Hostname: page-component-74656f9884-c52kr Total loading time: 0 Render date: 2026-11-06T15:34:43.257Z Has data issue: false hasContentIssue false
  • English
  • Français

The ain't constraint: Not-contraction in early African American English

Published online by Cambridge University Press:  16 February 2005

James A. Walker
James A. Walker
Affiliation:
York University
Rights & Permissions [Opens in a new window]

Abstract

Studies of negation in African American English (AAE) typically focus on its most salient exponents, ain't and negative concord. Because ain't arose during the development of auxiliary- and not-contraction in Early Modern English, an interesting question is whether constraints on ain't can be attributed to more general constraints on contraction. This article examines the constraints on not-contraction in three varieties argued to be representative of Early AAE. Although the analysis is complicated by the ever-narrowing variable context of ain't and by the competition of not-contraction with auxiliary contraction, results are largely parallel across the three varieties, pointing to a common origin. The parallels between ain't and not-contraction provide evidence that ain't is the extension of more general processes of contraction. The most consistent effect, the presence of negative concord, is argued to reflect a recurrent process of reinforcement in the history of English negation.The data on which this study is based were extracted from corpora housed in the Sociolinguistics Laboratory at the University of Ottawa. I gratefully acknowledge Professor Shana Poplack's permission to use these data. Earlier versions of this article were presented at the annual meeting of the American Dialect Society (Washington, DC, January 2001) and the third U.K. Language Variation and Change conference (York, U.K., July 2001). The analysis benefited from discussions with and comments from Greg Guy, Dennis Preston, Jennifer Smith, and Gerard Van Herk, as well as several anonymous reviewers. Special thanks go to Sali Tagliamonte and Malcah Yaeger-Dror, whose comments substantially improved the article, and to Anthony Warner for help with translating the Old English examples. Any remaining errors are my own responsibility.

Information

Type
Research Article
Information
Language Variation and Change , Volume 17 , Issue 1 , March 2005 , pp. 1 - 17
Copyright
© 2005 Cambridge University Press

Negation has recently come under scrutiny in the debate over the origins of African American English (AAE; cf., Howe, 1997; Howe & Walker, 2000; Weldon, 1994; Winford, 1998:107–109). Paradoxically, the negation system of AAE has been characterized both as one of “the chief areas in which it shows traces of its creole origin” (Winford, 1992:350) and as “thoroughly nonstandard English in character” (Winford, 1998:109). However, studies of negation in AAE have largely focused on only a handful of its morphosyntactic exponents, such as ain't, negative concord,1

“Negative concord” is a term used by Labov (1972) to describe constructions in which negation spreads over elements in the clause or sentence other than the auxiliary. This phenomenon is referred to by other terms in the historical and grammatical literature, such as “redundant negation” (Ramsey, 1892), “negative attraction” (Jespersen, 1917), “negative agreement” (Kaplan, 1989:17–18; Labov et al., 1968), “cumulative negation” (Barber, 1997), “pleonastic negation” (Curme, 1977; Martin, 1992), and “negative spread” (den Besten, 1986).

and negative inversion, as exemplified in (1).
2

Examples from the diaspora corpora are identified by corpus, speaker number, and line number in the transcription. Corpora include:

As I have argued for other variants, such as zero copula (Walker, 2000b) and nonstandard verbal -s (Walker, 2001), the near-exclusive focus on such features of AAE has no doubt arisen because of their salience to speakers of more standard varieties of English, in which these variants are either absent or occur at negligible rates.

However, as over thirty years of variationist research has shown, we can only understand the patterning of variants by examining their role in the entire linguistic system. To this end, this article widens the study of ain't and narrows the study of the negation system by investigating the factors conditioning the variable realization of not-negation in three varieties argued to represent an earlier stage of AAE. After briefly reviewing the variationist literature on negation in AAE, I provide a historical perspective outlining the development of not and not-contraction from Old and Middle English through Early Modern English to modern standard and nonstandard varieties. This review informs the discussion of my analysis of not-contraction in Early AAE. Here, as with other variationist work on these varieties (cf., the papers in Poplack, 2000; Poplack & Tagliamonte, 2001; Walker, 2000b, 2001), I invoke comparison across the different varieties to quantitatively determine which of the factors extracted from the literature condition the observed variability.

NEGATION IN AFRICAN AMERICAN ENGLISH

The study of negation in AAE has been dominated by attention to its most salient exponents, most notably, ain't and negative concord. Early studies suggested that these features predominate in AAE. For example, ain't was found to alternate not only with the negative forms of be and have, but also with did and (more infrequently) do (Fasold & Wolfram, 1975:69; Labov, 1972:284; Labov, Cohen, Robins, & Lewis, 1968:255–257; Weldon, 1994), a pattern not found in other varieties of English (Wolfram, 1991:293; but cf., Feagin, 1979:215). Studies of adolescents in Harlem, Detroit, and elsewhere have found high, in some cases near-categorical, use of negative concord in AAE (Baugh, 1983:82; Labov et al., 1968:277; Wolfram, 1969).

