4.2. Restructuring and substrate input in creole formation.
The less creole creators could continue to draw on superstrate input as they elaborated
their new language, the more they relied on L1 knowledge. The role of L1 or substrate
influence in creole formation has been convincingly demonstrated in many recent studies.
In the case of the Surinamese creoles, studies by Arends (1986), McWhorter (1992),
Sebba (1987) and Smith (1996) have argued for Kwa, especially Gbe, substrate influence
on serial verb constructions. Research by Bruyn (1994) points to influence from Gbe (and
to some extent Kikongo) on complex prepositional phrases in Sranan. Finally, Migge has
argued for Gbe influence on various Paamaka constructions, including “give”-type
SVC’s (1998), attributive (property) predication (2000) and the copula system (to
appear). There are many similar studies that argue convincingly for significant substrate
input to creole formation.
Strong syntactic parallels like these led Sylvain (1936) to assert that Haitian Creole
was a language with Ewe grammar and French words, thugh that claim has been called
into question (Chaudenson 2002). More recently, Lefebvre, Lumsden and their associates
have argued that most of Haitian Creole grammar derives more or less directly from Gbe
languages such as Fongbe (Lefebvre & Lumsden 1994). Their “Relexification
Hypothesis” of HC genesis has been challenged for (among other things) its failure to
take account of all substrate inputs as well as the significant contribution of provincial
French dialects to the formation of HC (Chaudenson 2002).
14
Still, the evidence presented by studies like these suggests that, in general, creoles do
preserve elements of substrate grammar in varying degrees, though they hardly replicate
such elements exactly.
4.3. Restructuring and internal developments.
The elaboration of early creole grammar, like that of developing interlanguage and
expanding pidgins, involves innovations driven by tendencies already present in the
developing system. For instance, Kouwenberg (1996), while acknowledging Kalabari
(Eastern Ijo) as the source of several aspects of Berbice Dutch grammar, also points to
several others that cannot be attribute to either Ijo or Dutch influence. These include
invariant SVO order, preverbal auxiliaries and negative marker, predicate cleft, and a
serial verb construction in which a verb “say” introduces complement clauses. Migge
(2003) also discusses several aspects of the copula system of Paamaka which appear to be
innovations in the creole, though much the system seems to be modeled on that of the
Gbe substrates.
Innovative features like these appear to arise from processes of internal restructuring
similar to those found in developing IL and in the elaboration of pidgins. In all cases,
speakers exploit intake from both L1 and L2 sources to expand their grammar, and this
often leads to new structures peculiar to the developing I-language. Such tendencies can
be seen at work even in “prototypical” pidgins, as we saw in the case of Russenorsk
earlier. Developments of a similar kind will be discussed below, in relation to the
emergence of TMA and other functional categories in creoles.
The overview of creole formation that we have given so far makes it clear that this
process was a complex one, involving a variety of linguistic inputs and strategies of
restructuring. In the following sections, I examine these strategies in more detail.
5. TMA and the process of restructuring in creole formation.
15
As Bickerton (1988:278) noted, the elimination of inflectional morphology in the
early stages of creole formation results in, among other things, a loss of TMA markers.
Hence these have to be reconstituted in the elaboration of creole grammar. The specific
sources of these TMA markers, and the nature of the processes involved in their
emergence in creole verbal morphosyntax, have long been matters of controversy. Again,
most of the disagreement revolves around the comparative contributions of superstrate
input, substrate influence and creative innovation to the restructuring of creole TMA
systems. There is clear evidence that these three types of contribution varied significantly
from one case of creole creation to another. Such differences can readily be found if we
compare the evolution of the Haitian Creole (HC) TMA system with that of Sranan
Tongo (SN).
5.1. The emergence of the Haitian Creole TMA system.
The major functional categories of the Haitian Creole TMA system are shown in
Table 1, which is based on DeGraff (1999b) and Spears (1990).
Table 1. Haitian Creole TMA categories.
Tense/aspect.
