Introduction
The phenomenon of linguistic diversity has assumed unprecedented salience in contemporary globalised societies, wherein the co-existence of multiple languages within a single community or individual speaker is no longer regarded as an anomalous condition but rather as a normative feature of human communication. Across Europe and beyond, patterns of migration, transnational mobility, and the pervasive influence of digital media have rendered bilingualism and multilingualism defining characteristics of the modern sociolinguistic landscape. Poland, historically characterised by relative linguistic homogeneity, has in recent decades experienced a profound transformation in its relationship with English, a language that now occupies a prominent position within Polish educational institutions, professional environments, and popular culture alike. The result of this sustained contact between Polish and English is not merely a population of speakers capable of operating in either language independently, but rather a substantial and growing community of bilinguals for whom both languages constitute integrated components of a unified communicative repertoire.
Within the study of bilingual communication, few phenomena have attracted as sustained or as theoretically productive a degree of scholarly attention as code-switching — the practice whereby bilingual or multilingual speakers alternate between two or more languages within the course of a single conversation, utterance, or discourse context. Far from representing a failure of linguistic competence or an indicator of insufficient mastery in either language, code-switching has been recognised by researchers since at least the foundational work of John Gumperz in the 1970s as a sophisticated communicative strategy governed by systematic structural constraints and motivated by a complex array of social, pragmatic, and interactional functions. The decision to switch languages — whether at the level of the word, the phrase, the clause, or the extended discourse segment — is rarely arbitrary; it is instead embedded within a dense matrix of contextual variables that include the speaker's social identity, the nature of the interlocutor relationship, the conversational topic, and the broader community norms governing language use.
The present thesis is concerned specifically with the code-switching behaviour of Polish-English bilingual speakers as it manifests in informal conversational contexts. The choice to focus upon informal interaction, as opposed to institutional or formal settings, is deliberate and theoretically motivated: it is in informal conversation that speakers exercise the greatest degree of freedom from prescriptive linguistic norms, and it is therefore in such contexts that the full range of code-switching functions — from the negotiation of social identity to the management of interpersonal solidarity — may be most clearly observed. Polish-English bilingualism represents a particularly fertile context for such an investigation, not only because of the typological distance between the two languages and the distinct structural challenges this distance poses for language alternation, but also because of the complex sociohistorical and ideological conditions that shape attitudes toward English in Polish society, conditions that inevitably inflect the meanings and motivations associated with switching behaviour.
The research questions that animate the present investigation may be formulated as follows. First, what structural patterns of code-switching are most prevalent among Polish-English bilingual speakers in informal conversation, and how may these patterns be accounted for within established typological and formal frameworks? Second, what social and pragmatic functions does code-switching serve in such contexts, and in what ways do variables such as age, gender, and level of bilingual proficiency condition the frequency and character of switching behaviour? Third, how do the empirical findings generated by this study compare with, and in what respects do they extend or complicate, the theoretical models of code-switching proposed by researchers such as Shana Poplack, Carol Myers-Scotton, Pieter Muysken, John Gumperz, and Peter Auer? In addressing these questions, the thesis aims to contribute both to the descriptive documentation of Polish-English code-switching and to the broader theoretical conversation concerning the structural and functional dimensions of language alternation in bilingual communities.
The significance of the present study is justified on multiple grounds. From a theoretical perspective, the investigation offers an opportunity to examine the applicability of models developed primarily on the basis of other bilingual communities — most notably Spanish-English bilingualism in North American contexts — to a typologically distinct language pair characterised by significant differences in morphosyntactic structure, orthographic conventions, and sociolinguistic history. The structural complexity of Polish, with its rich inflectional morphology and relatively free word order, raises specific questions regarding the integration of English elements into Polish-matrix utterances that are not fully addressed by existing theoretical frameworks. From a sociolinguistic perspective, the study speaks to broader questions concerning the role of language choice in the construction and performance of identity, the negotiation of group membership and social solidarity, and the expression of speaker stance and affect — questions whose relevance extends well beyond the specific community under investigation. From a practical perspective, the findings have potential implications for language education, for the design of bilingual language policies, and for the understanding of language change processes in Polish society more generally.
