To verify anecdotal claims regarding the nature of unstressed vowel deletion in Azorean (European) Portuguese, conversational data from a native speaker of the island of Faial have been analyzed to determine the segmental and prosodic contexts favoring elision. Results of a quantitative analysis indicate that unstressed [u] and schwa are the most likely vowels to be deleted; moreover, deletion is highly favored when the unstressed vowel occurs in word-final position at the end of an utterance. Factors such as rhythmic preservation, syllable structure, and functional load are discounted in the analysis, suggesting that vowel deletion is essentially a word-based variable process in the language.

English coronal stop deletion is constrained by the preceding segment, so that stops and sibilants favor deletion more than liquids and nonsibilant fricatives. Previous explanations of this constraint (e.g., the sonority hierarchy) have failed to account for the details, but we show that it can be comprehensively treated as a consequence of the obligatory contour principle (OCP). The OCP, introduced to account for a variety of categorical constraints against adjacent identical tones, segments, and so forth, can be generalized as a universal disfavoring of sequences of like features: *[αF] [αF]. Therefore, coronal stop deletion, which targets the set of segments /t, d/ defined by the features [−son, −cont, +cor], is favored when the preceding segment shares any of these features. But this requires adopting the assumption of inherent variability and interpreting the OCP as a probabilistic constraint with cumulative effects (the more shared features, the greater likelihood of deletion). This suggests an attractive theoretical integration of categorical and variable processes in the grammar.

This study investigates the contraction of negatives in a carefully chosen corpus of discourse and writing, to permit comparison of the relative influences of various linguistic and social parameters on contraction. Evidence is presented that negative contraction is conditioned by interactional and other register variables. The point is made that the pragmatic as well as morphological interpretation of negatives entails that negative contraction and auxiliary contraction should be distinguished from each other. Although a Cognitive Prominence Principle predicts noncontraction when the negative conveys semantically focal information, a Social Agreement Principle predicts contraction. This is because it would be face-threatening (and, therefore, in conversation analysis terms “dispreferred”) to focus on disagreement, which is most often the semantic information conveyed by negatives. This hypothesis is examined using corpora which differ along several dimensions. The most important of these (for this study) appear to be the interactional versus informational register dimensions (Finegan, 1994). Data from instructional (workshop presentations), confrontational (political debates), and casual conversational material are contrasted with comparable reading style materials. The following general results are predicted. The Cognitive Prominence Principle will take over in informational contexts when disagreement is acceptable or neutralized. The Social Agreement Principle will take over in more interactional contexts where disagreement is not acceptable. The results are of interest to the student of focus, the sociolinguist concerned with dialect, register, and style variation, and even the speech technician.

Data from fieldwork and a telephone survey is used to demonstrate that the merger of the vowels in words such as cot and caught, traditionally considered a defining characteristic of the speech of western Pennsylvania, is well established in the former anthracite mining towns of eastern Pennsylvania. Demographic and linguistic data indicate that the merger arose independently in the anthracite towns as a result of massive in-migration of speakers of Polish and other non-Germanic languages around the turn of the century. Having identified the proximate cause of the merger makes it possible to predict the presence or absence of the merger in various towns using demographic data alone.

The migration of people to the Sunbelt in the United States constitutes a major demographic shift, but has received little attention from language variationists. In Texas, this migration has led to a split of the Anglo population of the state into two dialects, a rural dialect and a metropolitan dialect. Evidence from a random-sample survey of Texas and from a systematic set of surveys of high schools in the state shows that young rural Anglos preserve two stereotypical features of the Texas accent, monophthongal /ai/, as in night, and lowered onsets of /e/, as in day, while young Anglos from metropolitan centers lack these features. This difference, which is absent among middle-aged and older native Texan Anglos, appears to have resulted from the fact that in-migration from other parts of the country is concentrated in metropolitan centers, especially suburbs.