Parallels have been drawn between patterns of negation in AAE and those in the English-based creoles (EBCs) of the Caribbean. In light of its alternation with a wider array of auxiliaries than is found in other varieties of English and its apparently indiscriminate occurrence across temporal and aspectual contexts (DeBose & Faraclas, 1993:370; cf., Mufwene, 1993:100, for Gullah), DeBose (1994:128) argued that ain't in AAE is a monomorphemic negation marker, analogous to that found in EBCs (e.g., Bickerton, 1975; Winford, 1998:108). However, studies of AAE have provided quantitative evidence against the hypothesis that ain't is the relic of a prior creole system (Brewer, 1974; Ewers, 1996:231–232; Howe, 1997; Howe & Walker, 2000; Schneider, 1989; Weldon, 1994;). In contrast with DeBose's (1994) claim, these studies all show that ain't is basically restricted to negating have and be and occurs largely in the present tense, just as in nonstandard varieties of English. Weldon's (1994) finding that ain't also negates did in modern AAE, a pattern not reported for Early AAE (e.g., Ewers, 1996; Howe, 1997; Howe & Walker, 2000), suggests that the more widespread use of ain't in AAE is a relatively recent development.

In addition to the type of auxiliary,3

In this article I use the cover term “auxiliary” to refer to all forms that function to express mood, tense, and aspect, including the copula and modals. I distinguish among these forms where relevant.

these studies propose other factors as constraints on the occurrence of ain't. Weldon (1994; following Wolfram, 1973) found the presence of negative concord to promote ain't, especially in the context of negated have. A frequently noted consideration is verbal aspect, though the purported effects are inconsistent. DeBose and Faraclas (1993:370) argued that ain't negates nonhabitual predicates, which would explain its exclusion from the (usually habitual) context of do. In contrast, Bolton (1982:323) argued that ain't occurs in punctual contexts, with don't be the negator of durative contexts (e.g., She don't be sick, she don't be going.). Weldon (1994) and Howe (1997) both found verbal stativity to correlate with ain't, though Howe noted that it also correlates with forms of negated be.

These studies have elucidated several linguistic constraints on ain't, such as the type of auxiliary, the presence of negative concord, and verbal aspect. However, they leave open the question of whether such constraints are in fact specific to ain't, or instead reflect more general constraints on the process of not-contraction. To answer this question, it is first necessary to examine the historical development of not and the constraints on its contracted form.

A BRIEF HISTORY OF ENGLISH NOT

Jespersen (1917, 1966) characterized the history of English negation as cyclic, with negating elements arising via grammaticization from lexical elements which themselves then become weakened and require reinforcement by other elements in the sentence. It is by this route that not entered English as a form of negation and then itself became contracted.

In Old English (OE), the verb was negated by immediately preposing ne (Strang, 1970:312; Mitchell & Robinson, 1992:101–102). Although this element could be stressed (as na or no), more frequently it cliticized to the verb and could be contracted if the verb began with a vowel, [h] or [w]:

No doubt because of the prosodic or perceptual weakness of ne, not began to grammaticize from the adverb naht, nought, not, nat, etc. (‘not at all’) as a reinforcing element for negation (Strang, 1970:209, 312; Rissanen, 1999:192, 200). Whereas not occurred sporadically in OE, by Middle English (ME) it had become the norm (Burrow & Turville-Petre, 1996:52–53; Mossé, 1952:112; Strang, 1970:312), cooccurring variably with ne (Jack, 1978:62):

As not eventually became indispensable in expressing negation, the clitic ne became redundant, surviving only in a few fossilized forms (Rissanen, 1999:201; Strang, 1970:209).

Not itself began to be contracted in Early Modern English, a process that followed closely on the development of auxiliary contraction (Rissanen, 1999:195; Strang, 1970:151; Warner, 1993:208). Contracted -n't is generally considered to have entered spoken English around 1600, appearing in writing half a century later (Barber, 1997:254; Jespersen, 1966:117; Rissanen, 1999:196; though Brainerd [1989] provided evidence from slightly earlier). The first citation in the Oxford English Dictionary dates from 1652:

The irregular third-person singular form don't (< does + not) also occurs as early as the 1660s and is common in the writings of Defoe, Goldsmith, and Sheridan (Lass, 1999:180).

In this period, ain't (or rather, its spelling variants han't and en't) makes its first appearance, as a contraction of are + not, have + not and am + not, in the writings of permissive grammarians such as Miege (1688:110–111) and in plays, as the examples in (5) illustrate.

Contractions were initially associated with colloquial speech, usually in the mouths of disreputable or bucolic characters (Brainerd, 1989:182; Haugland, 1995:166–167), though contracted -n't survived the prescriptivist “war upon contractions” of the 18th century (Haugland, 1995:172). Lass (1999:180) noted that variants of ain't, such as ha'n't (> have + not) and an't (> am + not), were also both common in this period.

While not-contraction gradually became acceptable at all social levels, ain't eventually disappeared from respectable middle-class speech and is now a shibboleth of nonstandard English (cf., Wolfram & Schilling-Estes, 1998:158). In fact, Strang (1970:151) argued that the modern forms haven't and hasn't were reformed by analogy with the positive forms have and has once ain't became stigmatized.

Varieties of Modern English are said to differ regionally in their preference for not-contraction. Biber (1987) stated that American English tends to favor not-contraction more than British English (cf., Greenbaum, 1977). Within British English, a north–south split is noted, with speakers of southern varieties preferring not-contraction (e.g., I won't) and speakers of Scottish and northern varieties preferring auxiliary contraction (e.g., I'll not) (Aitken, 1984; Cheshire, 1991:63; Rissanen, 1999:202; Trudgill, 1978:13, 1984:33). Some studies confirm this perception, especially in earlier centuries (e.g., Kjellmer, 1998:179; Yaeger-Dror, Hall-Lew, & Deckert, 2002). Tagliamonte and Smith's (2002) recent empirical analysis of seven varieties of English in Britain and Northern Ireland revealed “no clearcut north–south divide,” but rather a more complicated regional picture.