Perfective aspect Unmarked
(Relative) Past te
Prospective Future (a)pral(e)
Progressive/Immediate Future ap
Terminative (Perfect) fin(i)
Modal categories
Possible Future va (a/av/va)
Expectation/likelihood pu
16
Two radically different accounts of the sources of these TMA markers have been
suggested. On the one hand, Lefebvre (1996) argues that they arose through a process of
“relexification” by which substrate categories are relabeled with the phonetic shapes of
superstrate lexical items. Lumsden (1999) revises this somewhat so as to restrict the
process of relexification to the creation of a new vocabulary of lexical categories.
According to Lumsden, some of these lexical categories are subsequently re-interpreted
as functional morphemes via a process of “reanalysis.” For both Lefebvre and Lumsden,
the semantic and syntactic properties of the newly-created functional heads are more or
less directly modeled on substrate TMA categories.
On the other hand, researchers like Chaudenson (1992), Fattier (1998), DeGraff (to
appear) and others have argued that HC TMA markers as well as other functional heads
are derived more or less directly from 17th century French cognates with which they share
semantic and distributional properties. DeGraff (to appear) offers a succinct summary of
the key points of evidence for this view, which is quite compelling. He offers various
comparisons of HC and (earlier) regional French verb structures, which demonstrate
close correspondences between HC TMA markers and elements used in periphrastic
strategies for marking TMA meanings in the French dialects.
A few examples of this will suffice. Note first of all the clear similarities between
Past te and French était in sentences like the following (from DeGraff to appear: 39)
(6) a. HC. Li te (deja) ale
3sg PAST (already) go
“He had (already) gone”
b. FR. Il était (déjà) allé
3sg masc was (already) go (PP)
A similar correspondence is found between te and the French past participle été as
seen in the following (DeGraff to appear, 39-40).
(7) a. HC Li te malad
17
3sg PAST sick
“S/he was/has been sick”
b. FR. Il a été malade
“He has been sick”
Detgers (2000:150), following Chaudenson (1981:206f) suggests that était in French
periphrastic constructions such as il était à écrire “he was writing” was the source of Past
te in French creoles. No doubt all of these inputs complemented each other. It seems clear
that te has its source in the French past imperfect étais/étais, with possible reinforcement
from past participial été.
Similar correspondences can be found between HC modal pu and its French cognate,
the preposition pou “for” and between Future va and French va(s), the present singular
forms of aller, used in the Future construction aller + V “be going to V.” The following
examples from DeGraff (to appear, 40) illustrate.
(8) a. Mwen pou marye semen pwochèn (HC)
1sg. for marry week next
b. Je suis pour me marier la semaine prochaine (Canadian French)
1sg am for me marry the week next
“I am to get married next week”
(9) a. Ou (a)va ale demen (HC)
You FUT go tomorrow
b. Tu vas aller demain.
“You will go tomorrow”
Similar (regional) French cognates can be found for other HC TMA markers. For
example, Progressive marker ap(e) has its source in the preposition après, employed in
the earlier French construction être après à +V “to be V-ing.” HC Prospective (a)pral(e)
can be traced to the progressive construction après (de/à) aller + V “to be going to V”.
Terminative Perfect fin(i) similarly derives from the lexical verb finir.
18
Table 2 summarizes the correspondences between the TMA markers of HC and their
regional French cognates.
Table 2. Sources of main HC TMA markers.
HC catetory HC marker Regional French sources
Perfective Unmarked Infinitival/3rd sing/particle
(Relative) Past te Imperf. était / PP été
Prospective Future (a)pral(e) après (de/à) aller
Progressive/Immediate Future ap être après à +V
Terminative (Perfect) fin(i) finir “finish”
Possible Future va (a/av/va) va(s) + V
Expectation/likelihood pu être pour + V
In fact, as DeGraff (to appear: 39) points out, “Most of the HC functional heads…
have French cognates with which they share substantial distributional and semantic
properties.”
