When one form replaces another over time in a changing language, the new form does not occur equally often in all linguistic contexts. Linguists have generally assumed that those contexts in which the new form is more common are those in which the form first appears and in which it advances most rapidly. However, evidence from several linguistic changes (most importantly the rise of the periphrastic auxiliary do in late Middle English) shows that the general assumption is false. Instead, at least for syntactic cases, change seems to proceed at the same rate in all contexts. Contexts change together because they are merely surface manifestations of a single underlying change in grammar. Differences in frequency of use of a new form across contexts reflect functional and stylistic factors, which are constant across time and independent of grammar.

A set of ordered rules for generating variants of a variable determines (a) underlying/surface distinctions among some of the variants and (b) a hierarchical classification of the variants. In the analytical framework of variable rules, frequency data on variant occurrences in context bear only on (b) and not on (a). We provide a combinatorial characterization and enumeration of the set of classifications on n variants, the set of underlying/surface configurations, and the set of rule orders. We describe the statistical and computational techniques for generalizing variable rule analysis to the inference of rule order. These procedures are applied to the problems of the reduction of syllable-final consonants <s>, <n>, and <r> in Caribbean Spanish (n = 3, 4, 6 variants, respectively). Previous analyses have tended to assume that successive weakenings occur in an intrinsic order determined by phonological strength. Our results show that aspiration and deletion can indeed be seen to be intrinsically ordered in both <s> and <r> reduction, though an unordered analysis is equally likely in the case of <s>. On the other hand, velarization and deletion of <n> are unordered, while vocalization is a subsequent process, independent of the other two. Similarly, spirantization, aspiration, and lateralization of <r> are unordered, as confirmed by data sets from both Puerto Rican and Panamanian speakers. Furthermore, with both <n> and <r>, intrinsically ordered rule schemata proved to be extremely unlikely by statistical criteria. Syllable-final consonant reduction then consists of largely independent processes, most of which occur simultaneously.

Although it has generally been recognized that Old English was a verb-final language with verb-seconding, the existence of clauses with main verb complements and adjuncts appearing after the otherwise clause-final verb seems to contradict the hypothesis that the language was strictly verb-final in underlying structure. There are three possible analyses to explain these clauses: variable word order in the base, leftward verb movement, and rightward movement of NPs and PPs. In this article, we demonstrate that only the third analysis adequately explains the data of the Early Old English poem Beowulf. Moreover, by investigating the mapping between syntactic structures and metrical units, we provide evidence for two types of rightward movement with two distinct structures: heavy NP shift, with a characteristic major intonational boundary between the main verb and the postposed NP, and PP extraposition, where the intonational boundary was much less common.

Speaker's sex has emerged as one of the most important social factors in the quantitative study of phonological variation. However, sex does not have a uniform effect on variables or even on variables that represent sound change in progress. This is because sex is not directly related to linguistic behavior but reflects complex social practice. The correlations of sex with linguistic variables are only a reflection of the effects on linguistic behavior of gender — the complex social construction of sex — and it is in this construction that one must seek explanations for such correlations. Sociolinguists generally treat sex in terms of oppositional categories (male/female), and the effects of sex on variation are generally sought in linguistic differences between male and female speakers. However, because gender differences involve differences in orientation to other social categories, the effects of gender on linguistic behavior can show up in differences within sex groupings. Data on sound changes in progress (the Northern Cities Chain Shift) among Detroit area adolescents show that gender has a variety of effects on variables and that the significance of gender in variation cannot be reduced to notions of male or female speech as “more or less conservative.”

À partir de deux ensembles de données recueillies principalement auprés des mêmes locuteurs à 13 ans d'intervalle, le cheminement de cinq marqueurs discursifs dans la communauté francophone de Montréal est analysé. II s'agit de alors, tu-sais, je-veux-dire, tu-sais-veux-dire et disons. L'alternance des formes restrictives juste, rien que et seulement est aussi examinée. Les résultats obtenus montrent clairement que la différenciation selon l'âge observée à partir de données synchroniques peut (a) correspondre à des emplois qui se modifient au cours de la vie des locuteurs: c'est le cas de alors dont le taux augmente en vieillissant; (b) indiquer un changement en cours: c'est le cas de tu-sais qui a connu une augmentation significative entre les deux années, et de juste qui tend a supplanter les autres variantes; (c) traduire une mode associée à une génération spécifique (cf. disons). Par ailleurs, certaines variables connaissent une progression importante dans la communauté sans être associées à un sousgroupe social particulier (le marqueur tu-sais-veux-dire).

Some recent research proposes that changes observed in the morphology and syntax of Vernacular Black English (VBE), combined with the failure of VBE to participate in regional changes in the pronunciation of white speech, constitutes evidence for the divergence of VBE from white and standard varieties. The present article presents quantitative evidence for a change in VBE which appears to involve increasing similarity between VBE and other varieties: a marked tendency toward higher, fronter realizations of the final /i/ in words like happy in the speech of younger blacks. Evidence for the character of the change comes from English dialectology, historical phonology, and research conducted by the author on the speech of residents of East Palo Alto, California. It is proposed that whereas black and white varieties may remain distinct, undergoing some changes separately, this need not be regarded as absolute divergence. Distinct varieties in contact may, for complex sociological and linguistic reasons, be expected more realistically to appear to diverge and converge simultaneously.

