Multilingual Code-Switching in Public Services: A Field Study

Thesis Abstract

In multilingual urban centers, public service interactions routinely involve speakers of diverse languages, yet many encounters rely on monolingual communication that can marginalize customers and reduce service quality; this study investigates how multilingual code-switching facilitates or hinders effective public service delivery across health, social welfare, and municipal administrative settings. The aim is to examine the patterns, functions, and outcomes of code-switching in public service interactions and to assess their implications for client comprehension, satisfaction, and access to services. Specific objectives are (1) to characterize linguistic configurations (languages, registers, and switches) used by frontline staff and clients; (2) to identify communicative functions of code-switching (e.g., topic shifting, stance-taking, user engagement, relief of linguistic constraint) within service encounters; (3) to evaluate the relationship between code-switching practices and service outcomes such as task completion time, error rate, and client satisfaction; (4) to examine organizational and policy contexts that shape code-switching behaviors; and (5) to develop an evidence-based framework for optimizing multilingual service delivery. The study adopts an ethnomethodological field design, integrating conversation analysis (CA), discourse analysis, and mixed-methods triangulation. The population comprises frontline public service staff (n=60) and clients (n=600) in three municipal service domains healthcare clinics, social welfare offices, and urban administrative centers in a multicultural city. A stratified purposive sampling approach ensures representation across language repertoires, service types, and encounter outcomes. Data collection employs (a) video-recorded service encounters (approximately 300 hours), (b) audio-recorded semi-structured interviews with staff (n=40) and clients (n=40), (c) organizational documents and policy memos, and (d) a post-encounter satisfaction survey (response rate ? 70%). Instruments include a coding scheme for language alternation (sequence of switches, language pairing, and prosodic cues), CA-based transcripts with multi-tier labeling, and a standardized satisfaction index. Validity and reliability are addressed through intercoder reliability checks (Cohen’s kappa ? 0.80 for coding categories), pilot testing of instruments, and triangulation across data sources. Analytical procedures comprise (i) quantitative analysis using multilevel regression to assess the association between code-switching frequency and service outcomes, controlling for domain, encounter duration, and client language proficiency; (ii) ANOVA to compare outcomes across service domains and language groups; (iii) thematic analysis of interview data to identify perceived functions and organizational constraints; (iv) CA to examine interactional sequences and turn-taking in switch moments; and (v) a realist synthesis to integrate findings with organizational policy and training implications. A conceptual model will be iteratively refined to map how sociolinguistic factors, institutional norms, and domain-specific demands interact to produce observable service outcomes. Expected findings include (a) distinct code-switching patterns associated with success in each domain, with higher client satisfaction when staff employ client-language switches to facilitate comprehension and legitimacy; (b) functional categories of switches—phatic support, technical clarification, affective alignment, and resource coordination—each differentially predicting task efficiency and accuracy; (c) organizational constraints such as time pressure and policy restrictions shaping the likelihood and timing of switches; and (d) a positive correlation between staff language proficiency and favorable outcomes, moderated by training exposure to multilingual communication strategies. The study anticipates variation in switch functions across language pairs and service domains, with CA revealing micro-interactional structures that either promote inclusive access or inadvertently reproduce language-based barriers. The contribution to knowledge lies in presenting a unified empirical account of multilingual code-switching in public service contexts, linking interactional linguistics with public administration and service design. The research advances theory by integrating CA, discourse-pragmatic perspectives, and sociolinguistic repertoires into a practical framework for multilingual service delivery, and it informs policy by identifying training needs, staffing considerations, and protocol adjustments that support equitable access. The main conclusion is that deliberate, context-aware code-switching enhances communicative efficiency and client trust when aligned with staff proficiency, domain demands, and supportive organizational policies. Recommendations include incorporating language-macial training (language-switch competencies, cultural brokerage, and ethical guidelines) into staff development programs, implementing flexible time allowances for multilingual encounters, and developing standardized yet adaptable code-switching protocols to reduce miscommunication and improve public service equity.