Studies of modern varieties of English have proposed several linguistic constraints on both not-contraction and ain't. First is the type of auxiliary, featuring wide variance in their occurrence with not-contraction, even within the same dialect. Not-contraction is said to be preferred with have and will, whereas auxiliary contraction is preferred with be (Cheshire, 1991:62; Greenbaum, 1977; Hiller, 1987; Kjellmer, 1998:169; Selkirk, 1981:114; Tagliamonte & Smith, 2002). Ain't is similarly conditioned by the type of auxiliary, restricted to have and be, though more frequent with have (Cheshire, 1991:57). A related consideration is the grammatical person of the subject (Wolfram, 1973), though this may have to do with the irregular paradigm of auxiliary verbs (cf., Hudson, 2000). The type of clause has also been found to be important: not-contraction occurs more frequently in questions than in declaratives (Rissanen, 1999:198) and ain't is higher in tag questions than other types of questions and declaratives (Cheshire, 1982, 1991:58).

Obviously, while not-contraction exists in just about every variety of English,4

One exception to this generalization is vernacular Scots, which expresses negation via affixation of -na/nae rather than -n't (Smith, 2000).

standard or nonstandard, ain't is highly stigmatized. Both are constrained by a number of similar factors, most notably the type of auxiliary and the type of clause. An interesting question that has not been explored, though it holds possible implications for the AAE origins debate, is whether the constraints that have been observed to condition the occurrence of ain't also affect not-contraction, and vice versa. In the following section, I report on my analysis of both not-contraction and ain't in varieties of African American English that have been argued to represent an earlier stage of AAE.

NOT-CONTRACTION IN EARLY AFRICAN AMERICAN ENGLISH

Data

The data on which this study is based were taken from varieties of AAE spoken in disparate communities, corpora of which are housed in the Sociolinguistics Laboratory of the University of Ottawa. The first two corpora represent African American diaspora communities in North America, founded in the late 18th and early 19th centuries by escaped African American slaves and freedmen (Walker, 1992). The varieties spoken in these communities have been argued to represent an earlier form of AAE because of their social and geographic isolation (Poplack, 2000; Poplack & Sankoff, 1987; Poplack & Tagliamonte, 1991, 2001). African Nova Scotian English (ANSE; Poplack & Tagliamonte, 1991, 2001) is spoken in two communities in Nova Scotia (Guysborough and North Preston), which have until recently been isolated from the neighboring white communities (Walker, 1992:27–30, 388–391; Poplack & Tagliamonte, 1991:307–310, 2001:49–68). The African Nova Scotian English Corpus consists of sociolinguistic interviews conducted in 1990–1991 by local community members (Poplack & Tagliamonte, 1991, 2001). From this corpus I selected 24 interviews for this analysis, representing speakers from each of the two communities. Samaná English (SAM; Poplack & Sankoff, 1987) is spoken on the isolated northeastern peninsula of the Dominican Republic, in a community founded in 1824. The Samaná English Corpus consists of 21 interviews conducted in 1982 (Poplack & Sankoff, 1987; Poplack & Tagliamonte, 2001:10–38). The third corpus, the Ex-slave Recordings (ESR; Bailey, Maynor, & Cukor-Avila, 1991), represents a sampling of AAE speakers from the continental United States, which serves as a baseline for the diaspora varieties. This corpus consists of interviews with 11 former African American slaves born between 1844 and 1861, recorded between 1935 and 1974 (mostly in 1940–1942) by the Federal Writers' Project.

Although I will refer to these corpora collectively as Early African American English (Early AAE), it should be noted that their representativeness has been criticized in the literature on various grounds, such as the origins of the communities' settlers and the possibility of subsequent linguistic change caused by language contact or internal divergence, as well as the extent to which the interviewers succeeded in tapping the speakers' vernaculars (e.g., Hannah, 1997; Holm, 1988:505; Rickford, 1998; Singler, 1997). However, the variationist approach to historical reconstruction that I adopt in this study lessens, if not obviates, these concerns. Rather than relying on the overall rates of occurrence of a particular form in each of these varieties (which can fluctuate for a variety of language-external reasons; cf., Rickford & McNair-Knox, 1994), I compare the hierarchy of linguistic constraints on variability across all varieties. The constraint hierarchy serves not only as a diagnostic of system membership, but also as evidence of genetic affiliation. Given the very different origins, development, and subsequent language-contact scenarios of these communities, not to mention the different circumstances under which each data set was collected, any similarities in constraint hierarchies that cannot be attributed to universal linguistic principles can only be adduced as evidence for a shared ancestral grammatical system (cf., Poplack, 2000:13; Tagliamonte, 2002).

Defining the variable context

In contrast with previous work on negation in these varieties (Howe, 1997; Howe & Walker, 2000), the focus of this study is not the entire negation system, but rather, not-negation (Tottie, 1991:15), clauses in which negation is expressed via the postauxiliary negator not. After creating a concordance (Rand, 2001) of all three corpora, I extracted every instance of not-negation from the selected interviews. I excluded environments where variability does not occur or is unclear, such as ellipses (6a), in which there is neither a subject nor an auxiliary (and thus where not-contraction cannot occur); false starts (6b), which represent performance errors; and frozen expressions (6c–d), because these forms are not productive.