These facts are in keeping with the view that the input to early HC came from first
and second language varieties of regional French dialects that remained available as
models during the first stages of HC formation. As DeGraff (to appear), Chaudenson
(1995), Detgers (2000) and others have argued, the emergence of these TMA categories
can be accounted for in terms of simplification and reanalysis of the French models by
second language learners. These processes and outcomes can be found in other cases
where Africans and other learners acquire French as a second language (Mather 1995;
Prévost & White 2000). All of these situations also allow for a certain degree of L1
influence in the restructuring of the TMA system. As Mather (1995:259) points out:
19
“Once the French periphrastic constructions were stripped of their inflectional
endings by the first generation of creole speakers, they could be reinterpreted as
preverbal TMA markers by adult and children speakers of Kwa languages, who
identified them with their own L1 TMA markers.”
This might allow for some compromise between the strict “superstratist” account of
HC formation offered by Chaudenson (1995) and the quite different “substratist” account
offered by LeFebvre (1998).
We would expect that, in cases where superstrate input is more limited, creole
creators would compensate for this by drawing more heavily on L1 knowledge as well as
the internal resources of their developing IL system. A case in point is Sranan Tongo.
5.2. The emergence of TMA in Sranan Tongo.
Like Haitian, Sranan Tongo employs preverbal free forms to express temporal, aspectual and modal meanings. One exception is the perfect marker kaba, which always occurs in VP-final position. The inventory of the major tense/aspect categories and the forms that express them in SN are shown in Table 3. Note that Potential sa is more of a modal than a marker of just future time reference, but it is included here for purposes of comparison with o, the other marker of futurity.
Table 3. Major Tense/Aspect categories in Sranan (Winford 2000)
Aspect: Perfective ø (the unmarked verb).
Imperfective e
Terminative Perfect VP-final kaba.
Tense: Relative Past. benPredictive Future o
MoodPotential Future sa
20
Obligation musuRoot Possibility kanPhysical ability manPermission mag
Sentences (10 - 15) illustrate the use of each of the tense-aspect categories respectively.
The relevant forms are in boldface.
(11) A djuku wan man boro en here bere.
3sg. stab Art. man cut.open 3sgposs. whole belly.
‘He stabbed a man and cut open his entire belly’
(12) Wan tu fu den pikin fu owma e wroko gron now ooktu?
One two of the-PL child of granny IMP work ground now too
“Are some of granny’s children also cultivating the land now too?”
(13) A alen disi kan stop now. Yongu, a kon tumsi furu kaba, yere.
The rain this can stop now. Man it come too full already, hear
“This rain can stop now. Man, it has already rained more than enough.”
(14) A ben taigi mi a o kon na fesisey baka. Mi no sabi efu a go ete.
he PAST tell me he FUT come LOC front.side back. I NEG know if he go yet
“He told me he would come to the front again. I don’t know if he’s gone yet.”
(15) Efu yu no wroko, dan yu no o nyan, tog.
If you NEG work, then you NEG FUT eat, TAG
“If you don’t work, then you won’t eat, right?”
21
(16) Dan te mi miti en mi sa aksi en.
Then when I meet him I POT ask him
“Then when I meet him I will ask him”
It is clear from the above examples that, unlike the preverbal markers of Haitian
Creole, those in Sranan Tongo have no cognates in any English tense aspect markers.
One possible exception to this is the potential marker sa, which some have claimed to be
a form of English shall. However, it is much more probable that it derives from Dutch
zal. Table 4 provides an overview of the actual sources of the Sranan tense/aspect
markers.
Table 4. Major Tense/Aspect categories in Sranan and their English sources.
Sranan category Marker Source
Perfective Unmarked Bare verb
Imperfective e < de English there
Terminative Perfect kaba (VP-final) Portuguese acabar “finish”
Relative Past. ben Eng. been
Predictive Future o Eng. go
Potential sa Dutch zal
Obligation musu Eng. must (Dutch?)
Root possibility kan Eng. can.
Physical Ability man Eng. man.
Permission mag Dutch mag.
