When 17th century personal correspondence was studied, it was observed that women used the evidential expression I THINK more often than men. A closer analysis showed that women also used other 1st person evidential verbs as well as the 1st and 2nd person pronouns more frequently than men. This male/female difference was maintained even in different registers, although both sexes have higher frequencies of I THINK in more intimate circumstances, such as when the informants are writing to their friends or close family members. The male/female differences in frequencies are explained as a difference in the style of communication. Women's style is more “involved” and interactive: personal point of view is frequently expressed, and both the writer and the addressee are overtly included in the communication situation. Interestingly, similar differences have also been found in Present-Day English.

This article demonstrates the importance of investigating language variation and change both within and across ethnic groups, especially those that have been relatively insular historically. The focus is on the variable patterning of /ay/ in the variety of English spoken by the Lumbee Indians in tri-ethnic Robeson County, North Carolina. (The Lumbee refer to themselves as “Indians” rather than “Native Americans”; I use their term when referring to their tribe.) The analysis reveals that the Lumbee have been surprisingly innovative and heterogeneous. Explanations are both linguistic and extralinguistic. Insular groups do not face linguistic pressure to level intra- and inter-community differences or to curb internal innovations. In addition, insular groups are often more concerned with intra- than inter-group relations and hence with intra-group social and linguistic distinctions. The study also shows a lessening of inter- and intra-group dialect differences with increased inter-group contact. However, the Lumbee still preserve a degree of dialectal distinctiveness, indicating that the need to preserve cultural uniqueness may outweigh linguistic pressure to level out differences.

This article examines syntactic variation in an English dialect that has emerged from the process of language shift. South African Indian English has a wide variety of relative clauses that can be categorized as follows: (a) standard, (b) almost standard, (c) substrate influenced, (d) discourse governed. Of some interest is the occasional appearance of left-branching relative clauses and correlatives normally associated with object-verb (OV) languages. Social patterning of the relative clause is examined according to the parameter standard/nonstandard and according to the use made of certain relativization strategies by specific groups of speakers.

Traditional dialect boundaries in the United States have received renewed attention (Labov, 1991, 1994). Labov outlined three dialects of English (the Northern Cities Chain Shift, the Southern Chain Shift, and the Third Dialect), the boundaries of which are defined by chain shifts in the vowel system and roughly correspond to traditional dialectal boundaries defined through the bundling of lexical items (Kurath, 1949) and phonological isoglosses (Kurath & McDavid, 1961). Other research has suggested that the Third Dialect may be the most heterogeneous of these dialects, with speakers in different areas displaying widely disparate behaviors (see, e.g., Clarke, Elms, & Youssef, 1995; Di Paolo, 1988; Di Paolo & Faber, 1990; Labov, 1996; Moonwomon, 1987). The present article contributes towards a richer picture of the Third Dialect by offering the first systematic variationist analysis of speech in Pittsburgh, with a particular focus on three phonological processes: vocalization of /l/, laxing of /i/ before /l/, and laxing of /u/ before /l/. I argue that Veatch's (1991) model of English syllable structure provides a unified account of these seemingly unrelated phonological changes in Pittsburgh; the implications of this argument for further research on Pittsburgh speech are also noted.

One striking feature of Vimeu Picard concerns the regular insertion of epenthetic vowels in order to break up consonant clusters and to syllabify word-initial and word-final consonants. This corpus-based study focuses on word-initial epenthesis. It provides quantitative evidence that vowel epenthesis applies categorically in some environments and variably in others. Probabilistic analysis demonstrates that the variable pattern is constrained by a complex interplay of linguistic factors. Following Labov (1972a, 1972b) and Antilla and Cho (1998), I interpret such intricate grammatical conditioning as evidence that this variation is a reflection of a grammatical competence that generates both categorical and variable outputs, and I propose an account within the framework of Optimality Theory. An analysis of individual patterns of epenthesis by members of the community reveals that, even though all speakers share the same basic community grammar, their use of epenthesis differs qualitatively as well as quantitatively. I show that individual grammars can be derived from the community grammar, and that Optimality Theory allows us to formalize the idea that individual grammars constitute more specific versions of community grammars.