These protocols yielded a total of 4,880 tokens, divided by community as shown in the first column of Table 1.

Overall distribution of tokens of not-negation in Early African American English, by community

Coding

I began by classifying each token into five possible forms of not-negation. There may be no contraction at all (7a), or the auxiliary may be contracted instead of not (7b). In these varieties, some auxiliaries may also be deleted altogether (7c).

Because my concern here is with the variable contraction of not, which requires an auxiliary host to cliticize or affix to, I excluded tokens of auxiliary deletion for the present analysis. In cases of not-contraction (7d–e), either -n't may be affixed to the auxiliary or the suppletive form ain't may be used.

However, not all types of contraction can occur in all environments. As Figure 1 shows, there is an ever-narrowing context of variability, which entails consequences for the method of calculating rates of occurrence for each of these processes. I will return to this consideration later.

Variable contexts for not-contraction, aux-contraction and ain't.

In addition to noting the form of not-negation, I also coded each token for a number of factors hypothesized in the literature to condition ain't.5

An additional possible conditioning factor is the interactive nature (e.g., supportive, remedial, informative) of the conversational turn in which not occurs (cf., Pomerantz, 1984). However, Yaeger-Dror (personal communication) suggested that conversational turns in sociolinguistic interviews are predominantly informative or supportive. Because of this consideration and the difficulty of coding such distinctions reliably, I do not consider this factor here.

Auxiliary or modal

The first factor group coded was the (potential) host of -n't, which includes modal and tense-aspect auxiliaries and the copula (i.e., be, have, do, will, would, can, could, may, might, should, must), as well as whether this host occurred in full or contracted form.

Type of clause

The type of clause was coded as declarative or as one of several nondeclarative types: questions (8a), imperatives (8b), tag questions (8c), and negative inversions (8d).

Because initial analysis showed not-contraction to be categorical in the last three clause types, they were excluded from further analysis. Restricting the context to declaratives reduced the total number of tokens to 4,627, as shown in the second column of Table 1.

Type of subject

The type of subject has proven important for auxiliary contraction (cf., McElhinney, 1993; Walker, 2000a; Weldon, 1994), with pronominal subjects favoring contraction. Because auxiliary contraction is often cited as competing with not-contraction, each token was coded for whether the subject was a noun phrase or pronoun.

Presence of negative concord

Given the finding that the presence of negative concord promotes the occurrence of ain't (Weldon, 1994; Wolfram, 1973), I coded each token for the presence of negative concord within the same clause. Negative concord was considered to be present if the subject (9a), object (9b), or other complement (9c) was negated.

Lexical stativity

As discussed previously, verbal aspect has been cited as an important consideration in constraining ain't. In this analysis I focus on considerations of stativity, that is, whether the verb describes a state or an event. Following Howe (1997; Howe & Walker, 2000), I assume that auxiliaries take on the stativity of the main verb and that copular be is inherently stative. However, there are various and often conflicting criteria that have been proposed for classifying verbs as stative or nonstative (e.g., Mufwene, 1984:2; Scheffer, 1975:61–64; Winford, 1992), which makes it difficult to code this factor in an objective and replicable manner.

In line with my previous work in this area (Walker, 2000b, 2001), here I adopt a model of aspectual semantics in which the underlying aspectual representation of verbs can be augmented (but not deleted) by contextual elements of the proposition (Levin, 1993; Olsen, 1997; Smith, 1997). By distinguishing between lexical aspect and propositional (or sentential) aspect (Brinton, 1987:204), this model allows us to account for the fact that the same verb can occur in a range of aspectual contexts. I coded each verb for whether it was lexically stative (10a) or nonstative (10b), regardless of whether or not its use in the sentence was stative.

Analysis

All of the above factors were considered individually and together using the multiple-regression feature incorporated in goldvarb 2.1 for the Macintosh (Rand & Sankoff, 1990).

Overall distribution

The final column of Table 1 shows the overall rate of not-contraction in these varieties (including both -n't and ain't). Note that not-contraction is preferred almost categorically in all communities. However, a breakdown of the percentages by potential host, as shown in Table 2, reveals that the pattern in Table 1 does not apply across all forms. Contraction is categorical (or nearly so) for most modals and some auxiliaries, especially can, could, should, do, and had, but it never occurs with may (perhaps because of the phonotactic complexity of mayn't or its relative rarity in informal speech). In fact, most of the variability occurs with the present-tense forms of the auxiliaries be and have and the modal will. These forms, along with would, also happen to be the only auxiliaries that can be contracted, suggesting that auxiliary contraction and not-contraction are alternative strategies for reducing clausal complexity (cf., Walker, 2000a).

Distribution of not-contraction (including ain't) by auxiliary/modal in Early African American English

Auxiliary contraction vs. not-contraction

To test the competing effects of auxiliary contraction and not-contraction, I narrowed the variable context to only those auxiliaries in which both strategies of contraction are possible (see Figure 1). Thus, all auxiliaries except present-tense be, have, and will were excluded (would never occurred in its contracted form in these data and thus was also excluded).6

For comparability with other studies (e.g., Tagliamonte & Smith, 2002; Yaeger-Dror, Hall-Lew, & Deckert, 2002), I exclude full forms (i.e., tokens in which neither the auxiliary nor not was contracted) from the calculations. Comparison of calculations with and without full forms reveals little difference in the results.

