{
  "id": "code-switching",
  "title": "Code-switching",
  "lang": "en",
  "source_lang": "en",
  "summary": "Code-switching is the use of two or more languages or language varieties within an interaction, text, or stretch of discourse.",
  "category": "language",
  "subcategory": "",
  "type": "concept",
  "aliases": [
    "Code switching",
    "Codeswitching",
    "Language switching"
  ],
  "generated_at": "2026-11-03",
  "model": "pro/gpt-6.1-sol",
  "url": "https://aiwiki.page/en/code-switching/",
  "translations": {
    "zh-Hans": {
      "url": "https://aiwiki.page/zh/code-switching/",
      "json": "https://aiwiki.page/zh/code-switching.json"
    }
  },
  "sections": [
    {
      "heading": "",
      "text": "Code-switching is the use of two or more languages or language varieties within a conversation, text, or other communicative event. Switching may occur between speakers’ turns, between sentences, or within a single sentence. The term is especially associated with bilingual and multilingual communication, although some definitions also include alternation between dialects or styles. In sociolinguistics, code-switching is studied as a structured communicative practice shaped by grammatical possibilities, social relationships, and the organization of interaction, rather than simply as evidence of incomplete language knowledge."
    },
    {
      "heading": "Scope and terminology",
      "text": "In this context, a code is a language or recognizable variety available to participants. It need not be a standardized or officially recognized language. Speakers may switch between a national language and a regional dialect, between two unrelated languages, or between closely related varieties.\n\nDefinitions differ across research traditions. Some scholars use code-switching broadly for multilingual alternation of many kinds. Others distinguish it from code-mixing, often reserving the latter for combinations within a sentence or for particular patterns of grammatical integration. These terms do not have universally accepted boundaries.\n\nThe relevant unit is not necessarily a complete sentence. A switch may involve a phrase, a quotation, a discourse marker, or an entire stretch of conversation. It can also occur in writing, including letters, fiction, advertising, and online messages. Because identifying a switch requires identifying the varieties involved, classification can be difficult where varieties overlap or participants do not treat them as sharply separate.\n\nCode-switching forms part of a speaker’s linguistic repertoire. Its frequency and meanings vary across communities: what appears conspicuous to an outsider may be an ordinary, relatively unmarked way of speaking among participants."
    },
    {
      "heading": "Main structural patterns",
      "text": "A widely used classification distinguishes switching according to its position in discourse.\n\nIntersentential switching\n\nIntersentential switching occurs between sentences or clauses. An illustrative English–Spanish example is:\n\n“The meeting starts at nine. No llegues tarde.”\n\nThe Spanish sentence means “Don’t arrive late.” Each sentence can largely be analyzed through the grammar of its own language.\n\nSwitching may also occur at a change of speaker, although researchers differ over whether every exchange in which speakers use different languages should count as code-switching or whether the term should require alternation within an individual’s contribution.\n\nIntrasentential switching\n\nIntrasentential switching occurs within a sentence or clause. For example:\n\n“We can meet después de clase.”\n\nHere the Spanish phrase means “after class” and occupies a position that an English prepositional phrase could occupy.\n\nSuch combinations are not generally random collections of words. Their organization can depend on syntax, morphology, lexical properties, and established community practices. Different language pairs present different possibilities because their grammatical structures do not align in identical ways.\n\nTag and discourse-marker switching\n\nShort expressions such as greetings, exclamations, tags, and discourse markers can be inserted into otherwise relatively monolingual speech. Because many such expressions are loosely attached to the surrounding clause, they may be less constrained by its syntax than embedded phrases are.\n\nNevertheless, their status may be ambiguous. A frequently used expression can become a conventional part of local speech and need not be experienced by speakers as a momentary change of language.\n\nSigned languages and code-blending\n\nSwitching also occurs between signed languages. In communication involving a signed and a spoken language, some bilinguals can produce speech and signs simultaneously. This simultaneous production is commonly called code-blending, distinguishing it from sequential switching. The availability of two articulatory channels makes these patterns different from ordinary alternation between two spoken languages."