22
This presents a very different picture from that we saw earlier for Haitian creole. In the
first place, there are few, if any distributional or semantic similarities between the SN
markers and their English cognates. Second, two markers, sa and kaba, have been
adopted, not from English, but from Dutch and Portuguese respectively. This can only be
explained in terms of the limited access to and input from, varieties of English among the
Africans who created Sranan Tongo.
These facts suggest, first, that the varieties of English that made up the superstrate
input to the creation of Sranan did not have a TMA system or set of TMA markers that
could serve as a model for those of Sranan. This of course distinguishes SN clearly from
Haitian Creole. As noted earlier, it seems likely that the English input to Sranan consisted
of a pidgin-like variety, similar perhaps to the one that provided the input to Hawai‘i
Creole English. This pidgin, like Russenorsk, Hawai‘i Pidgin English and others (e.g,
early Melanesian Pidgin) had the potential for elaboration of its resources via, for
example, processes of reanalysis of available items. In general, the cases of reanalysis
that created the TMA system involved substantial substrate influence. In addition, Sranan
expanded this system (and other aspects of its grammar) via borrowing (mostly from
Dutch), as well as through internally driven processes of change over time.
Two categories that are clearly due to borrowing from non-English sources include
Potential sa and Perfect kaba. The former appears as early as 1718, and may have been
the only marker of futurity until English go was grammaticized to Future (g)o.
Unfortunately we have no clear proof of this so far. Perfect kaba (which derives from
Portuguese acabar “finish”) is modeled on the semantics and distribution of the verb
“finish” in Gbe languages, which occurs in VP-final position and also seems to have
been grammaticalized into a kind of Terminative Perfect (Winford, in preparation).
Still other TMA categories emerged gradually over time, due to processes of
grammaticalization. Some of these, like Relative Past ben < been appear to have emerged
earlier than others, e.g., Future o < go and Imperfective e. The latter appears to have
23
emerged via extension of the meaning of an earlier Progressive marker de to cover
habitual and generic meanings as well. This kind of development that is has parallels in
the development of progressives into imperfectives cross-linguistically. It is likely that de
‘there’ was first reanalyzed as a locative copula and then developed into a progressive
marker by analogy with the locative copulas of Gbe languages, which also function as
Progressive markers (see, e.g., Jondoh 1980:46). A similar development of locative
copula de also occurred in Western Caribbean creoles such as Jamaican and Belizean.
Contemporary Sranan also has other TMA markers that evolved via gradual
grammaticalization (e.g. modal man ‘be able to’) or were borrowed from Dutch (e.g.,
modal mag ‘may’). Fuller details of these and other developments in the restructuring of
Sranan TMA can be found in Winford (in preparation).
The creation of Sranan grammar was clearly a gradual process, in which successive
generations of learners contributed in different ways to the elaboration and
systematization of the grammar. Children acquiring Sranan Tongo no doubt played an
important role in these developments. As Arends (1989) and Baker (1995) have both
argued, we cannot assume that all of the features of creole grammar were established in a
single generation, as Bickerton (1984) claimed, or even in the first 70 years of settlement.
The evidence from the available historical texts suggests that many features, including
TMA markers, emerged over a much longer period of time. For SN, the past marker ben
and future marker sa were first attested by 1718 (Baker 1995). Van den Berg (2000:50)
quotes uses of VP-final kaba from court records in 1745 and of the progressive marker de
in 1762. She also provides examples of two-marker combinations (e.g. sa ben) from
1757.
6. Conclusion.
The foregoing comparison of the emergence and development of TMA systems in
Haitian Creole and Sranan Tongo demonstrates that no single formula can be found to
explain creole formation. But all cases of creole formation are in varying ways similar to
cases of second language acquisition in ‘natural’ settings. This paper has attempted to
justify and elaborate this view by examining more closely the similarities in the
24
developmental stages, processes and principles that characterize the two cases of
language acquisition.