The results of the multivariate analysis are shown in Table 3.7

The small numbers in Samaná English and the ESR preclude the use of the multiple-regression feature.

First, note the differential patterning according to auxiliary. There is a virtually categorical preference for not-contraction with have and will, which means that the competition between auxiliary contraction and not-contraction occurs with one auxiliary, be. The results for subject type are consistent across all three varieties, with noun phrase subjects promoting not-contraction. This result is unsurprising, given the well-known tendency of pronouns to favor auxiliary contraction in affirmative contexts, as well (e.g., Blake, 1997; McElhinney, 1993; Walker, 2000a). Also consistent across varieties is the tendency for not-contraction to correlate with negative concord, as Weldon (1994) and Wolfram (1973) found for ain't. Finally, not-contraction tends to be preferred with nonstative verbs. These results are revealing of the interplay between auxiliary contraction and not-contraction. However, because ain't was included under not-contraction, it is unclear to what extent the effects in Table 3 can be attributed to the variable context.

Factors contributing to the occurrence of not-contraction (including both -n't and ain't) against auxiliary contraction in three varieties of Early African American English

not-contraction vs. ain't

To examine -n't and ain't separately, we need to consider two issues. First, as Howe and Walker (2000) found, ain't alternates primarily with forms of be and have in these varieties, which requires further reducing the variable context for ain't to these two auxiliaries, as shown in Figure 1.

A second consideration is the method of calculation for ain't. To my knowledge, this issue has largely been neglected, in sharp contrast with studies of the copula, in which the method of calculation remains controversial (Blake, 1997; Rickford, Ball, Blake, Jackson, & Martin, 1991; Walker, 2000a, 2000b). For example, Rickford et al. (1991) showed that the choice of method can drastically affect the results obtained. However, as I have noted elsewhere (Walker, 2000a, 2000b), the choice of method should be made on statistical grounds, independent of the analyst's theoretical preferences. This consideration involves the use of procedures for inferring rule ordering from statistical analysis, as outlined by Sankoff and Rousseau (1989:4–7). In line with this work, I ran every possible combination of methods of calculation in goldvarb and compared the log likelihoods of each combination (indicating how well each analysis fit the data). I found that the most appropriate method of calculation for these data was to oppose ain't to both full and not-contracted forms in the context of be and have. To calculate rates of not-contraction without the effect of ain't, I excluded all tokens of ain't and included every auxiliary or modal that exhibited variation between full and not-contracted forms. Finally, because preliminary analysis for ain't revealed that the effects of the type of auxiliary and the stativity of the verb were not independent, these factor groups were combined into one.

Separate multivariate analyses of -n't and ain't, shown in Table 4,8

The small numbers in Samaná English preclude the use of the multiple-regression feature. I exclude the ESR from this table because in this variety there is no not-contraction that is not ain't.

reveal that ain't is preferred with be, especially in nonstative contexts. The favoring effect of nonstatives is also consistent in -n't across all varieties. Surprisingly, the type of subject is significant for both ain't and -n't in the two communities, with pronouns favoring. Finally, with the exception of -n't in Samaná English, there is a tendency for negative concord to favor both -n't and ain't.

Factors contributing to the occurrence of not-contraction and ain't in not-negation contexts in two varieties of Early African American English

DISCUSSION AND CONCLUSION

The results of this analysis demonstrate that the linguistic factors conditioning not-contraction and ain't are parallel, both to each other and across the three varieties of Early AAE considered here. Early AAE conforms to other English varieties in its preference for not-contraction with have and will and for auxiliary contraction with be (Cheshire, 1991; Greenbaum, 1977; Hiller, 1987; Kjellmer, 1998; Selkirk, 1981; Tagliamonte & Smith, 2002), and to other American English dialects in a general preference for not-contraction (Biber, 1987; Greenbaum, 1977). Early AAE differs from other varieties in its preference for ain't with be over have (cf., Cheshire, 1991). This study differs from others (Howe, 1997; Howe & Walker, 2000; Weldon, 1994) in the finding that ain't is favored with nonstative verbs, though this difference may reflect different definitions of stativity. Nevertheless, in this study the effect of stativity is parallel between not-contraction and ain't. Negative concord also largely favors both not-contraction and ain't across these varieties, in line with Weldon's (1994) and Wolfram's (1973) findings.

The results of this analysis raise questions for the methodology of studying not-contraction. The preponderance of ain't in these varieties, which presents problems for the calculation of not-contraction, auxiliary contraction, and ain't, may have distorted many of the effects observed in studies of varieties in which ain't is absent, making it difficult to compare results across studies. Taken together with the higher rates of not-contraction with have and will, the preponderance of ain't could be explained as an instantiation of the morphological “blocking effect” (Aronoff, 1976; Aronoff & Anshen, 1998; Bresnan, 2001), in which forms with relatively opaque morphology and separate lexical entries (e.g., ain't, won't) block more transparent and productive morphological processes (e.g., -n't-affixation). This interpretation could also have consequences for the way we analyze nonstandard third-person singular don't, which is often considered in studies of nonstandard agreement such as verbal -s (e.g., Christian, Wolfram, & Dube, 1988:111, 123–124; Fasold, 1972:123; Wolfram & Christian, 1976:78, 83), but could be interpreted as yet another instantiation of the blocking effect.

