Recent research has been concerned with whether speech accommodation is an automatic process or determined by social factors (e.g. Trudgill 2008). This paper investigates phonetic accommodation in New Zealand English when speakers of NZE are responding to an Australian talker in a speech production task. NZ participants were randomly assigned to either a Positive or Negative group, where they were either flattered or insulted by the Australian. Overall, the NZE speakers accommodated to the speech of the AuE speaker. The flattery/insult manipulation did not influence degree of accommodation, but accommodation was predicted by participants' scores on an Implicit Association Task that measured Australia and New Zealand biases. Participants who scored with a pro-Australia bias were more likely to accommodate to the speech of the AuE speaker. Social biases about how a participant feels about a speaker predicted the extent of accommodation. These biases are, crucially, simultaneously automatic and social. (Speech accommodation, phonetic convergence, New Zealand English, dialect contact)*

This article investigates the enregisterment of an internet-specific language variety and its features. The enregisterment of internet language is explored through several sites of metadiscourse: academic scholarship about computer-mediated communication, uses of the metalinguistic terms netspeak and chatspeak in print media, and online comment threads about language and the internet. This metadiscourse provides evidence of a shared concept of internet language as comprising distinctive written features, primarily acronyms, abbreviations, and respellings. Internet language's enregisterment emerges from standard language ideology and deterministic views of technology, where the construal of these features as both nonstandard and internet-specific articulates the perceived distinctiveness of internet interactions. Yet empirical evidence shows that these features are relatively rare in instant messaging conversations, one form of interaction to which internet language is attributed; this discrepancy has implications for the application of indexical order to enregisterment. (Enregisterment, language ideology, computer-mediated communication, internet, metadiscourse, indexical order, Standard English, technological determinism, mass media)*

In the Tanzanian Bongo Flava youth music scene, Abel Motika is a popular artist who uses both verbal and visual markers of Kisongo Maasai ethnicity to style himself as “the Maasai rapper” with the stage name “Mr. Ebbo.” Through analysis of his 2002 song “Mi Mmasai” ‘I am Maasai’, this study investigates his ethnic stylizing in playful use of Maa pronunciation and an understudied Swahili language game known as kinyume ‘backwards style’. The study finds that while Ebbo strategically disrupts the sociolinguistic order that privileges Standard Swahili, the Maasai persona he projects is humorously stylized as unable both to speak Standard Swahili and to engage with the urban lifestyle associated with Tanzania's de-ethnicized Swahili modernity, thereby leaving dominant ideologies of language and ethnicity intact. Moreover, in arguing that Motika's stylization of ethnicity has a contradictory effect, both affirming a local ethnic identity and preserving the logic of ethnolinguistic stereotyping, the study critiques approaches to hip hop that privilige authorial intent and assume linguistic subversiveness. (Swahili, Maa, Bongo Flava, parody, ethnicity, rap, kinyume)*

Although an increasing number of sociolinguistic researchers consider functions of voice qualities as stylistic features, few studies consider cases where voice qualities serve as the primary signs of speech registers. This article addresses this gap through the presentation of a case study of Lachixío Zapotec speech registers indexed though falsetto, breathy, creaky, modal, and whispered voice qualities. I describe the system of contrastive speech registers in Lachixío Zapotec and then track a speaker on a single evening where she switches between three of these registers. Analyzing line-by-line conversational structure I show both obligatory and creative shifts between registers that co-occur with shifts in the participant structures of the situated social interactions. I then examine similar uses of voice qualities in other Zapotec languages and in the two unrelated language families Nahuatl and Mayan to suggest the possibility that such voice registers are a feature of the Mesoamerican culture area. (Voice quality, register, performance, metapragmatics, Mesoamerica, Zapotecan, Mayan, Nahuatl)*