The thesis is organised into three substantive chapters, each of which addresses a distinct but interconnected dimension of the phenomenon under investigation, preceded by the present introduction and followed by a conclusion that synthesises the principal findings and reflects upon their theoretical implications. Chapter One, entitled Theoretical Foundations of Code-Switching in Bilingual Communication, provides the conceptual and terminological groundwork upon which the subsequent analysis rests. It examines the principal definitions and typologies of code-switching that have been proposed in the scholarly literature, with particular attention to the contributions of Poplack, Myers-Scotton, and Muysken, and reviews the major sociolinguistic models — including Gumperz's notion of conversational code-switching, Myers-Scotton's Matrix Language Frame model, and Auer's interactional approach — that have been developed to account for both the structural patterning and the functional motivations of switching behaviour. The chapter also situates Polish-English bilingualism within its broader sociohistorical context, providing the background necessary for an informed interpretation of the empirical data.
Chapter Two, Sociolinguistic and Pragmatic Dimensions of Code-Switching, moves from the structural and definitional concerns of the first chapter to a more detailed examination of the social and pragmatic functions that code-switching serves in bilingual interaction. The chapter explores the role of language alternation in the construction and negotiation of speaker identity, in the expression of solidarity and social distance, and in the dynamics of power and authority within conversational exchanges. It further examines the pragmatic dimensions of code-switching, including its use as a device for emphasis and rhetorical effect, for the mitigation of face-threatening acts, and for the management of conversational structure and interactional alignment. The chapter also considers the attitudes that speakers and communities hold toward code-switching, and examines the relationship between overall proficiency in each language and the character and frequency of switching behaviour. Chapter Three, Empirical Analysis of Code-Switching in Polish-English Informal Conversations, presents the original empirical investigation undertaken for the present thesis, involving a sample of twenty bilingual adults whose informal conversational interactions were recorded, transcribed, and analysed. The methodological approach, the structural patterns identified in the data, and the functional analysis of switching instances are presented in detail, and the findings are discussed in relation to the theoretical frameworks reviewed in the preceding chapters. The concluding chapter offers a synthesis of the principal results, reflects upon the theoretical contributions of the study, acknowledges its limitations, and identifies directions for future research.
It is the contention of this thesis that an examination of code-switching among Polish-English bilinguals in informal conversational contexts yields insights of considerable theoretical and empirical value, illuminating both the structural regularities that govern language alternation and the social meanings that such alternation encodes and communicates. The study proceeds from the conviction, well-established in the sociolinguistic literature, that the languages in a bilingual's repertoire do not constitute hermetically sealed and independent systems but are instead dynamically integrated resources whose deployment in interaction is shaped by, and in turn shapes, the social world within which speakers are embedded. Chapter One begins by establishing the theoretical foundations necessary for the analysis that follows.
Chapter 1: Theoretical Foundations of Code-Switching in Bilingual Communication
1.1. Defining Code-Switching: Terminological and Conceptual Distinctions
The concept of code-switching has undergone substantial theoretical refinement since its initial systematic treatment in the mid-twentieth century, reflecting broader shifts in sociolinguistic theory from structuralist to interactional and, more recently, post-structuralist paradigms. The term "code" itself was introduced into linguistics through a functionalist tradition that conceptualised language varieties as discrete systems employed by speakers in accordance with contextual demands. Bernstein's influential work in the 1960s drew attention to the socially stratified nature of linguistic codes, distinguishing between restricted and elaborated variants associated with different social positions and communicative contexts [13]. Although Bernstein's framework was primarily directed at monolingual variation within English, it established a theoretical vocabulary that would prove productive for subsequent bilingual research. The decisive extension of code theory to bilingual speech communities was accomplished principally by Gumperz, whose foundational study of Norwegian speakers in Hemnesberget, conducted in collaboration with Blom, demonstrated that speakers systematically alternated between local dialect and standard Norwegian in ways that conveyed social meaning beyond mere referential content [14]. This observation established the principle that shifts between codes are not random deviations from a monolingual norm but are instead structured communicative acts governed by social and contextual variables. The theoretical implications of this insight have shaped decades of subsequent research and continue to inform contemporary debates about the nature of bilingual speech.