Both involve an initial or early stage of interlanguage (IL) creation followed by
elaborative stages in which three major sources of input are involved. These include input
(intake) from native and non-native varieties of the lexifier language, L1 influence, and
internally driven changes that regularize and expand the grammar. The interaction
between L1 knowledge, intake from superstrate sources and creative adaptation operates
within the developing IL system itself – or more accurately, within the minds of
individual learners creating IL systems or I-languages.
The process of elaboration involves various types of restructuring, in the sense in
which scholars of first and second language acquisition define the term. This
restructuring of available materials into a creole grammar involves processes familiar in
SLA. These include reductive and elaborative simplification, processes of reanalysis due
to “transfer” or substrate influence, and processes of regularization and leveling that yield
a uniform and transparent grammar. Various mechanisms and principles similar to those
that operate in SLA guide these processes. The role of such principles is to constrain the
processes of restructuring by which superstrate and substrate inputs (intakes) are shaped
into a viable grammar – one that conforms to universal principles of language design.
Such principles play a role in all phases of creole formation – the early pidginization
stage, the elaborating stages and the later developmental stages.
The creation of creoles differs from more usual cases of SLA in certain respects. For
instance, there are differences in the nature of the target language and the kinds of input
from that source. Another major difference lies in the perseverance of L1-based strategies
and other internal innovations in creole formation, by contrast with SLA, which, as it
progresses, typically involves replacement of such strategies (and other compensatory
ones) by those adopted from the TL. Creoles whose creators have had more access to
superstrate sources exploit those resources more fully, and as a result, approximate
superstrate grammars more closely than others. Some, like Sranan Tongo, depart more
radically from the lexifier language because of the need to rely more on L1 knowledge
and internal innovations, due to restricted availability of superstrate models. In general,
25
however, we conclude that creole formation was essentially a process of SLA with
(usually) restricted TL input under unusual social circumstances.
26
References.
Andersen, Roger W. 1983. Transfer to Somewhere. In S. Gass & L.Selinker (eds.)
Language Transfer in Language Learning, 177-201. Rowley, Mass.: Newbury
House.
Andersen, Roger W. 1984. The One to One principle of interlanguage construction.
Language Learning 34, 77-95.
Andersen, Roger W. 1990. Models, processes, principles and strategies: second language
acquisition inside and outside the classroom. In Bill VanPatten & James F. Lee
(eds.) Second Language Acquisition/Foreign Language Learning, 45-66.
Philadelphia: Multilingual Matters Ltd.
Arends, Jacques. 1989. Syntactic developments in Sranan. Ph.D. dissertation, University
of Nijmegen.
Arends, Jacques. 1995. Demographic factors in the formation of Sranan. In Jacques
Arends (ed.) The early stages of creolization, 233-85. Amsterdam: Benjamins.
Baker, Philip. 1990. Off target. Journal of Pidgin and Creole Languages 5: 107-19.
Baker, Philip. 1995. Some developmental inferences from the historical studies of pidgins
and creoles. In Jacques Arends (ed.) The early stages of creolization, 1-24.
Amsterdam: Benjamins.
Bickerton, Derek. 1984. The language bioprogram hypothesis. The Behavioral and Brain
Sciences 7, 173-88.
Bickerton, Derek. 1988. Creole languages and the bioprogram. In Frederick J. Newmeyer
(ed.) Linguistics: The Cambridge Survey: vol II. Linguistic theory: Extensions and
applications, 268-284. Cambridge:Cambridge University Press.
Bickerton, Derek. 1999. How to acquire language without positive evidence: What
acquisitionists can learn from creoles. In Michel DeGraff (ed.), 1999a, pp. 49-74.
Brown, H. D. 1980. Principles of language learning and teaching. Englewood Cliffs, NJ:
Prentice Hall.
Bruyn, Adrienne. 1994. Some remarkable facts in Sranan: a discussion of possible
accounts. Paper presented at the joint meeting of the SCL and SPCL, Georgetown,
Guyana.
27
Chaudenson, Robert. 1981. Textes creoles anciens ( La Réunion et Île Maurice).
Comparaison et essai d’analyse. Hamburg: Buske.