    },
    {
      "heading": "Communicative and social functions",
      "text": "A switch has no fixed meaning independent of context. Its interpretation depends on who is speaking, to whom, about what, and within which activity.\n\nCommon functions include:\n\n- Accommodating an interlocutor: selecting a language understood or preferred by a participant.\n- Managing participation: addressing a particular person, incorporating a new arrival, or changing the intended audience.\n- Quoting or voicing another person: reproducing speech in its original language or distinguishing another speaker’s perspective.\n- Organizing discourse: marking a topic change, a parenthetical comment, a reformulation, or a contrast.\n- Expressing social alignment: invoking shared experience, familiarity, institutional authority, or distance.\n- Accessing particular expressions: using a term associated with a profession, institution, cultural practice, or personally familiar experience.\n\nThese functions belong partly to pragmatics, because they concern what language choices accomplish in context. Switching may help frame an utterance as a joke, an instruction, a personal aside, or reported speech without explicitly labeling it as such.\n\nAn influential distinction separates situational switching, associated with changes in participants, setting, or activity, from metaphorical switching, in which a language change evokes a different social relationship or interpretive frame within an otherwise continuing situation. In practice, the categories can overlap.\n\nNo language inherently expresses intimacy, authority, rationality, or emotion. Such associations arise from social histories and local conventions. Nor must every switch carry a distinct social message: in some communities, sustained multilingual alternation is itself the expected conversational pattern."
    },
    {
      "heading": "Grammatical explanations",
      "text": "Research asks both where switches occur and why some combinations are more acceptable or frequent than others.\n\nOne influential approach, associated with Shana Poplack, proposed constraints on switching. The equivalence constraint predicts that switches are favored where the surface structures of the languages align, so that switching does not violate the ordering requirements of either language. The free morpheme constraint concerns restrictions on switching between bound and free morphemes, with qualifications involving integration into the receiving language.\n\nThese proposals stimulated extensive comparative research. They are not accepted as exceptionless laws: findings depend on the language pair, the construction, and how switching and borrowing are distinguished.\n\nCarol Myers-Scotton’s Matrix Language Frame model instead emphasizes asymmetry within mixed clauses. A matrix language supplies the principal morphosyntactic frame, while an embedded language contributes material within it. The model makes predictions about word order and particular grammatical morphemes. Identifying a single matrix language is more straightforward in some mixed structures than in others.\n\nOther approaches draw on generative grammar, lexical properties, or usage-based patterns. A central question is whether multilingual combinations require special grammatical mechanisms or can be explained through the same principles that organize monolingual language.\n\nThese approaches describe different aspects of the evidence. Structural constraints, processing demands, conventional expressions, and community norms can all contribute to the distribution of switches."
    },
    {
      "heading": "Distinction from related phenomena",
      "text": "Lexical borrowing\n\nLexical borrowing occurs when a word or expression from one language becomes part of another language’s vocabulary. A conventional loanword can be used by speakers who do not know the source language.\n\nCode-switching, by contrast, usually involves the active use of more than one linguistic system within an interaction. The boundary is nevertheless difficult to draw for isolated words. Pronunciation, grammatical integration, frequency, community-wide use, and speakers’ judgments can provide evidence, but no single criterion resolves every case. A one-time insertion may display receiving-language morphology, while an established loanword may preserve features associated with its source.\n\nDiglossia and language shift\n\nDiglossia concerns a relatively stable distribution of varieties across social functions, such as formal institutions and everyday conversation. It describes the organization of a linguistic community, rather than the moment-to-moment alternation itself. Diglossic arrangements can encourage particular switching patterns, but code-switching also occurs without diglossia.\n\nLanguage shift is a longer-term change in which a community increasingly uses one language in place of another. Frequent switching alone does not establish that a language is being abandoned; stable bilingual communities may maintain switching practices over generations.\n\nStyle-shifting and translanguaging\n\nStyle-shifting commonly refers to changes in formality, pronunciation, or register within a language. Some broad uses of code-switching include such changes; narrower linguistic definitions distinguish them from alternation between languages.\n\nTranslanguaging emphasizes how multilingual people draw on an integrated repertoire rather than treating their practices only as movement between separately named languages. It overlaps empirically with code-switching but may adopt a different theoretical perspective. The relationship between the terms remains debated, particularly over how linguistic categories should relate to speakers’ experience and to socially established language boundaries."