These results also have implications for the debate over the origins and development of AAE. The generally parallel results for auxiliary contraction, -n't, and ain't across the three varieties add to findings from other areas of the linguistic system (e.g., Poplack, 2000; Poplack & Tagliamonte, 2001; Walker, 2000b) that these varieties share a common origin. The similarity of these findings with studies of other varieties of English confirms the contention (Howe, 1997; Howe & Walker, 2000) that the negation system of Early AAE is basically that of nonstandard English. In addition to supporting previous studies, this analysis contributes the unexpected findings that not-contraction tends to be favored in nonstative contexts and that both ain't and not-contraction are similarly constrained by negative concord (though not entirely consistently). Weldon (1994) suggested that the latter effect can be attributed to stylistic considerations, whereby nonstandard forms cooccur in less formal contexts. However, following Jespersen (1917, 1966), I view this effect as the preservation in nonstandard varieties of the recurrent and natural cycle of weakening and reinforcement in the history of English negation.

References

REFERENCES

Aitken, A. J. (1984). Scottish accents and dialects. In P. Trudgill (ed.), Language in the British Isles. Cambridge: Cambridge University Press. 95–114.
Aronoff, Mark. (1976). Word formation in generative grammar. Cambridge, MA: MIT Press.
Aronoff, Mark, & Anshen, Frank. (1998). Morphology and the lexicon: Lexicalization and productivity. In A. Spencer & A. Zwicky (eds.), The handbook of morphology. Oxford: Blackwell. 237–247.
Bailey, Guy, Maynor, Nathalie, & Cukor-Avila, Patricia (eds.), (1991). The emergence of Black English: Texts and commentary. Amsterdam: Benjamins.
Barber, Charles. (1997). Early Modern English. Edinburgh: Edinburgh University Press.
Baugh, John. (1983). Black street speech: Its history, structure, and survival. Austin: University of Texas Press.
Biber, Douglas. (1987). A textual comparison of British and American writing. American Speech 62:99–119.Google Scholar
Bickerton, Derek. (1975). Dynamics of a creole system. New York: Cambridge University Press.
Blake, Renee. (1997). Defining the envelope of linguistic variation: The case of “don't count” forms in the copula analysis of African American Vernacular English. Language Variation and Change 9: 57–79.Google Scholar
Bolton, W. F. (1982). A living language: The history and structure of English. New York: Random House.
Brainerd, Barron. (1989 [1993]). The contractions of not: A historical note. Journal of English Linguistics 22:176–196.Google Scholar
Bresnan, Joan. (2001). Explaining morphosyntactic competition. In M. Baltin & C. Collins (eds.), The handbook of contemporary syntactic theory. Oxford: Blackwell. 11–44.
Brewer, Jeutonne. (1974). The verb be in early Black English: A study based on the WPA ex-slave narratives. Doctoral dissertation, University of North Carolina at Chapel Hill, NC.
Brinton, Laurel. (1987). The aspectual nature of states and habits. Folia Linguistica 21: 195–214.Google Scholar
Brome, Richard. (1635). The sparagus garden. Reprinted (1980) in D. S. McClure (ed.), A critical edition of Richard Brome's The weeding of Covent Garden and The sparagus garden. New York: Garland.
Burrow, J. A., & Turville-Petre, Thorlac. (1996). A book of Middle English, Second Edition. Oxford: Blackwell.
Cheshire, Jenny. (1982). Variation in an English dialect: A sociolinguistic study. Cambridge: Cambridge University Press.
Cheshire, Jenny. (1991). Variation in the use of ain't in an urban British English dialect. In P. Trudgill and J. K. Chambers (eds.), Dialects of English: Studies in grammatical variation. Singapore: Longman. 54–73.
Christian, Donna, Wolfram, Walt, & Dube, Nanjo. (1988). Variation and change in geographically isolated communities: Appalachian English and Ozark English. Tuscaloosa: University of Alabama Press.
Curme, George. (1977). A grammar of the English language. Essex, CT: Verbatim.
DeBose, Charles E. (1994). A note on ain't vs. didn't negation in African American Vernacular. Journal of Pidgin and Creole Languages 9:127–130.Google Scholar
DeBose, Charles E., & Faraclas, Nicholas. (1993). An Africanist approach to the linguistic study of Black English: Getting to the roots of the tense-aspect-modality and copula systems in Afro-American. In S. Mufwene (ed.), Africanisms in Afro-American language varieties. Athens: University of Georgia Press. 364–387.
den Besten, Hans. (1986). Double negation and the genesis of Afrikaans. In P. Muysken & N. Smith (eds.), Substrata versus universals in creole languages. Dordrecht: Foris. 185–230.
Duffett, Thomas. (1674). The mock-tempest. Reprinted (1972) in R. E. DiLorenzo (ed.), Three burlesque plays of Thomas Duffett. Iowa City: University of Iowa Press.
Ewers, Traute. (1996). The origin of American Black English: Be-forms in the HOODOO texts. Berlin: Mouton de Gruyter.
Fasold, Ralph. (1972). Tense marking in Black English: A linguistic and social analysis. Washington, DC: Center for Applied Linguistics.