This article contributes to the understanding of the origin and function of verbal -s marking in the Black English grammar by systematically examining the behaviour of this affix in two corpora on early Black English. To ascertain whether the variation observed in (early and modern Black English) -s usage has a precedent in the history of the language, or is rather an intrusion from another system, we focus particularly on the linguistic and social contexts of its occurrence, within a historical and comparative perspective. Our results show that both third person singular and nonconcord -s are subject to regular, parallel environmental conditioning. The finding that both insertion and deletion are conditioned by the same factors suggests that verbal -s marking is a unitary process, involving both concord and nonconcord contexts. Moreover, the (few) variable constraints on verbal -s usage reported throughout the history of the English language remain operative in early Black English. These results, taken in conjunction with indications that -s marking across the verbal paradigm was a prestige marker in the dialect at some earlier point in time, lead us to hypothesize that the contemporary pattern might be a synchronic reflex of the constraint ranking on -s usage in the varieties of English that provided the linguistic model for the slaves. Many of the conditioning effects we report would have been subsequently overridden by the grammaticalization of -s as the Standard English agreement marker. We conclude that present-tense marking via verbal -s formed an integral part of the early Black English grammar.

Spanish dialects may differ according to the behavior of voiced obstruents following r, l, s, and the glides. For most dialects, for example, Mexican, fricatives occur in these environments; in some, such as highland Colombian, stops tend to occur. Another feature distinguishing dialects is s-aspiration/deletion. Several Central American dialects are of particular interest because they are both s-aspirating/deleting and stop-conserving in the series of postconsonantal environments (Canfield, 1961/1981). Detailed examination of spirantization in these dialects, as well as of the particular question of what occurs after /s/ → [s] [z] [h] [Ø] has been nonexistent. Obviously, if s is deleted before a voiced obstruent, the resulting environment of the obstruent is intervocalic, which is a favored environment for spirantization. For this investigation, 14 Honduran speakers were interviewed using standard sociolinguistic interview techniques. Tokens were transcribed and categorized according to whether s-aspiration/deletion had applied, whether spirantization had applied, and by preceding/following environment. It is shown that (a) in Honduran Spanish the group r, l, glides, and s inhibit spirantization variably, much as they do in highland Colombian Spanish; (b) neither voicing, aspiration, nor deletion of (s) favor spirantization, despite the apparently favorable (surface) environment created when deletion applies; (c) even though an analysis of (b) that orders spirantization before deletion is apparently explanatory, a view that attributes spirantization to syllable structure offers a more comprehensive explanation of what happens when the two variable processes intersect by applying to adjacent segments.

This article analyzes the assibilation of /r/ among young people in San Luis Potosí, Mexico, in light of previous research on women's speech in language change. It is demonstrated that assibilation, an innovation known to have first appeared in the speech of women of the middle and upper social echelons, is closely associated with sex, sociocultural level, and attitude toward traditional male and female roles. These attitudes are suggested as a factor that plays an important role in the dynamics of the change, showing opposite effects among young men and women. That is, young men with traditional attitudes assibilate least, whereas young women with traditional attitudes assibilate most frequently. Parallels between this study and one of a similar innovation in Argentinian Spanish suggest a generalized pattern of change in which variables introduced by women of the middle and upper social echelons become markers of gender display in the lower classes, where they grow to be favored by women and avoided by men. The discovery of the role of attitude toward traditional sex roles in this pattern of change is unique to the present study.

Determining the function of verbal -s in the Black English Vernacular (BEV) has been a major problem in sociolinguistics. Linguists have offered four answers to questions about the feature's origins and function, with -s seen as a case of hypercorrection, as a marker of durative/habitual aspect, as a variable marker of present tense (with the variation stemming from dialect mixture), and as a marker of historical present regardless of person and number. This article argues that confusion about -s results largely from four mistaken assumptions about it: (1) that -s did not exist in earlier varieties of BEV, (2) that the use of -s in the plural bears no relation to its use in the singular or to other processes such as the use of is as a plural or copula and auxiliary absence, (3) that -s has always functioned simply as a person/number marker in white vernaculars, and (4) that any role white vernaculars may have played in the variability of -s was a consequence of regional variation brought to the United States. This article addresses these assumptions by examining the function of -s, both in the singular and plural, and of plural is in the Cely Letters, written between 1472 and 1488, and by comparing the results to similar data in black and white vernaculars. The analysis shows that in the Cely Letters the presence of an NP subject strongly favors the occurrence of both singular and plural -s and also plural is. The same constraint operates on the same forms in older black and white vernaculars, and it affects copula and auxiliary absence as well. In the speech of younger blacks and whites, this constraint has begun to disappear as -s has become solely a marker of person/number agreement in white speech and as -s itself has disappeared in BEV.

Though the diachronic dimension of linguistic variation is often identified with linguistic change, many stable linguistic variables with no synchronic motivation show historical continuity with little change over long periods of time. Children acquire at an early age historically transmitted constraints on variables that appear to have no communicative significance, such as the grammatical conditioning of (ing) in English. Studies of (td) and (ing) in King of Prussia families show that children have matched their parents' patterns of variation by age 7, before many categorical phonological and grammatical rules are established. Some dialect-specific and socially marked constraints are acquired before constraints with general articulatory motivation. Constraints on (td) appear in the speech of a 4-year-old, but there is no evidence in the productions of a 2-year-old child in the same family.

Brazilian Portuguese has a vowel harmony rule which consists of the optional raising of the pretonic e and o vowels to i and u, respectively, if there is a high vowel in the following syllable. This is an old rule, the origin of which can be traced back to the Latin of the 4th century a.d. Because the rule is subject to multiple conditioning by both linguistic and social factors, it can be treated as a variable rule.

This article investigates the grammaticalization process of liao as a main verb in 10th-century vernacular texts to le as an aspectual particle in modern Chinese. I propose that two processes are involved. First, with the “resultative construction” coming into existence in the language, some instances of liao were reanalyzed as the phase complement of the new morphological construction. Second, other instances of the verb began to lose their verbness by taking sentential subjects and occurring in temporal clauses only. These processes gave rise to the positional change of liao from after the complement of the verb to before the complement of the verb of le.