In the study of past be, most researchers have assumed a two variant distinction for negative forms (i.e., wasn't and weren't). In Warren County, North Carolina, evidence suggests a three variant distinction (i.e., wasn't, weren't, and wont). Throughout the history of sociolinguistic investigation, two types of variants have been noted: a sociolinguistic variant and a linguistic variant. In Warren County, wont functions as both types, as demonstrated through acoustic, morphophonological, and sociolinguistic evidence. Through the analysis of negative past be, guidelines are suggested for determining a particular form as a separate linguistic and sociolinguistic variant.

The English of bilingual Cajuns living in southern Louisiana has been pejoratively depicted as an accented English; foremost among the stereotypes of Cajun English is the use of tink and dat for think and that. We present a variationist study of /θ/ and /ð/ in the speech of bilingual Cajuns in St. Landry Parish. The results show a complex interrelationship of age, gender, and social network. One of the major findings is a v-shaped age pattern rather than the regular generational model that is expected. The older generation use more of the dental variants [t, d] than all others, the middle-aged dramatically decrease their use, but the young show a level of usage closer to the old generation. The change is attributed to both language attrition and the blossoming of the Cajun cultural renaissance. Interestingly, neither young men in open networks nor women of all ages in open networks follow the v-shaped age pattern. In addition, they show opposite directions of change: men in open networks lead the change to [d], whereas women in open networks drop the dental variants of [t, d] almost entirely.

This sociolinguistic investigation analyzes an innovative usage of Spanish motion verbs, demonstratives, and locatives in Barcelona that involves crosslinguistic pragmatic transfer. Speakers in the two social networks examined (N = 58) use these Spanish deictics following pragmatic rules that generally correspond to the rules for their Catalan counterparts. Quantitative analysis demonstrates that this innovative usage of the Spanish deictics is not predictable from the lexical form of the deictic systems in both languages. Multiple regression analysis (SPSS) demonstrates that as relative exposure to Catalan increases so does the amount of innovative usage observed, although degree of integration into a Catalan social network and degree of Catalanist ideology do not affect such usage. Qualitatively, this innovative usage is a linguistic marker of a unique contact variety of Spanish spoken in Catalonia (i.e., Catalan Spanish). In this variety, such usage represents a potential resource for performing Catalan identity.

It has been claimed that the ellipsis of postpositional particles is one of the properties of Japanese women's language. The validity of this claim, however, is compromised by several methodological problems in prior research in this area: (1) the subjective comparison of bare percentages without any statistical verification of male–female differences; (2) the neglect of the intersecting relationships among a variety of potential constraints which simultaneously influence the speaker's choice of the variable; and (3) the exclusive focus on single-sex interactions. Conducting multivariate analyses of conversational data from both single-sex and cross-sex peer interactions, this preliminary study demonstrates the overgeneralization of gender-linked differentiation in past work due to the neglect of the relative strength of potential intersecting factors. The results also reveal statistically significant degrees of stylistic intragender variability across different types of gender composition, which counters the static approach to a social variable of gender in the traditional sociolinguistic paradigm.