In studying language change, variationists are, naturally perhaps, more interested in the new, innovative form than in the old conservative one, and because of the actuation problem, investigations of changes in progress very rarely are able to shed light on the change in its very earliest stages. In this article, I suggest that we should perhaps pay more attention than we have at present to the origins of the change (in addition to its route and destination) and the nature of the conservative form if we are to chart ongoing changes in an accurate way. Here, I highlight an example of a feature of New Zealand English (NZE) (realizations of the mouth diphthong with front mid-open onsets) that has, until recently, been assumed to have resulted from a change of the Southern Shift-kind—a raising and fronting to [εʊ~εƏ]—but which, as I demonstrate using contemporary and past dialectological, as well as sociodemographic evidence, did not undergo this change in this way. Indeed, the supposedly conservative [aʊ] form has barely been used at all as a conversational vernacular variant in NZE. I argue here that the present-day NZE realization is far more likely to be the outcome of a process of dialect leveling operating on the mixture of forms brought to New Zealand by British and Irish migrants in the 19th century. The moral of the story is that if we think we observe a change in progress from A to B, we need to provide evidence not just of the existence of B, but also of the prior existence of A.

A social dialect survey of a New Zealand community documented a change in progress in the pronunciation of the vowels in words such as air and ear. The data support a tentative interpretation that a shift to the variant with the closer onset for AIR words was initiated by middle-aged Pakeha women. This was not followed in any strength by other social groups. More recently there has been a change in the opposite direction, with EAR words being realized by a variant with a more open onset. This is being led by young, working class Pakeha speakers, possibly the women. For some speakers, merger of the two diphthongs is close to completion. The change thus seems to have begun in the past 40 years, but its direction has been reversed in the past 20 years. With the strong following among young speakers for the downward shift toward [eə], this seems the likely direction in which change will continue.

This paper is equal parts methodological recommendation and an empirical investigation of the time dimensions of linguistic change. It is increasingly common in the sociolinguistic literature for researchers to utilize speech data that was collected over the course of many decades. These kinds of datasets contain three different time dimensions that researchers can utilize to investigate language change: (i) the speakers’ dates of birth, (ii) the speakers’ ages at the time of the recording, and (iii) the date of the recording. Proper investigation of all three time dimensions is crucial for a theoretical understanding of the dynamics of language change. I recommend utilizing two-dimensional tensor product smooths, fit over speakers’ date of birth and the year of the recording, to analyze the contribution of these three time dimensions to linguistic changes. I apply this method to five language changes, based on data drawn from the Philadelphia Neighborhood Corpus. I find relatively weak evidence for lifespan effects in these changes, robust generational effects, and in one case, evidence of a zeitgeist effect.

This paper reports a speaker evaluation experiment that investigated the competition between three regional accents of Standard Dutch and references to the speaker's profession as determinants of attitude formation. A stratified sample of listener-judges rated speech stimuli that were presented in two guises, a neutral guise and a teacher guise (the latter containing multiple references that revealed the speaker to be a high school teacher of Dutch). The experimental findings corroborate our earlier claim that regional flavoring is embedded in lay conceptualizations of Standard Dutch. Although teachers of Dutch may be the last “gatekeepers” of the standard in the Low Countries, they are not automatically downgraded when they have a regional accent: What matters is, clearly, which accent they have. Analysis of the ratings further suggests a hierarchical relation between accent and occupation as perception triggers: Even though regional accent clearly is the stronger attitude determinant, it does not suppress occupational information but interacts with it to generate richer social meaning.

In the wake of numerous analyses of vowels in African American English (AAE), this study examines acoustically the phonetic production of a consonant—the word-initial lateral /l/—across several generations of speakers from a long-standing African American community in central North Carolina. The results of the study show that /l/ is darker in younger AAE speakers than in older ones, independent of phonetic context. Comparisons with ex-slave recordings suggest that a light variant of /l/ may be a substrate feature of AAE that has changed in recent decades. Additional comparisons with regional European Americans suggest that the darkening may be due to convergence with majority American English dialects.