    },
    {
      "heading": "Historical development",
      "text": "Descriptions of multilingual alternation predate its establishment as a major research topic. During the mid-twentieth century, work on language contact and bilingualism helped make it an object of systematic analysis.\n\nJohn J. Gumperz and collaborators were influential in showing how language alternation helps participants interpret social relationships and conversational meaning. Later research developed increasingly detailed accounts of grammatical structure, particularly through studies of naturally occurring bilingual speech.\n\nThe field expanded to include experimental research, child language, classroom interaction, migration, digital communication, and signed languages. This development challenged accounts that treated mixed speech mainly as corruption, carelessness, or linguistic deficiency."
    },
    {
      "heading": "Acquisition and processing",
      "text": "Children exposed to more than one language may switch early in development. Their language choices can be influenced by vocabulary availability, the person being addressed, exposure patterns, and the practices of caregivers. Mixing by itself is not evidence that a child cannot distinguish the languages. Research on language acquisition distinguishes knowledge of linguistic systems from developing control over when and with whom to use them.\n\nIn adult processing research, switching is often investigated through tasks in which participants name objects in different languages. A change of language can produce a measurable delay relative to repeating the same language, commonly called a switch cost. Its size varies with task design, proficiency, preparation, and the direction of switching.\n\nLaboratory tasks do not reproduce all conditions of everyday conversation. Cued alternation between isolated words differs from voluntary switching in a familiar interaction. Experimental costs therefore cannot be treated as direct measures of how difficult or disruptive natural code-switching must be."
    },
    {
      "heading": "Education, institutions, and public attitudes",
      "text": "In education, switching can be used to explain terminology, compare expressions, manage classroom interaction, or connect new material with familiar knowledge. Its role depends on instructional goals, learners’ proficiency, and the surrounding language policy. Switching alone does not demonstrate either effective teaching or inadequate command of the instructional language.\n\nPublic attitudes are shaped by language ideologies, including beliefs that legitimate language use should remain monolingual or conform to a standard variety. The same practice may be praised as cosmopolitan in one setting and stigmatized in another. Such judgments reflect social hierarchies as well as linguistic behavior.\n\nIn broader public discussion, especially concerning workplaces and racialized identities, code-switching can refer to changes in accent, vocabulary, manner, or self-presentation across social settings. This usage extends beyond the narrower linguistic sense of alternating languages. Distinguishing the senses is important when comparing research claims or interpreting public debates."
    },
    {
      "heading": "Research methods and limitations",
      "text": "Researchers use recorded conversations, interviews, ethnographic observation, written materials, acceptability judgments, and controlled experiments. Corpus linguistics supports analysis of switch locations, grammatical patterns, and differences across speakers or settings.\n\nA major methodological problem is deciding what counts as a switch. Shared vocabulary, established loanwords, closely related varieties, and uncertain word histories can complicate annotation. Written language labels may also obscure pronunciation and other cues available in speech.\n\nReliable interpretation therefore requires more than counting words assigned to different languages. Sequential context, participants’ responses, community conventions, and the grammatical properties of the varieties involved can all affect whether an alternation is analytically significant."
    }
  ],
  "links": [
    "language",
    "bilingualism",
    "multilingualism",
    "sociolinguistics",
    "dialect",
    "syntax",
    "morphology",
    "sign-language",
    "pragmatics",
    "word-order",
    "generative-grammar",
    "register-linguistics",
    "language-acquisition",
    "education",
    "language-policy",
    "standard-language",
    "corpus-linguistics"
  ],
  "sources": [],
  "license": "https://creativecommons.org/licenses/by/4.0/"
}