The definitional landscape of code-switching is complicated by its proximity to a cluster of related but analytically distinct phenomena, each of which requires careful delineation. The term code-mixing has been employed in some traditions, particularly in South Asian linguistics, to refer to the intersentential and intrasentential blending of two or more languages within a single utterance or discourse stretch, without necessarily implying any degree of speaker intentionality or sociolinguistic motivation [19]. However, the majority of contemporary researchers prefer to subsume such phenomena under the broader rubric of code-switching, reserving "mixing" for descriptive rather than analytical purposes. Borrowing, by contrast, is conventionally understood to involve the incorporation of lexical items from one language into the morphophonological and syntactic system of another, typically at the level of nonce or established loanwords [36]. Haugen's early work on Norwegian-American bilingualism was instrumental in establishing this distinction: he observed that the integration of English-origin words into Norwegian morphology constituted borrowing at the lexical level rather than switching between grammatical systems. Weinreich similarly distinguished between interference — defined as instances of deviation from the norms of either language resulting from a bilingual's familiarity with multiple codes — and deliberate alternation between codes as a pragmatically purposive act [52]. These distinctions, while contested in subsequent literature, continue to provide a productive baseline for terminological discussion.
The concept of translanguaging, developed most prominently by García, represents a more radical departure from the code-switching framework in that it challenges the very premise of discrete, bounded language systems [32]. For translanguaging theorists, bilinguals do not possess two separate systems between which they alternate; rather, they deploy a unified, flexible repertoire from which features are selected according to communicative need. This perspective aligns with Grosjean's holistic model of bilingualism, which similarly criticises the "fractional view" that judges bilinguals against monolingual norms in each of their languages, arguing instead that bilinguals constitute a distinct and complete communicative type in their own right [34]. The deficit framing of code-switching — the view, prevalent in earlier educational and popular discourse, that alternating between languages reflects linguistic inadequacy or incomplete acquisition — has been decisively refuted by empirical research demonstrating that code-switching is systematically governed by complex grammatical and pragmatic constraints accessible only to those with high competence in both languages [1]. Contemporary research thus adopts a resource-oriented perspective, understanding code-switching as a sophisticated communicative resource through which bilingual speakers exercise considerable pragmatic and identity-related agency, rather than as a symptom of linguistic deficiency or failure to conform to monolingual norms.
Auer's influential contribution to the definitional debate introduced the term language alternation as a cover term encompassing all instances of bilingual speech behaviour, with code-switching proper constituting one subtype characterised by its sequential organisation within conversation [22]. This terminological choice reflects Auer's commitment to a participant-relevant, interactional definition of switching: only those language alternations that are oriented to by participants as meaningful sequential phenomena qualify as code-switching in the strict sense. Poplack's seminal study of Spanish-English bilinguals in New York established both an influential typological framework and a set of grammatical constraints on switching that remain foundational to the field [51]. For the purposes of the present thesis, code-switching is defined as the systematic alternation between two or more linguistic codes — Polish and English in the corpus under examination — within a single conversation or utterance by a speaker who possesses functional competence in both codes. This definition excludes single-word borrowings that have been fully integrated into Polish morphology and does not require intentionality on the part of the speaker, recognising that both deliberate and automatic switching constitute valid objects of sociolinguistic analysis. The chosen definition thus integrates structural, functional, and interactional criteria in a manner appropriate to the empirical and theoretical objectives of the present investigation.