One of the core problems in second language acquisition theory is how to describe and explain the highly variable (yet rule-governed) speech of second language learners. Is such variation simply random and most likely due to the first language's interference, or is it governed (at least in part) by general rules that reflect language universals? Within a multidisciplinary approach to the analysis of variability in second language acquisition, this article addresses these questions in the context of a cross-sectional study involving the acquisition of word-final stops by Brazilian Portuguese speakers learning English in a classroom environment. The study follows a sociolinguistic approach for data collection and the analysis is couched within a stochastic version of Optimality Theory.

Sociolinguistic situations that involve interplay between standard and nonstandard varieties have posed certain difficulties for the analysis of variation at the morphosyntactic level (Harris, 1984; Milroy, 1987). In particular, there is the problem of defining the scope and relevant contexts of morphosyntactic variables—a problem closely linked to that of identifying the semantic equivalence in terms of which the variable is defined (Cheshire, 1982; Romaine, 1984). This article addresses such issues by examining variation in the use of perfect have and its alternatives in the Trinidadian creole continuum. The alternatives to have include other Standard English (SE) forms (e.g., the past), as well as forms like perfective Ø, completive done, and others which function in the Trinidadian Creole (TC) system. The analysis is based on data from a sample of speakers from different social backgrounds. The variation described here is part of a wider pattern involving different strategies for expressing the various meanings associated with the perfect in SE, namely, the “continuative,” “experiential,” and “resulative” interpretations of have. TC employs different strategies for expressing these meanings. Perfective Ø is used (along with apporopriate adverbial specifications) to convey the experiential meaning; progressive -in, Ø copula, or Ø-marked statives convey the continuative sense; and completive done competes with prefective Ø to convey the resultative sense. All of these may vary with have. I discuss at length the procedures that are used to define the variable and the equivalents of have that are identified. A quantitative analysis reveals clear patterns of variation according to class, subcategory of perfect, and predicate type. The differences in the use of have are explained in terms of varying degrees of influences from vernacular norms among the social groups, as well as the availability of alternative choices to have within the pragmatic and grammatical environments. Finally, I indicate the implications of this investigation for the study of morphosyntactic variation in other divergent dialect situations. Specifically, the findings are relevant to situations, such as AAVE, which involve variation between standard and nonstandard strategies for expressing the meanings associated with the perfect and other morophosyntactic categories.

This study examines cross-generational differences in the realization of an English phonological contrast by bilingual Polish Americans in New York City. I analyze the production of voice onset time (VOT) in underlying stops, as in tin and den, and stops derived from interdental fricatives, as in [t]in for thin and [d]en for then, in an English-only reading task. Generation one exhibits VOT “interference” for both stop types, with a bias toward interference for voiced stops. Generation two “transfers” Polish-like VOTs to derived stops. I argue that the cross-generational progression from the global effects of interference to the focused presence of transfer is filtered through L1 markedness and reflects speakers' growing sensitivity to L2 phonology and social considerations. The observed asymmetries in the distribution of interference/transfer are unaccountable by existing models of L2 acquisition and motivate a view of L1/L2 phonetic categories as governed by a variable grammar with access to phonological and social information.

This paper examines the phonological leveling and diffusion of variants of /t/ in Liverpool, northwest England, and two localities in its hinterland. We show that lenited realizations of /t/, thought to be historically restricted to Liverpool, are increasing over time and spreading over geographical space. We explore Labov's (2007) claims that linguistic changes that progress via transmission, within a speech community, are reproduced in all their structural complexity, whereas changes that spread across speech communities, via diffusion, are “simplified” en route. We find support for these hypotheses. Using a comparative sociolinguistic methodology, we show that the linguistic constraints operating on the realization of /t/ as [h] in Liverpool have remained stable over time, while those in a nearby town—Skelmersdale—seem to have simplified. However, we show that not all speakers from Skelmersdale share the same constraints on this variable form, and we connect this with speakers' positive or negative attitudes toward Skelmersdale or Liverpool.