We examine a pattern of end-of-word deletion in Faetar, a Francoprovençal dialect spoken in southern Italy, considering synchronic variants like [brókələ] ˜ [brókəl] ˜ [brókə] ˜ [brok] ‘fork’. We use the word “deletion” as a synchronic description of the facts; speakers do not always phonetically produce everything in the input form, assuming that the input form is the longest form ever produced. Optimality Theory accounts for this type of variation by positing different rankings of the constraint hierarchy, each of which produces a different optimal output. The predication of alternate constraint rankings within a single dialect, however, poses problems for Optimality Theory as it has been formulated, necessitating numerous grammars for each speaker. We propose floating constraints (Reynolds, 1994), whereby some particular constraint within a single grammar may be represented as falling anywhere within a designated range in the ranking hierarchy. In a previous study (Reynolds & Nagy, 1994) we showed that this model accounts for the distribution of types of output forms produced. Here, we analyze a corpus of 624 tokens from 40 speakers and show that the pattern of distribution of tokens is accounted for as well: the number of rankings that produce each output form is closely correlated to the number of output forms that occur in the data set.

Optimality theory (OT) (McCarthy & Prince, 1993; Prince & Smolensky, 1993) has been proposed as a constraint-based theory of phonology in which the phonological facts of each language are accounted for by a language-specific ordering of a universal inventory of constraints. The constraints, expressing desirable (i.e., optimal) phonological states, evaluate possible candidate forms, selecting the optimal output. Any constraint may be violated by a surface form if it is overridden by a higherranked constraint; the ordinal sequence of constraints provides a weak quantification of constraint effects. Variability has been treated within OT by varying constraint orders. This model is analogous in several important respects to the variable rule model (VR) of Labov (1969) and Cedergren and Sankoff (1974). In VR, variable constraints express desirable phonological states which are variably realized on the surface, when not overridden by other constraints; constraints are probabilistically quantified. This article compares the OT and VR models, arguing that the VR model is superior on theoretical and empirical grounds: constraint effects in VR are stable, transparent, and learnable. Moreover, the probabilistic treatment of constraint effects allows VR to model successfully cases in which multiple violations of a single constraint lead to a cumulative reduction in likelihood of a form; such cases cannot be efficiently treated in OT.

Ever since Labov, Cohen, Robbins, and Lewis's (1968) pioneering study, it has been commonplace to set aside certain tokens in analyzing variability in the English copula as “don't count” (DC) forms. These cases are most often occurrences of the copula that exhibit categorical behavior (as with the full copula in clause-final position), as well as those that are ambiguous or indeterminate. In this article, I propose a set of copula forms that should be set aside from variable analysis as instances of DC forms to allow for systematic comparisons among studies. I review the major alternative descriptions of DC copula cases in the literature and analyze the behavior of the traditional DC categories. New data are presented to support the exclusion of particular DC cases from analyses of copula variability. Among the conclusions are that [was], [thas], and [is] should be excluded from quantitative analyses of variation in the copula because of their invariant status, and that a number of tokens commonly included (e.g., questions) should be excluded on various grounds.

Use of discourse markers by 17 speakers of Anglophone Montreal French (AMF) showed great variation in individual repertoires and frequency of use. Only five subjects manifested rates of usage comparable to those of native speakers or to their own LI usage in English. In decreasing order of frequency, the speakers used tu sais ‘y'know’; là ‘there’ (the most frequent among L1 Montreal French speakers); bon ‘good’, alors ‘so’, comme ‘like’, and bien ‘well’; and the local discourse conjunction fait que ‘so’. The subjects occasionally made use of the English markers you know, so, like, and well. Québécois French markers with no English equivalent were used by the speakers who had been exposed to French in their early childhood environment. The one marker that showed influence from English was comme, apparently calqued on English like. Overall, frequent use of discourse markers correlated only with the speakers' knowledge of French grammar – evidence that a higher frequency of discourse marker use is the hallmark of the fluent speaker. As a feature that is not explicitly taught in school, mastery of the appropriate use of discourse markers is thus particularly revealing of the speakers' integration into the local speech community.

This article, which examines the system of relative markers in Early African American English as documented in the Ex-Slave Recordings (Bailey et al., 1991), is intended as a contribution to two areas of research: African American Vernacular English and the system of relativization in English. We found a significantly higher incidence of zero marking in adverbial relatives than in non-adverbial relatives. Among non-adverbial relatives, a variable rule analysis showed that non-humanness of the head as well as the function of the head as subject complement or subject in an existential sentence strongly favored zero relatives, and that prepositional complement heads disfavored zeroes. The lack of wh-relatives aswell as the frequency of zero subject relatives is interpreted as evidence that African American Vernacular English is a dialect of English.