Fasold, Ralph, & Wolfram, Walt. (1975). Some linguistic features of the Negro dialect. In P. Stoller (ed.), Black American English: Its background and its usage in the schools and in literature. New York: Dell. 49–89.
Feagin, Crawford. (1979). Variation and change in Alabama English. Washington, DC: Georgetown University Press.
Greenbaum, Sidney. (1977). Judgement of syntactic acceptability and frequency. Studia Linguistica 31:83–105.Google Scholar
Hannah, Dawn. (1997). Copula absence in Samaná English: Implications for research on the linguistic history of African-American Vernacular English. American Speech 72:339–372.Google Scholar
Haugland, Kari E. (1995). Is't allow'd or ain't it? On contraction in early grammars and spelling books. Studia Neophilologica 67:165–184.Google Scholar
Hiller, Ulrich. (1987). She isn't studying vs. she's not studying: An investigation into the choice between the two contracted variants of negated English auxiliaries. Die Neueren Sprachen 86:531–553.Google Scholar
Holm, John. (1988). Pidgins and creoles, Volume 1: Theory and structure. New York: Cambridge University Press.
Howe, Darin M. (1997). Negation and the history of African American English. Language Variation and Change 9:267–294.Google Scholar
Howe, Darin M., & Walker, James A. (2000). Negation and the creole-origins hypothesis: Evidence from Early African American English. In S. Poplack (ed.), The English history of African American English. Malden, Oxford: Blackwell. 109–140.
Jack, G. B. (1978). Negation in later Middle English Prose. Archivum Linguisticum, n.s. 9:58–72.Google Scholar
Jespersen, Otto. (1917). Negation in English and other languages. Copenhagen: A. F. Host.
Jespersen, Otto. (1966). Negation in English and other languages ( 2nd ed.). Copenhagen: Munksgaard.
Kaplan, Jeffrey P. (1989). English grammar: Principles and facts. Englewood Cliffs, NJ: Prentice-Hall.
Kjellmer, Göran. (1998). On contraction in Modern English. Studia Neophilologica 69:155–186.Google Scholar
Labov, William. (1972). Language in the inner city. Philadelphia: University of Pennsylvania Press.
Labov, William, Cohen, Philip, Robins, Clarence, & Lewis, John. (1968). A study of the non-standard English of Negro and Puerto Rican speakers in New York City. Co-operative Research Report 3288, Vol. I. Philadelphia: U.S. Regional Survey.
Lass, Roger. (1999). Phonology and morphology. In R. Lass (ed.), The Cambridge history of the English language, volume 3: 1476–1776. Cambridge: Cambridge University Press.
Levin, Beth. (1993). English verb classes and alternations: A preliminary investigation. Chicago: University of Chicago Press.
Martin, Stefan E. (1992). Topics in the syntax of nonstandard English. Doctoral dissertation, University of Maryland.
McElhinney, Bonnie S. (1993). Copula and auxiliary contraction in the speech of White Americans. American Speech 68:371–399.Google Scholar
Miege, Guy. (1688 [1969]). The English grammar. Menton, UK: Scolar Press.
Mitchell, Bruce, & Robinson, Fred C. (1992). A guide to Old English ( 5th ed.). Oxford: Blackwell.
Mossé, Fernand. (1952). A handbook of Middle English. Baltimore: Johns Hopkins University Press.
Mufwene, Salikoko S. (1984). Stativity and the progressive. Bloomington: Indiana University Linguistics Club.
Mufwene, Salikoko S. (1993). Scope of negation and focus in Gullah. In F. Byrne & D. Winford (eds.), Focus and grammatical relations in creole languages. Amsterdam: Benjamins. 95–116.
Olsen, Mari Broman. (1997). A semantic and pragmatic model of lexical and grammatical aspect. New York: Garland.
Oxford English Dictionary (2nd ed.). (1989). Oxford: Oxford University Press.
Pomerantz, Anita. (1984). Agreeing and disagreeing with assessments: Some features of preferred/dispreferred turn shapes. In J. M. Atkinson & J. Heritage (eds.), Structures of social action: Studies in conversation analysis. Cambridge: Cambridge University Press. 57–101.
Pope, J. C. (ed.). (1981). Seven Old English poems. New York: Norton.
Poplack, Shana (ed.). (2000). The English history of African American English. Oxford: Blackwell.
Poplack, Shana, & Sankoff, David. (1987). The Philadelphia story in the Spanish Caribbean. American Speech 64:291–314.Google Scholar
Poplack, Shana, & Tagliamonte, Sali. (1991). African American English in the diaspora: The case of old-line Nova Scotians. Language Variation and Change 3:301–339.Google Scholar
Poplack, Shana, & Tagliamonte, Sali. (2001). African American English in the diaspora: Tense and aspect. Oxford: Blackwell.
Ramsey, S. (1892 [1968]). The English language and English grammar: An historical study of the sources, development, and analogies of the language and of the principles governing its usages. New York: Haskell House.
Rand, David. (2001). Concorder: Concordance software for the Macintosh. Université de Montréal: Centre de Recherches Mathématiques.
Rand, David, & Sankoff, David. (1990). GoldVarb: A variable rule application for the Macintosh. Université de Montréal: Centre de Recherches Mathématiques.
Rickford, John R. (1998). The creole origins of African-American Vernacular English: Evidence from copula absence. In S. S. Mufwene, J. R. Rickford, G. Bailey, & J. Baugh (eds.), African-American English: Structure, history, and use. London: Routledge. 154–200.
