Sociolinguistic patterns in language change are largely based on generalizations from linguistic variables consisting of lexemes or morphemes. This article takes a diachronic, corpus-based approach to the diffusion of a change in a more extensive morphosemantic function consisting of several linguistic subsystems. It focuses on the pronoun variants he, they, and those used for human indefiniteness in two contexts: (a) epicene anaphoric uses with indefinite pronouns and (b) cataphoric personal references. The quantitative corpus analyses show that the pronoun selection in Early and Late Modern English developed a greater tendency to use one pronoun type over the other in both contexts. The main data come from the Corpus of Early English Correspondence and its Extension. Statistical analyses compare the observed correlations of the pronouns with a set of social, external variables and language-internal factors. This article concludes that it is possible to establish sociolinguistic patterns in larger shifts if we account for the closely related internal developments in the language.

Standard Hungarian has a short front unrounded low vowel phoneme /ε/, but certain dialects have a mid /e/ as well. The standard /ε/ evolved from a merger of the two phonemes. A formal reading of word groups and a same/different listening test revealed that Hungarian–Americans in South Bend, Indiana, exhibit a continuum between the /ε/ dialect and the /ε/versus /e/ dialect, showing important differences between the informants' perception and production. The Hungarian–American data and metropolitan Hungarian data were compared to Labov, Karen, and Miller's (1991) findings on near-mergers. The importance of investigating perceptual and productive systems separtely is emphasized.

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.

Although the sound system of Indian English has been the object of numerous publications over the years, there has been a remarkable scarcity of variationist sociolinguistic research carried out on the topic. The present study addresses this gap by describing the short front vowels of 22 lifelong English-speaking Delhi residents born between 1948 and 1992. Focusing more specifically on variation in the relative configuration of trap /æ/, dress /ɛ/, and kit /ɪ/, the study provides apparent-time evidence for a series of interrelated changes affecting the system. Those include an ongoing lowering of /æ/ and /ɛ/, as well as age-related variation in a previously unreported allophonic split of /ɪ/. I argue that these apparent-time patterns are amenable to an analysis in terms of chain shift, and I discuss the implications of such a claim, linking the phenomenon described to similar patterns reported in various other parts of the English-speaking world.

In this study, we investigate first-person-singular subject expression in Louisiana French. This variety is undergoing language death and features extreme variation, with twelve first-person-singular subject forms identified within our corpus. We demonstrate that variationist methods are robust for examining such variation in obsolescing languages, and we provide a model for undertaking such analyses. Examining different aspects of our data, we fit two mixed-effects models, one that analyzes the four most frequent phonological variants of the atonic pronoun je ‘I’ and the other that focuses on the tonic pronoun mon ‘me.’ Several linguistic and social factors predict the use of these subject forms, supporting the claim that variability in declining languages is systematic, just as variation in healthy languages is. We argue that variationist methodologies have contributions to make to research on obsolescing languages and that variationist examinations of endangered and minority languages can provide methodological and theoretical contributions to the study of language variation and change more broadly.

This article focuses on change in the “short-a system” of New York City English (NYCE). Recent results suggest that a complex set of tensing rules traditionally described for NYCE are being replaced by several simpler systems. This article reports on a study of this change using a recently developed large audio-aligned parsed speech corpus (CoNYCE). This change is similar to the simplification reported for Philadelphia by Labov et al. (2016). Unlike in the Philadelphia case, however, our results do not show evidence of a single abstract process of change. Our findings, rather, suggest at least two separate changes in the community—one affecting short-a in prenasal contexts and a second affecting pre-oral obstruent contexts. In addition, the results suggest an additional independent process of lowering and retraction affecting short-a sounds in contexts not targeted by the process of phonological reorganization, that is, “trap-backing.”

The role of outliers in vowel distributions is examined through an experiment that registers subjects' assessments of symmetrical and asymmetrical distributions. Subjects first scaled their subjective impressions of a series of six resynthesized tokens of the word bad ranging from low front to upper mid front. They were then asked to register on this scale their overall impressions of four series of five phrases including bad: low symmetrical, low with a very low outlier, high symmetrical, high with a very high outlier. Subjects show the capacity to integrate outliers into their overall assessments in a manner consistent with their acoustic properties. The effect of low outliers was significantly greater, reflecting the socially marked status of this form in the Philadelphia community.