1.2. Typologies of Code-Switching: Intersentential, Intrasentential, and Tag-Switching
The typological classification of code-switching has been a central preoccupation of bilingualism research, with several competing frameworks proposing different criteria for categorising the formal and positional properties of language alternation. Poplack's tripartite distinction between intersentential, intrasentential, and tag-switching remains the most widely cited and applied taxonomy in the empirical literature [50]. Intersentential switching occurs at clause or sentence boundaries, such that each clause belongs unambiguously to one language system; intrasentential switching, the most grammatically complex type, occurs within a single clause or sentence and requires the simultaneous activation of two grammatical systems; and tag-switching involves the insertion of a formulaic tag — an interjection, sentence adverb, or idiomatic phrase — from one language into an utterance otherwise produced in the other [3]. Research consistently identifies intrasentential switching as the most frequent type in fluent bilinguals, accounting for as much as 68 per cent of switching instances in some corpora, though the proportion varies significantly across speech communities and interactional contexts [3]. The relative rarity of tag-switching in more proficient bilinguals is interpreted as evidence that this type functions as an introductory strategy for speakers with lower competence in the embedded language, who are not yet equipped for the more demanding intrasentential variety.
Poplack also proposed two major grammatical constraints on code-switching: the Free Morpheme Constraint and the Equivalence Constraint. The Free Morpheme Constraint holds that a switch may not occur between a bound morpheme and a lexical form unless the latter has been phonologically integrated into the language of the bound morpheme; the Equivalence Constraint states that switches may occur only at points in discourse where the surface structures of both languages are mapped onto each other, such that the word order at the switching point is grammatical in both languages simultaneously [50]. These constraints have been extensively tested and partially contested in subsequent research. The application of the Equivalence Constraint to Polish-English switching presents particular theoretical challenges, given the substantial typological distance between the two languages: Polish is a synthetic language characterised by a rich nominal and verbal inflection system and relatively free word order governed by information-structural principles, while English is an analytic language with fixed SVO word order and minimal inflectional morphology. Switching at the verb phrase boundary, for instance, may require reconciliation of Polish aspect marking with English tense and aspect constructions — a point of non-equivalence that generates important empirical predictions regarding the distribution of switching in Polish-English bilingual speech and challenges the universalist aspirations of Poplack's original formulation.
Myers-Scotton's Matrix Language Frame (MLF) model offers a structural account of how two languages are integrated in intrasentential switching [19]. The MLF model posits that in any mixed-code utterance, one language functions as the Matrix Language (ML), providing the morphosyntactic frame — including the grammatical morphemes and word order — while the other language functions as the Embedded Language (EL), contributing content morphemes and lexical items. This asymmetric relationship between the two languages explains why switches are not randomly distributed across syntactic positions but tend to cluster at phrase boundaries and to respect the morphosyntactic requirements of the matrix language. Two associated principles — the Morpheme Order Principle, which holds that surface morpheme order in mixed constituents follows that of the ML, and the System Morpheme Principle, which holds that grammatical morphemes in mixed constituents come from the ML — generate specific predictions about the structural form of switching events in typologically asymmetric language pairs. In Polish-English bilingual speech, Polish typically functions as the matrix language in informal conversation among diaspora community speakers, with English content morphemes inserted into a Polish syntactic frame, though the directionality of this relationship shifts in speech communities where English has become the dominant language of professional life [1].
Muysken's alternative typology distinguishes three primary code-switching types: alternation, which corresponds broadly to Poplack's intersentential switching and involves the juxtaposition of complete stretches of two languages; insertion, which parallels intrasentential switching and involves the embedding of lexical or phrasal material from one language into the grammatical frame of another; and congruent lexicalisation, a third type in which two languages share a grammatical structure, allowing lexical items from either language to be deployed within the same syntactic frame [18]. This three-way distinction has proven particularly productive in research on language pairs whose typological properties differ significantly: insertion is predicted to predominate in structurally distant pairs, while congruent lexicalisation is expected in typologically similar languages. In Arabic-English bilingual speech, the typological distance between the two languages shapes the relative prevalence of each switching type, with insertion being the predominant mode given the considerable structural divergence between Arabic morphosyntax and English analytic grammar [2]. Auer's sequential approach, by contrast, brackets the question of linguistic structure and focuses on the interactional significance of code alternation, examining how participants orient to language switches as resources for managing turn organisation, topic transitions, and relational alignment [21].