Rickford, John R., Ball, Arnetha, Blake, Renée, Jackson, Raina, & Martin, Nomi. (1991). Rappin on the copula coffin: Theoretical and methodological issues in the analysis of copula variation in African American Vernacular English. Language Variation and Change 3:103–132.Google Scholar
Rickford, John R., & McNair-Knox, Faye. (1994). Addressee- and topic-influenced style shift: A quantitative sociolinguistic study. In D. Biber & E. Finegan (eds.), Sociolinguistic perspectives on register. New York: Oxford University Press. 235–276.
Rissanen, Matti. (1999). Isn't it? or is it not? On the order of postverbal subject and negative particle in the history of English. In I. Tieken-Boon van Ostade, G. Tottie, & W. van der Wurff (eds.), Negation in the history of English. Berlin: Mouton de Gruyter. 189–205.
Sankoff, David, & Rousseau, Pascale. (1989). Statistical evidence for rule ordering. Language Variation and Change 1:1–18.Google Scholar
Scheffer, Johannes. (1975). The progressive in English. Amsterdam: North-Holland.
Schneider, Edgar W. (1989). American earlier Black English. Tuscaloosa: University of Alabama Press.
Selkirk, Elisabeth O. (1981). The phrase phonology of English and French. Bloomington, IN: Indiana University Linguistics Club.
Singler, John V. (1997). On the genesis, evolution, and diversity of African American English: Evidence from verbal -s in the Liberian Settler English of Sinoe. Paper presented at the Society for Pidgin and Creole Linguistics, London, UK.
Smith, Carlota S. (1997). The parameter of aspect. Second edition. Dordrecht: Kluwer.
Smith, Jennifer. (2000). ‘You Ø na hear o' that kind o'? things’: Negative do in Buckie Scots. English World-Wide 21:231–259.Google Scholar
Strang, Barbara M. H. (1970). A history of English. London: Methuen.
Tagliamonte, Sali. (2002). Comparative sociolinguistics. In J. K. Chambers, P. Trudgill, & N. Schilling-Estes (eds.), The handbook of language variation and change. Oxford: Blackwell. 729–763.
Tagliamonte, Sali, & Smith, Jennifer. (2002). Either it isn't or it's not: NEG/AUX contraction in British dialects. English World-Wide 23:251–281.Google Scholar
Tottie, Gunnel. (1991). Negation in English speech and writing: A study in variation. San Diego, CA: Academic.
Trudgill, Peter (ed.). (1978). Sociolinguistic patterns in British English. London: Edward Arnold.
Trudgill, Peter. (1984). Language in the British Isles. Cambridge: Cambridge University Press.
Walker, James A. (2000a). Rephrasing the copula: Contraction and zero in Early African American English. In S. Poplack (ed.), The English history of African American English. Oxford: Blackwell. 35–72.
Walker, James A. (2000b). Present accounted for: Prosody and aspect in Early African American English. Doctoral Dissertation, University of Ottawa.
Walker, James A. (2001). Using the past to explain the present: Tense and temporal reference in Early African American English. Language Variation and Change 13:1–35.Google Scholar
Walker, James W. St. G. (1992). The Black Loyalists: The search for a promised land in Nova Scotia and Sierra Leone, 1783–1870. Toronto: University of Toronto Press.
Warner, Anthony. (1993). English auxiliaries: Structure and history. Cambridge: Cambridge University Press.
Weldon, Tracey. (1994). Variability in negation in African American Vernacular English. Language Variation and Change 6:359–397.Google Scholar
Winford, Donald. (1992). Back to the past: The BEV/creole connection revisited. Language Variation and Change 4:311–357.Google Scholar
Winford, Donald. (1998). On the origins of African American Vernacular English—A creolist perspective, Part II: Linguistic features. Diachronica 15:99–154.Google Scholar
Wolfram, Walt. (1969). A sociolinguistic description of Detroit Negro speech. Washington, DC: Center for Applied Linguistics.
Wolfram, Walt. (1973). Sociolinguistic aspects of assimilation: Puerto Rican English in New York City. Washington, DC: Center for Applied Linguistics.
Wolfram, Walt. (1991). Dialects and American English. Englewood Cliffs, NJ: Prentice-Hall.
Wolfram, Walt, & Christian, Donna. (1976). Appalachian speech. Arlington, VA: Center for Applied Linguistics.
Wolfram, Walt, & Schilling-Estes, Natalie. (1998). American English: Dialects and variation. Oxford: Blackwell.
Yaeger-Dror, Malcah, Hall-Lew, Lauren, & Deckert, Sharon. (2002). It's not or isn't it? Using large corpora to determine the influences on contraction strategies. Language Variation and Change 14:79–118.Google Scholar
Figure 0

Overall distribution of tokens of not-negation in Early African American English, by community

Figure 1

Variable contexts for not-contraction, aux-contraction and ain't.

Figure 2

Distribution of not-contraction (including ain't) by auxiliary/modal in Early African American English

Figure 3

Factors contributing to the occurrence of not-contraction (including both -n't and ain't) against auxiliary contraction in three varieties of Early African American English

Figure 4

Factors contributing to the occurrence of not-contraction and ain't in not-negation contexts in two varieties of Early African American English