In Spanish, subject pronouns may be realized phonetically or as null. Previous research on a wide range of dialects has established a rich patterning of constraints on this variation, with switch reference as the first order linguistic constraint. Recently, however, Paredes Silva (1993), in a study of written Brazilian Portuguese, suggested a more fine-grained analysis of null subject pronoun variation based on a model of discourse connectedness. This study tests Paredes Silva's model on the oral and written Spanish narratives of northern California Mexican-descent preadolescents. Results of multivariate analysis indicate that discourse connectedness provides a more fine-grained account of pronoun variation in the Spanish of these children than switch reference. The study also considers the effect of morphological ambiguity. We suggest that tense and aspect features provide a better explanation for the higher incidence of overt pronouns with imperfect, conditional, and subjunctive verb forms than the functional compensation hypothesis. Finally, we examine pronoun variation across immigrant generations. The results indicate that children with the greatest depth of ties to the United States are less likely to use overt pronouns than children born in Mexico.

Long, complex phrases tend to come at the ends of clauses; this is called “endweight.” A variety of characterizations of weight have been proposed in the literature, but none has been sufficient to cover the full range of attested cases of end-weight. Corpus data on heavy NP shift, the dative alternation, and particle movement indicate that there are several structural measures of weight that are highly correlated with constituent ordering. Proposed explanations for endweight have been based on parsing considerations, largely ignoring the speaker; but what facilitates parsing does not always help in production. Examination of phenomena where these interests do not coincide indicates that the demands of sentence planning provide a better explanation for end-weight than parsing. Finally, accounts of end-weight cannot be purely structure-based, but must take lexical factors into consideration.

This article provides evidence from a small city in the northern United States for Canadian raising, the raising of /au/ and /ai/ to [AU] and [Ai]. As hypothesized, /ai/-raising is much more frequent and regular, but /au/-raising also occurs. Speech style is statistically significant only for /ai/-raising. The raising of /ai/ is not significant for gender or age, but /au/-raising seems to be led by young women. This evidence suggests that, while /ai/-raising has been fully entrenched in this phonological system for some time, /au/-raising is a new and separate change in progress. The /au/-diphthong is also often fronted when raised, a finding similar to several Canadian studies (Chambers, 1981a; Hung, Davison, & Chambers, 1993). This suggests that the Canadian varieties studied may not, as previously suggested (Chambers, 1981b), be conforming to the U.S. norm, but instead that the two varieties are converging.

Recent work in the acquisition of dialect has shown that children acquire patterns of stable variation and completed sound changes at a very early age. The present study looks at the acquisition of new sound changes by 3- and 4-year-old children. The results were consistent with previous findings in that, with critical exceptions, the children had acquired the changes. These results underscore the point that the dialect transmission period begins early – before the age of maximal peer group influence. In addition, as Labov (1994) suggested, one possible explanation for the frequent finding that women lead in changes from below is the asymmetry of the early childcare situation, which could advance female-dominated changes. The finding that children begin to learn their community norms at a very early age supports the possibility that this childcare asymmetry affects the progress of sound changes over time.

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.

This article describes the use of negation in three corpora representative of early to mid-19th century African American English: the Ex-Slave Recordings (Bailey, Maynor, & Cukor-Avila, 1991), the Samaná Corpus (Poplack & Sankoff, 1981), and the African Nova Scotian English Corpus (Poplack & Tagliamonte, 1991). The specific structures studied are the negative form ain't, negative concord to indefinites and to verbs, negative inversion, and negative postposing. It is found that Early African American English (i) is far more conservative than modern African American Vernacular English; (ii) is generally similar to Southern White Nonstandard English; and (iii) displays no distinct Creole behavior. In other words, our study suggests that the negation system of Early African American English derived directly (i.e., without approximation or creolization) from colonial English, contrary to the findings of Rickford (1977, 1995), Labov (1982), Winford (1992), De Bose and Faraclas (1993), DeBose (1994), and others.