Chaudenson, Robert.1992.Des îsles des hommes, des langues. Paris: L’Harmattan.
Chaudenson, Robert. 1995. Les Créoles. Paris: Presses Universitaires de France.
Chaudenson, Robert. 2001. Creolization of language and culture. (Revised version of
Chaudenson 1992, in collaboration with Salikoko Mufwene) London and New
York: Routledge.
DeGraff, Michel (ed.) 1999a. Language Creation and Language Change: Creolization,
Diachrony and Development. Cambridge, Mass.: The MIT Press.
DeGraff, Michel. 1999b. Creolization, Language Change, and Language Acquisition: An
Epilogue. In Michel DeGraff (ed.) 1999a, pp. 473-543.
Detgers, Ulrich. 2000. Two types of restructuring in French creoles: A cognitive
approach to the genesis of tense markers. In Newman-Holzschuh & Schneider
(eds.), pp.135-62.
Fattier, Dominique. 1998. Contribution à l’Étude de la Genèse d’un Créole: L’Atlas
Linguistique d’Haïti, Cartes et Commentaires. Université de Provence: Doctoral
dissertation (Distributed by Presses Universitaires du Septentrion, Villeneuve
d’Ascq, France.
Felix, Sascha W. 1977. Early syntactic development in first and second language
acquisition. In C. A. Henning (ed.) Proceedings of the Los Angeles Second
Language Research Forum, 147-59. Los Angeles.
Hesseling, Dirk Christiaan. 1897. Het Hollandsch in Zuid-Afrika. De Gids 60 (1), 138-
162. (Reprinted in English in Dirk Hesseling (1979) On the origin and formation of
creoles: a miscellany of articles. Ann Arbor: Karoma.
Hymes, Dell. 1971. Section III. General Conceptions of Process: Introduction. In Dell
Hymes (ed.) Pidginization and Creolization of Languages, 65-90. Cambridge:
Cambridge University Press.
Jahr, Ernst Håkon. 1996. On the pidgin status of Russenorsk. In Ernst Håkon Jahr &
Ingvild Broch (eds.) Language Contact in the Arctic: Northern pidgins and contact
languages, 107-22. Berlin: Mouton de Gruyter.
28
Jespersen, Otto. 1922. Language: its nature, development, and origin. London: Allen and
Unwin.
Jondoh, Edina Elemawusi Ayaba. 1980. Some aspects of the predicate phrase in Gengbe.
Ph.D. dissertation, Indiana University.
Jordens, Peter. 1996. Input and instruction in second language acquisition. In Peter
Jordens & Josine Lalleman (eds.) Investigating second language acquisition, 407-
49. Berlin: Mouton de Gruyter.
Klein, Wolfgang & Clive Perdue. 1997. The Basic Variety (or: Couldn’t natural
languages be much simpler?). Second Language Research 13:4, 301-47.
Kotsinas, Ulla-Britt. 1996. Aspect marking and grammaticalization in Russenorsk
compared with Immigrant Swedish. In Ernst Håkon Jahr & Ingvild Broch (eds.)
Language Contact in the Arctic: Northern pidgins and contact languages, 123-54.
Berlin: Mouton de Gruyter.
Kouwenberg, Silvia. 1996. Short Note: Substrate or superstrate: What’s in a name? JPCL
11:2, 343-347.
Lalleman, Josine. 1996. The state of the art in second language acquisition research. In
Peter Jordens and Josine Lalleman (eds.) 1996, Investigating Second Language
Acquisition. 3-69. Berlin: Mouton de Gruyter.
Lefebvre, Claire. 1996. The tense, mood, and aspect system of Haitian Creole and the
problem of transmission of grammar in creole genesis. JPCL 11:2, pp. 231-311.
Lefebvre, Claire & John S. Lumsden. 1994. Relexification in creole genesis. Paper read
at the MIT Symposium on the role of relexification in creole genesis: The case of
Haitian Creole.
Lumsden, John S. 1999.Language acquisition and creolization. In Michel DeGraff (ed.)
1999, pp. 129-57.