| Framework | Primary Types Proposed | Core Analytical Focus | Grammatical Constraints | Applicability to Polish-English |
|---|---|---|---|---|
| Poplack (1980) | Intersentential; Intrasentential; Tag-switching | Structural position of switch within utterance | Free Morpheme Constraint; Equivalence Constraint | Partially applicable; Equivalence Constraint challenged by morphological asymmetry of Polish and English |
| Myers-Scotton (1993) | Matrix Language frame vs. Embedded Language insertions | Morphosyntactic frame assignment and asymmetry | Morpheme Order Principle; System Morpheme Principle | Highly applicable; Polish typically serves as ML in diaspora informal speech |
| Muysken (2000) | Alternation; Insertion; Congruent Lexicalisation | Typological properties of the language pair | Type prevalence conditioned by structural distance | Insertion most prevalent; congruent lexicalisation limited by structural distance between Polish and English |
| Auer (1998) | Language alternation (cover term); code-switching as interactionally relevant subtype | Sequential organisation and participant orientation in conversation | No structural constraints proposed; interactional relevance criterion applied | Productive for analysing conversational significance of switches in informal interaction |
The methodological challenges involved in classifying switching instances according to these typologies are considerable and require explicit treatment before any empirical analysis is undertaken. The boundary between borrowing and code-switching at the lexical level is frequently contested, particularly in the case of established English-origin words in Polish professional discourse — such as meeting, deadline, or feedback — which may be variably treated as integrated loanwords or as switch points depending on the analytical framework adopted and the degree of morphophonological adaptation observable in the speech sample. Similarly, the identification of the matrix language in highly fluent bilinguals who switch frequently may be ambiguous, especially in cases where structural parallelism between the two languages makes it difficult to assign morphosyntactic frame-setting status to either with confidence. These challenges have significant methodological implications for the empirical component of the present thesis, which will adopt a combined analytical approach drawing on Poplack's typological categories and Muysken's refinements while remaining attentive to Auer's interactional perspective on the sequential significance of language alternation events.
1.3. Sociolinguistic Models of Code-Switching: Gumperz, Myers-Scotton, and Beyond
Gumperz's sociolinguistic approach to code-switching, developed most fully in his 1982 monograph Discourse Strategies, remains one of the most influential and widely applied frameworks for understanding the communicative functions of language alternation in bilingual interaction [16]. A central contribution of Gumperz's work was the distinction between situational and metaphorical code-switching. Situational switching refers to instances where language alternation is determined by changes in the external situation of utterance — the interlocutor, topic, setting, or activity type — and where a different choice of language would be perceived as inappropriate or socially marked. Metaphorical switching, by contrast, occurs within a single situation whose social parameters remain constant and serves to invoke the values, attitudes, or social identities associated with the switched language. Metaphorical switching is thus the more communicatively complex phenomenon: through a momentary switch to one language within a conversation otherwise conducted in another, speakers can signal solidarity, distance, authority, or group affiliation without these meanings being made explicit in propositional content. The analytical significance of this distinction lies in the observation that the communicative work performed by code-switching extends far beyond the merely referential, encompassing a rich range of social and relational meanings that are recoverable by participants through shared sociolinguistic knowledge.
