LANGUAGE USE AND CODE-SHIFTING AMONG PIDGIN (HAWAIʻI CREOLE) SPEAKING CHILDREN

Abstract

Children who speak non-standardized language varieties are at risk for both over- and under-referral to speech-language and special education services (e.g., Morgan et al., 2016; Pearson et al., 2014). Extensive research with bidialectal (Craig, 2016; Kohn et al., 2021) and bilingual (Goldstein, 2022; Paradis et al., 2021) children has shown the importance of distinguishing between language variation and language deficit when assessing children from culturally and linguistically diverse backgrounds. Unlike the extensive research available on the language development of African American Language bidialectal or Spanish/English bilingual children, research on children who speak Pidgin (Hawaiʻi Creole) is sparse and none of it is recent (Day, 1972; Purcell, 1984). This dissertation begins the process of describing the morphosyntactic features of Pidgin used currently by children in Hawaiʻi as well as how much or little children change their use of those features across different tasks and interlocutors. To accomplish this, two studies were conducted using two different data sets: Study 1 utilizes a previously collected data set of story retells collected from around Hawaiʻi in 2014 (Fiestas, 2015) and Study 2 utilizes a newly collected set of story retells and play-based conversations conducted with different interlocutors. Study 1: Because the Fiestas corpus represents samples from heterogeneous populations, a subset of the available recordings was selected to be rated by adult Pidgin speakers on a Likert-type scale from 1 (heavy Pidgin) to 4 (English) using procedures similar to Oetting and McDonald (2002). The four children whose audio clips had mean ratings closest to Pidgin (Pidgin-rated; PR) and English (English-rated; ER) were then selected from grades K through 3rd. The story retell recordings for the selected children (n = 32) were then fully transcribed and analyzed for Pidgin morphosyntax use. Results suggest that Pidgin-speaking children today may not produce “classical” Pidgin features like copula/progressive/completive ste, future/irrealis go, and clause final forms when speaking with unfamiliar adults. They also might use other classical Pidgin forms such as past tense wen and indefinite article wan with a low frequency. Results also suggest that plain form, zero-marked (zero BE), and invariant (invariant waz, invariant indefinite article a) forms are characteristics of typical Pidgin, rather than developmental forms, since only PR children used these forms in 2nd and 3rd grades, although both ER and PR children used them in K and 1st grades. Mean Pidgin feature use (Pidgin Density Measure (PDM); mean Pidgin tokens per utterance) for the PR children were statistically higher than the ER children showing that patterns of language use by PR children are distinguishable from patterns used by ER children. A statistically significant, large relationship was found between PDM and the mean rating by adult listeners. This result suggests that Pidgin/English-speaking adult listener ratings may be a relatively reliable way to identify child Pidgin speakers. Study 2: Drawing from methods and tools used with bilingual and bidialectal children, this study elicited language samples from Hawaiʻi children (n = 14) across two tasks (story retell and play-based conversation) and two types of interlocutors (White English speaker and Local Pidgin speaker). Data were gathered in two sessions: English-context where the child listened to an English story and talked with an English-speaking investigator and Pidgin-context with Pidgin story and Pidgin speaker. Additionally, each child’s Pidgin exposure was estimated from a parent interview. Due to the Covid-19 pandemic, all interactions with subjects were remote (video conferencing, phone, email). Overall, children’s PDMs were very low during play-based conversation for both the English- and Pidgin-contexts. For the story retells, children showed a trend toward higher PDMs during the Pidgin-context than in the English-context. There also was an overall trend towards a positive correlation between mean PDM and current exposure to Pidgin as indexed by parent report estimates. These results provide at least tentative evidence that child Pidgin-speakers adapt their language use according to context. Parent interviews may also be helpful for identifying child Pidgin speakers with the caveat that not all parents will be fully forthcoming about this often-stigmatized variety. Taken together, findings from these small-scale studies provide a more expansive description of child Pidgin than has been available to date and shows at least tentative evidence that child Pidgin-speakers adapt their use of language to situational context. Better understanding of these children's use of Pidgin has clinical and educational implications for the unbiased assessment of language development (de Villiers, 2017) in this historically underserved population.