McWhorter, John H. 1992. Substratal influence on Saramaccan serial verb constructions.
JPCL 7, 1-53.
Meisel, Jürgen M. 1977. Linguistic simplification: a study of immigrant workers’ speech
and foreigner talk. In S. P. Corder & E. Roulet (eds.) The Notions of Simplification,
Interlanguages and Pidgins in their relation to second language pedagogy, 88-113.
Geneve: Droz.
29
Meisel, Jürgen. 1983. Strategies of second language acquisition, more than one kind of
simplification. In Roger W. Andersen (ed.) Pidginization and Creolization as
language acquisition, 120-157. Rowley, Mass.: Newbury House.
Migge, Bettina. 1998. Substrate influence in creole formation: The origin of give-type
serial verb constructions in the Surinamese Plantation Creole. JPCL 13:2, 215-65.
Migge, Bettina. 2000. The Origin of the Syntax and Semantics of Property Items in the
Surinamese Plantation Creole. In John McWhorter (ed.), pp. 201-34.
Migge, Bettina. To appear. The origin of the copulas (d/n)a and de in the Eastern Maroon
Creole. To appear in Diachronica.
Mufwene, Salikoko. 1990. Transfer and the substrate hypothesis in creolistics. Studies in
Second Language Acquisition 12, 1-23.
Mufwene, Salikoko. 1996a. The development of American Englishes: Some questions
from a creole genesis perspective. In Edgar W. Schneider (ed.) Focus on the USA.
(Varieties of English Around the World. G16) Amsterdam: Benjamins, pp. 231-64.
Mufwene, Salikoko. 1996b. The Founder Principle in creole genesis. Diachronica 13: 83-
134.
Prévost, Philippe & Lydia White. 2000. Missing surface inflection or impairment in
second language acquisition? Evidence from tense and agreement. Second
Language Research 16.2, 103-133.
Poulisse, Nanda. 1996. Strategies. In Peter Jordens & Josine Lalleman (eds.)
Investigating second language acquisition, 135-63. Berlin: Mouton de Gruyter.
Schumann, John H. 1978. The pidginization process: A model for second language
acquisition. Rowley, MA: Newbury House.
Sebba, Mark. 1987. The syntax of serial verbs. Amsterdam: Benjamins.
Siegel, Jeff. 2000. Substrate influence in Hawai’I Creole English. Language in Society
29, 197-236.
Singler, John. 1990. On the use of sociohistorical criteria in the comparison of creoles.
Linguistics 28, 645-69.
Smith, Norval. 1996. WE-focus in Saramaccan: substrate feature or grammaticalization?
In Philip Baker and Anand Syea (eds.), Changing meanings, changing functions:
30
Papers relating to grammaticalization in contact languages, 113-128. London:
University of Westminster Press.
Spears, Arthur. 1990. Tense, Mood and Aspect in the Haitian Creole Preverbal Marker
System. In John Singler (ed.) Pidgin and Creole Tense-Mood-Aspect Systems, 119-
142. Amsterdam: Benjamins.
Sylvain, S. 1936. Le créole haïtien. Morphologie at syntaxe. Wetteren: Impr. De Meester.
van Buren, Paul. 1996. Are there principles of universal grammar that do not apply to
second language acquisition? In Peter Jordens and Josine Lalleman (eds.) 1996,
Investigating Second Language Acquisition, 187-207. Berlin: Mouton de Gruyter.
van den Berg, Margot. 2000. “Mi no sal tron tongo” Early Sranan in court records 1667-
1767. MA Thesis, Katholieke Universiteit Nijmegen.
Van Patten, B. 1996. Input Processing and Grammar Instruction in Second Language
Acquisition. Norwood, NJ: Ablex.
Winford, Donald. 2000b. Tense and aspect in Sranan and the creole prototype. In John
McWhorther (ed.) Language change and language contact in pidgins and creoles,
383-442. Amsterdam: John Benjamins.
Winford, Donald. In preparation. Sranan TMA and substrate influence.
31