The we-code and they-code distinction, developed by Gumperz in collaboration with Blom, provides a complementary framework for understanding the social meaning of code choice in bilingual communities [14]. The we-code is the minority or in-group language, associated with solidarity, intimacy, and ethnic or community identity, while the they-code is the majority or out-group language, associated with formal, institutional, and public domains. Applied to Polish-English bilingual speakers in a migrant context, Polish typically functions as the we-code — the language of home, family, and ethnic solidarity — while English functions as the they-code, associated with the workplace, public institutions, and majority society. However, this binary has been substantially complicated by subsequent research demonstrating that the symbolic values associated with particular languages are not fixed but are negotiated moment by moment in interaction and are subject to historical change as communities' social circumstances evolve. In communities where English functions both as a prestige international language and as the medium of everyday professional life, its we-code and they-code status may be ambiguous or contested, and speakers may exploit this ambiguity strategically in constructing hybrid, cosmopolitan identities that transcend simple us-versus-them binaries [1]. Urbaite's research on multilingual identity construction in European contexts illustrates precisely how Polish-English bilinguals employ code-switching not merely to signal membership in pre-existing ethnolinguistic categories but to construct fluid identities that draw simultaneously on multiple cultural and linguistic resources [1].
Myers-Scotton's Markedness Model, first presented in her 1993 study of bilingualism in sub-Saharan Africa, offers a rational-actor account of code choice in bilingual interaction that has been widely applied in diverse sociolinguistic contexts [19]. The model proposes that speakers are aware, through socialisation into their speech community, of the rights and obligations (RO) sets associated with different languages in that community — the social relationships, roles, and interactional frames that each language conventionally indexes in a given situation. Code choice is understood as a strategic act through which speakers negotiate their desired social identity and relational stance by selecting the code whose RO set most closely corresponds to their communicative intentions. An unmarked code choice is one that conforms to community expectations for a given situation; a marked choice deviates from these expectations and thereby foregrounds the speaker's communicative intentions, whether those involve the assertion of authority, the claim of in-group solidarity, or the negotiation of a novel relational frame. This framework has been criticised for overemphasising rationality and deliberate agency in code choice while underemphasising the role of automatic, habit-driven, or unconscious switching and the constraints imposed by processing factors, but it remains a productive analytical tool for examining the social motivations of systematic code choice patterns.
Auer's interactional sociolinguistic approach to code-switching diverges significantly from both Gumperz's and Myers-Scotton's frameworks in its methodological commitment to analysing language alternation from the participants' perspective rather than imposing external analytical categories derived from the analyst's knowledge of the social meanings associated with each language [21]. Drawing on conversation analysis, Auer argues that the meaning of a language switch is not predetermined by its structural position or by the social meanings abstractly associated with each language in the community, but is instead constructed locally and sequentially in the unfolding of interaction. A switch acquires meaning through the interpretive work of co-participants, who respond to language choice as a contextualisation cue that signals a shift in footing, topic, or relational alignment. Li Wei's "moment of analysis" concept extends this interactional perspective by focusing on the real-time cognitive and social processes through which bilinguals deploy their multilingual resources, emphasising the creative and agentive dimensions of code-switching as a form of innovative linguistic practice that resists reduction to pre-formed social categories or deterministic structural rules [17]. Together, these interactional approaches provide an important corrective to frameworks that treat social context as an independent variable that mechanically determines code choice.
The concept of translanguaging, developed extensively by Canagarajah and García, represents the most theoretically ambitious rethinking of the foundations of code-switching research [29]. Translanguaging theorists contest the premise, shared by all the frameworks discussed above, that bilingual speakers operate with two distinct, bounded linguistic systems between which they switch. Instead, they propose that multilingual individuals draw upon a single, integrated linguistic repertoire characterised by fluid movement across what are conventionally regarded as language boundaries, and that the very notion of "a language" as a discrete system is a socially constructed and politically maintained fiction rather than a psycholinguistic reality. This perspective has important implications for language education and critical linguistics, but it raises methodological challenges for empirical research that seeks to document and analyse specific switching events in terms of their formal and functional properties — challenges to which the present thesis will attend through a critical synthesis of structural, sociolinguistic, and interactional perspectives. The empirical framework adopted in Chapter 2 draws selectively on each of the models reviewed here, treating them as complementary analytical lenses that together provide a richer account of code-switching behaviour than any single framework can supply.
