YPL2 - Issue 14

Issue 14 (June 2015)

Editors: Rebecca Woods, Ella Jeffries, Mariam Dar and Aiqing Wang

Developing an Urdu CDI for early language acquisition

1–14

Mariam Dar, Huma Anwaar, Marilyn Vihman and Tamar Keren-Portnoy

Abstract

This paper reports on the development of a lexical checklist for parents to trace vocabulary advances in Pakistani children. Various cross-linguistic adaptations of the MacArthur-Bates Communicative Development Inventory (CDI) were consulted. Pilot testing was carried out using a sample of 17 children from middle-class homes, aged from 12 to 30 months. Cross-linguistic comparison reveals similarities between the vocabulary growth of the Pakistani children and children learning other languages. Plans for further pilot-testing and eventual validation are discussed. 

Pre-school children's identification of familiar speakers and the role of accent features

15–40

Ella Jeffries

Abstract

This paper explores pre-school children's abilities in recognising and identifying familiar speakers. Two experiments were run with a group of 22 nursery children from York, in the north of England, aged 2.4-4.9 years. The first experiment investigates the children's ability to identify six known nursery school teachers from short audio stimuli. The results of this experiment showed an improvement in the identification ability of the children correlating with age, greater exposure to the teachers, and a longer stimulus length. The pitch and voice quality of the speakers' voices were also found to play a role in how easily identified the individuals were. The second experiment considers the role of regional accent features in the recognition of one particular speaker from Yorkshire when compared to other, unfamiliar speakers with different regional accents of British English. This experiment presents more mixed results and different results between the sexes. Among the boys, the results showed an age-related improvement. Girls with parents from outside of Yorkshire performed better than those with parents from Yorkshire. Overall, the children were more likely to misidentify an unfamiliar speaker as the familiar speaker if the speakers had a similar regional accent. Furthermore, a disguise in accent by the familiar speaker resulted, as expected, in more children being unable to identify her. The results are discussed in terms of memory models of speaker recognition. 

The role of gender-correlated sociolinguistic variables in identifying speaker-indexical information

41–80

Ania Kubisz

Abstract

The present paper investigates perceptions of speaker social-indexical information, including gender, age and social-class, from smaller phonetic segments such as gender-correlated phonetic variants. Since fundamental frequency (F0) is not the only cue to speaker gender identification, the perceptions are examined using gender-ambiguous sounding speech. The results of the study show that while speaker social-indexical information is identifiable at the segmental level, some of the phonetic variants seem to encode social-indexical information to a more consistent degree than others. 

Implicit associations with Welsh in two educational contexts

81–105

Rachelle Lee

Abstract

This study explores whether adolescents in Welsh- and English-medium schools in Cardiff, Wales, differ in the strength of their positive and negative associations with the Welsh language. An Implicit Association Test (IAT), which uses latency times to determine associations with dichotomous target concepts, is employed. The results suggest that there are indeed differences in both the strength and direction of positivity toward the Welsh language between pupils at different types of schools. Pupils from the Welsh-medium school displayed significantly more positive associations with the Welsh language than by pupils from the English-medium school, even when controlling for first language and parental language effects. While self-rated proficiency in Welsh is positively correlated with implicit associations with the language at the Welsh-medium school, no such correlation was found among pupils at the English-medium school. Finally, pupils from the Welsh-medium school who opted to complete the IAT in Welsh had significantly more positive associations with Welsh than those who completed the task in English, which is evidence of an implicit association-behaviour link. The implicit association data suggests that there may be a substantial ideological divide between Welsh- and English-medium schools in Cardiff. 

Accent categorisation by lay listeners: which type of "native ear" works better?

106–131

Chen Shen and Dominic Watt

Abstract

Listeners, be they lay or expert, can to a greater or lesser extent distinguish and correctly identify different accents of familiar languages. This ability plays a central role across a spectrum of speech perception-based activities, not the least of which are speaker profiling and comparison of the sort carried out for forensic purposes. Factors that affect listeners' abilities to identify different accents, such as their native linguistic competence and their level of familiarity with the target variety, have been reported extensively in the literature. However, we currently lack data from studies that make direct comparisons of listeners' abilities to correctly categorise foreign-accented varieties of a language they speak natively (L1 listeners) against the competence of non-native speakers of that language (L2 listeners), where the aim of the listening task is to categorise familiar and unfamiliar foreign accents of the language. The present study is designed to investigate these issues by looking at how well a listener's "native ear" allows him or her to judge the origins of non-native speakers of English. By "native ear", we mean in the present case the sort of perceptual acuity that we might associate with being an L1 English speaker versus an L1 Chinese speaker who speaks English as an L2, in the situation where speakers of these two sorts are exposed to recordings of Chinese-accented English (CE). For the experiment reported in this paper, the speakers recruited were L2 English speakers from East Asia (Japan, South Korea, and three regions of China), with the principal interest being in those in the Chinese group. A total of 42 listeners (monolingual L1 English speakers vs. Chinese L2 English speakers) completed a dual-task forced-choice experiment in which they were exposed to samples of English spoken by the East Asian talkers. The results showed that the L1 Chinese listeners were significantly better than the L1 English listeners at distinguishing the CE speakers from the Japanese and South Korean ones. However, neither of the two types of "native ear" was better than the other (or indeed in absolute terms) at correctly categorising the three subvarieties of CE according to three broadly- defined geographical/dialect regions (the North, South, and West of China). We conclude by discussing the relevance of our results to forensic cases in which an analogous kind of categorisation task might be necessitated.