Code-Switching Patterns in Multilingual Urban Speech: A Corpus-Based Analysis of Social Context, Identity, and Language Attitudes

Project Abstract

This study investigates code-switching (CS) patterns in multilingual urban speech through a corpus-based approach to elucidate how social context, speaker identity, and language attitudes shape language choice in everyday communication. Drawing on a newly compiled, large-scale spoken corpus collected from diverse urban environments, including public transit, markets, educational settings, and online interactions, the research analyzes CS phenomena across multiple language pairs commonly found in urban multilingual ecosystems. Employing a mixed-methods design, the study combines quantitative coding of CS instances, syntactic and pragmatic analyses, and qualitative interviews to triangulate findings and capture subtleties in speaker intent and social meaning. The corpus is annotated for language segments, speaker metadata (age, gender, socioeconomic status, ethnicity), discourse function (topic shift, emphasis, irony, solidarity), and social context (formal/informal, institutional vs. informal domains). Advanced computational tools, including automatic language identification, n-gram frequency analysis, and topic modeling, are integrated with manual coding to ensure reliability and depth. The central aim is to model the relationship between CS frequency and social/contextual variables, identify prominent CS strategies (intersentential, intrasentential, and tag-switching), and examine how identity construction and stance-taking mediate language choices in real time. A key component of the methodology is the examination of attitudinal factors through perception surveys and ethnographic notes that assess attitudes toward each language, perceived prestige, and perceived interlocutor proximity, thereby linking CS behavior to broader sociolinguistic ideologies. The study also evaluates how CS functions as a socio-pragmatic resource for managing social distance, power relations, group membership, and credibility in urban interactions. Findings are expected to reveal context-dependent CS patterns, with higher rates of CS occurring in informal, multilingual domains and in interactions involving cross-group interlocutors, while more stable monolingual segments are associated with formal or institutional settings. The research contributes to theory by refining models of CS as a dynamic resource for social meaning, extending the understanding of how urban multilingualism interacts with mobility, digital communication, and age-related language repertoires. Practically, the results have implications for language education, public policy on multilingual communication, and the design of natural language processing systems that must accurately detect and interpret CS in diverse urban speech. The study also discusses ethical considerations in data collection, the representativeness of the corpus, and limitations related to annotation reliability and speaker self-report bias. Overall, the project advances knowledge on how code-switching functions as a nuanced, context-sensitive mechanism for indexing identity, solidarity, and social stance in contemporary urban multilingual settings.

Project Overview

What This Project Is About

---

This project examines how people switch between languages in busy city settings, and why they choose to switch in different social situations. It uses real language samples from conversations to see patterns of language choice, and what those choices say about identity and attitudes toward languages.

---

---

The Problem It Addresses

---

Many urban multilingual communities use more than one language in daily talk, but we don’t always understand when and why people switch. This study fills gaps by linking language choices to social context and personal attitudes, helping educators and policymakers appreciate the value of multilingual speech.

---

---

Objectives of the Project

---

1. Identify common language switches in everyday conversations.
2. Link switching patterns to social contexts (e.g., age, setting, and topic).
3. Explore how speakers describe their own language preferences and identities.
4. Provide practical insights for language education and community programs.

---

---

What You Will Do Step by Step

---

1) Collect short recording samples from consenting participants in public and private settings. 2) Transcribe the samples into text, noting language switches. 3) Classify contexts and speaker backgrounds. 4) Analyze how often and where switches occur. 5) Interpret why speakers switch, informed by simple theories of language and identity. 6) Present findings with clear examples and practical implications.

---

---

Expected Outcome

Clear patterns of when and why people switch languages, a better understanding of how multilingualism shapes identity, and practical guidance for teaching and community programming that respects language diversity.