This article discusses the notion of regional standards of a language with regard to modern Danish. A regional standard is defined relative to a spectrum of varieties ranging from the national standard to the classical local dialects. We define the national standard, in accordance with Brink (1988), as the variety which has no localizable features. A local dialect is defined, in accordance with Danish dialectological tradition, as a combination of features which are inherent and not derived from the national standard. With evidence from studies of Danish regional variation we show that, for a geographical entity, a statistically determined boundary can be established in the range between these two extremes. On the scale of varieties between the two, the regional standard is posited as the range closest to the national standard. Local varieties that are not strictly classical dialects may no longer be active in some areas — a phenomenon illustrated by the variation in an area relatively close to Copenhagen, the center of gravity of new national standard developments. The development of features is described in terms of Bailey's wave model (1973).

As part of a national effort to improve reading levels, spontaneous speech samples were collected from 630 Latino, African American, and white children in grades 2 through 4 in Georgia, California, and Pennsylvania. In this study, data was used from 126 Latinos, and a comparison group of 28 African American and 28 white children to study their use of 3rd person possessive pronouns, periphrastic of possessives, and attributive -s possessives. It was found that Latino children confused his for her and her for his; used more periphrastic of constructions; and omitted the attributive -s marker in noun + -s + noun constructions. Multivariate analyses revealed that beyond Spanish influence, speaker sex, language origin, and grade also affected the expression of possession. Most striking are the differences according to speaker sex, and between Mexican and Puerto Rico origin children, which are considered in light of the closer relationship between Puerto Ricans and African Americans in Philadelphia.The research on which this report is based was carried out at the Linguistics Laboratory of the University of Pennsylvania, supported by NSF and the Interagency Educational Research Initiative as proposal 0115676 and the Spencer Foundation under Grant 200200074.

In this article we use the syntax of the noun phrase to evaluate two competing hypotheses: the traditional account, that Middle English is a West Germanic language with Old English as its immediate ancestor, and Emonds and Faarlund's (2014) proposal, that Middle English is a North Germanic language, the direct descendant of Old Norse. The development of nominal syntax shows that the Middle English noun phrase can be derived only from Old English, not from Old Norse. We examine six nominal characteristics; in each case, we find in Middle English exactly the construction that one would expect given the nominal syntax of previous Old English stages. The evidence from Old Norse shows that, although some of the same constructions did develop in the same way in the attested Norse varieties, the development occurred only at a later stage, too late to have affected the syntax of Middle English.

We analyze the dynamics of dialect loss in a cluster of villages in rural northern Russia based on a corpus of transcribed interviews, the Ustja River Basin Corpus. Eleven phonological and morphological variables are analyzed across 33 speakers born between 1922 and 1996 in a series of logistic regression models. We propose three characteristics for a comparison of the rate of loss of different variables: initial level, steepness, and turning point. We show that the dynamics of loss differs significantly across variables and discuss possible reasons for such differences, including perceptual salience, initial variation in the dialect, and convergence with regionally or socially defined varieties of Russian. In conclusion, we discuss the pros and cons of logistic regression as an approach to quantitative modeling of dialect loss. Our paper contributes to the study and documentation of Russian dialects, most of which are on the verge of extinction.

Based on evidence in Zen dialogues, Bianwen and Chuanqi texts in the Tang Dynasty (a.d. 618–907), and other sources from earlier and later periods, this article, after dismissing the common misconception that during Tang ba () was frequently used in the disposal and serial-verb constructions, proposes that the grammaticalization of the jiang () serial-verb construction was encouraged by yi (), but jiang's disposal characteristic was modeled after its passive counterpart bei, and that it is via the process of lexical replacement that ba replaced jiang's prepositional functions in instrumental and disposal constructions. Ba later lost its instrumental function to na (). We thus resolve the debate between Huang (1986) and Bennett (1981). Moreover, from the recurring pattern of change in the historical development of yi, jiang, ba, and na, we observe two counteracting principles at work: functional refinement, which forces a one-to-one correspondence between a form and its function, and analogous development, which does just the opposite to create a one-to-many relationship.

This paper discusses a group of Brazilian Portuguese plurals that end with the nasal diphthong [–ão] in their singular form. When pluralized, these words can take three plural forms: [–ões], [–ãos], or [–ães]. Each of these words has an etymological source that defines which morpheme should be applied to the plural form. However, in some cases, a different plural form is used instead of the etymological one. A set of psycholinguistic experiments was carried out with Brazilian Portuguese speakers. The goal was to analyze whether token or type frequency affects the choice of plural form for these words. The results show an interaction between type and token frequency, in the sense that type frequency serves as the basis for analogy, and token frequency can determine which words will be affected by the variation.