This account deals with a stage of chain shifting in North American English in which the usual tendency to preserve margins of security does not seem to apply. Two vowels follow a collision course with a high degree of overlap. The question then arises as to whether a small difference in vowel length can effectively preserve a phonemic distinction.This paper was first given at NWAV 33, in Ann Arbor, Michigan.

The system of stative possession has been subject to variation and change since at least the Early Modern period, with have got rising in frequency in British and Antipodean varieties of English. In Canadian English, as represented by data from the largest city, Toronto, have predominates. Nonetheless, the full set of constraints previously reported for this variable are operative, corroborating the longitudinal maintenance of linguistic factors across time and space (Kroch, 1989). At the same time, variation among possessive forms is conditioned by robust sociolinguistic patterns. Have is correlated with education and with female speakers, whereas less-educated men favor have got and got. Such findings demonstrate that the domination of one form or another in a variable system can be the result of historical accident, in this case a founder effect at a particular point in history, and that the social value of forms is a product of local circumstances at the time of change.

This study examines subject personal pronoun expression in the Spanish of the west-central highlands of Puerto Rico. Although rates of s-deletion are comparably high, rates of overt subject expression are shown to be much lower than rates reported for varieties of coastal Puerto Rican Spanish and U.S. mainland Puerto Rican Spanish. The linguistic constraints on overt versus null pronoun usage in the data are shown to coincide to a very large extent with constraints identified for other Puerto Rican dialects and also Castilian Spanish in central Spain, whereas of the social factors, only the distinction between farmers and nonfarmers is significant. The study suggests that, if rates of personal subject pronoun expression are an indication of dialectal variation, the rates presented here for this syntactic phenomenon represent the continuing effects of a conservative dialect in the interior of the island of Puerto Rico.

In this article I examine the ongoing loss of palatalization assimilation in Modern Standard Russian within the framework of Optimality Theory (OT) (McCarthy & Prince, 1993a; Prince & Smolensky, 1993) and show that this theory offers new, meaningful explanations for the role of markedness and naturalness in the mechanism of a sound change. I also argue that OT provides new possibilities for relating quantitative patterns to the formal principles of grammatical organization. In particular, in OT, suggested cross-linguistically invariant relations between phonological factors predict a general pattern of influence that these factors have on the quantitative outcome of a change. I suggest that a change operates as a gradual weakening or strengthening of whole subhierarchies of constraints with universally fixed rankings (constraint families, which implement markedness scales in OT). In an examination of variable data, I argue that the major differences in quantitative patterns mirror the fixed constraint ranking within such constraint families and constraint violation/nonviolation in OT grammar. I also discuss the problem of modeling variable data within OT, which, like other formal phonological theories, permits no output variation. I examine a grammar competition model explored in OT literature and argue instead for constraint competition with constraint weight in production.

A number of studies of African American communities show a tendency to approximate the phonological patterns of the surrounding mainstream white community. An analysis of the vowel systems of 36 African American speakers in the Philadelphia Neighborhood Corpus compares their development over the 20th century with that of the mainstream community. For vowels involved in change in the white community, African Americans show very different patterns, often moving in opposite directions. The traditional split of short-a words into tense and lax categories is a more fine-grained measure of dialect relations. The degree of participation by African Americans is described by measures of bimodality, which are applied as well to the innovative nasal short-a system. The prototypical African American speakers show no bimodality in either measure, recombining the traditional tense and lax categories into a single short-a in lower mid, nonperipheral position. The lack of relation between the two short-a systems is related to the high degree of residential segregation, in that linguistic contact is largely diffusion among adults rather than the faithful transmission found among children.

The question of whether or not grammatical factors can condition or block sound change has been discussed from many perspectives for more than a century without resolution (Melchert, 1975). Here we consider studies of sound change in progress which show that words or phrases that are used frequently in the phonetic environment for change undergo the change before those whose use is less frequent in these contexts. Because words of different categories and with different structures also have different distributions, they may occur preferentially in certain phonetic environments. Thus, some apparent cases of influence by grammatical and lexical factors can be explained by phonetic factors if we expand our notion of phonetic environment to include frequency within the environment for change, which includes the segmental environment as well as factors that affect the degree of prominence a word receives in context.