For fifty years after World War II, Latvia was incorporated into the former Soviet Union. Although in theory the use of regional languages was not discouraged, in practice knowledge of Russian was obligatory. Since 1991, Latvian has again become the official language, and knowledge of Russian is widespread but optional. These political events have created a natural experiment in the effects of almost universal bilingualism on a language. To assess the impact on pronunciation, native speakers of Latvian, ranging from retirement age to teens, were recorded reading a word list and a short narrative. Vowel pronunciation differed across the generations both in quantity relationships and in formant structure.Research for this project was supported by a grant from the International Research and Exchange Board, with funds provided by the U.S. Department of State (Title VII Program) and the National Endowment for the Humanities. None of these organizations is responsible for the views expressed here.

It is well documented that the study of differences in grammaticality contrasts across the world's languages has implications for the synchronic study of preferential/frequency contrasts within a single language. Our paper extends this observation, arguing that the cross-linguistic study of both grammaticality and frequency contrasts can be crucial to the proper characterization of patterns of diachronic change. As an illustration of this proposal, we investigate patterns of synchronic and diachronic variation in the use of postverbal negative quantifiers (e.g., nothing, nobody, no book, etc., as in, I know nothing) versus negative polarity items under negation (e.g., not … anything, not … anybody, not … any book, etc., as in, I don't know anything) in English. We show how a detailed comparison with similar patterns found elsewhere in closely related languages can give us a better understanding of which linguistic factors condition the use of these different kinds of indefinites in Modern Spoken English and a new perspective on a well-studied proposed change in progress in the English quantificational system.

Beliefs about a language rarely correspond to how it is used. This is especially true for Hebrew, a language that has been subject to continued ideological “preservation” efforts ever since its (re)vernacularization in the early 20th century. Recently, attention has turned to the maintenance of Hebrew gender morphology, which is perceived in both scholarly and popular opinion as threatened by a process of leveling to gender syncretized forms across a range of word classes and inflectional paradigms. In this article, I investigate the extent to which sociolinguistic evidence supports this perception in cases of animate reference. I argue that while the claim of widespread gender neutralization of these forms is descriptively valid, its characterization as a change-in-progress is inaccurate. Rather, I suggest that Hebrew is already fully syncretized for gender in certain relevant morphological contexts and that the perception of an ongoing process of change reflects a prescriptive belief about how Hebrew should be, not how it actually is.

In this paper, we return to the well-studied yet still puzzling phenomenon of complementizer omission in a large spoken corpus of Quebec French, with the help of modern computational methods for annotation and mixed effects logistic regression models. Supporting previous work, our study reveals that complementizer que omission is conditioned by social factors and grammatical factors; however, we also find that que omission is conditioned by cognitive factors such as information density. Our paper thus illustrates an important way in which older variationist corpora can continue to be valuable resources for studying fine-grained patterns of variation, particularly in their cognitive aspects.

One of the most salient putative African features of Palenquero, an Afro-Hispanic creole spoken in northern Colombia, is the prenominal plural marker ma. However, plural number is not categorically marked with ma, which alternates with bare forms in plural contexts and also occurs in singular contexts. In a principled sample of noun phrases (n = 1,186) from the spontaneous speech of twenty-seven Palenquero-Spanish bilinguals, the rate of ma (versus zero) is 51% in plural and 13% in singular contexts. Singular ma is favored with subjects and specific objects, consistent with an association with definiteness. In plural contexts, where it is robust, selection of ma is favored with specific and generic referents in subject role. This conditioning indicates that plural marking is favored for discourse referential nouns, in accordance with the cross-linguistic generalization that morphological marking tends to appear on instances that approach the prototypical function of a category (Hopper & Thompson, 1984).

Neither historical nor linguistic records reveal exactly when the shift from Negerhollands to English and English Creole began in the Danish West Indies. In order to assess phonological continuity and change in the last stage of this moribund creole, the following discussion (1) contrasts earlier and current views of Negerhollands and sketches language contact in the Danish West Indies; (2) examines the language history and the vowel systems of the last speaker; and (3) assesses variation in a Negerhollands corpus produced by 10 elderly bilinguals. The article demonstrates how using a variety of approaches can enhance the investigation of an underdocumented language.

Spanish grammarians used to say that the demonstrative adverbs aquí ‘here’ and acá ‘here’ can alternate in some contexts. They attempt to explain the alternation by appealing to one of two hypotheses, which can be identified as the movement-to-a-goal hypothesis and the delimitation hypothesis. The former states that acá is associated with movement directed to a goal, whereas aquí is associated with all other possibilities. The latter states that the place denoted by aquí is less limited and defined than the place denoted by acá The present article evaluates both hypotheses in a corpus of Venezuelan Spanish. The evaluation takes into account the kind of place denoted by aquí and acá (goal of movement, source of movement, place of transit, and location) and also the prepositions and verbs used when the adverbs denote the goal of movement. The results of the quantitative analysis prove the validity of the delimitation hypothesis.