1.4. Motivations and Functions of Code-Switching in Informal Speech
The question of why bilingual speakers alternate between languages in informal conversation has occupied researchers since the earliest systematic studies of bilingual behaviour, and the literature has produced a rich taxonomy of functional motivations that spans cognitive, affective, social, and pragmatic dimensions. Gumperz's seminal analysis identified six primary conversational functions of code-switching: quotation, whereby a speaker switches to reproduce the language in which an original utterance was made; addressee specification, which signals that a particular turn is directed at a specific co-participant associated with the switched language; interjections, in which a switch introduces exclamations or discourse fillers; reiteration, where a message is repeated or reformulated in the second language for emphasis or clarification; message qualification, in which comments and elaborations are produced in the switched language; and personalisation versus objectification, which involves using one language to mark personal involvement and the other to create distance or generality [16]. This taxonomy has been broadly confirmed and extended by subsequent empirical research across diverse bilingual speech communities, including Indonesian university students whose code-switching on social media was shown to serve functions of identity marking, social closeness, humour, and topic shifting, with intrasentential switching overwhelmingly predominant [3]. The convergence across geographically and typologically diverse communities supports the hypothesis that many of the functional motivations for code-switching are universal features of bilingual communicative competence rather than community-specific conventions.
The relationship between code-switching and identity construction has been explored in depth by sociolinguists working within the interactional and post-structuralist traditions. Bucholtz and Hall's theoretical framework for the analysis of identity in interaction proposes that identity is not a pre-existing attribute that speakers bring to conversations but an emergent, intersubjective achievement accomplished through the deployment of linguistic and semiotic resources [28]. Code-switching is among the most salient of these resources: through the choice of language at particular interactional moments, speakers position themselves in relation to social categories, group affiliations, and ideological alignments. Urbaite's research on hybrid identity construction in European multilingual contexts demonstrates how code-switching serves as a primary marker of complex, non-unitary identities that cannot be reduced to affiliation with a single ethnolinguistic group, and how the act of switching itself constitutes a performance of hybridity and cosmopolitan orientation that distinguishes multilingual speakers from their monolingual peers [1]. For Polish-English bilinguals, switching between the two languages may serve simultaneously to claim membership in a Polish-speaking community while also signalling affiliation with a transnational social network in which English functions as a shared professional lingua franca.
The affective dimensions of code-switching have received sustained attention in the work of Pavlenko, who has argued that the two languages of a bilingual are not emotionally equivalent but carry different affective associations deriving from the contexts and relationships in which each language was acquired [48]. For speakers who acquired Polish as a first language in the domestic and community context and English as a second language in formal educational or professional settings, Polish may carry stronger affective resonance for the expression of intimate emotions, spontaneous reactions, and in-group solidarity, while English may be preferred for more distanced, analytical, or humorous registers. Code-switching thus enables bilinguals to modulate the emotional colouring of their speech with considerable precision: a switch from Polish to English in the middle of an emotionally charged exchange may signal ironic distance, a retreat to formality, or a humorous reframing of the situation. Conversely, a switch from English to Polish in a professional context may activate an intimate, in-group footing that transcends the formal demands of the setting and signals an appeal to shared cultural and emotional experience. These affective functions are particularly relevant to the analysis of informal conversation, where speakers are less constrained by institutional norms and more free to exploit the full range of social and emotional meanings that code-switching can convey.
The pragmatic economy of code-switching — its utility as a mechanism for accessing specialised lexical and conceptual resources concentrated in one of the bilingual's languages — constitutes a further important dimension of switching motivation in contemporary Polish-English speech. The extensive penetration of English-origin terminology into Polish professional, academic, and technological discourse has created a situation in which English lexical items frequently occupy semantic niches for which Polish equivalents are either absent, less precise, or stylistically marked as awkward calques [1]. In the domains of information technology, business, and academic research, Polish-English bilinguals may switch to English not only for lexical gap-filling but also to invoke the epistemic authority and international legitimacy associated with English-language professional discourse. Kniaź's research on Arabic-English switching in academic contexts similarly documents how bilingual speakers switch to English when discussing topics for which English-language terminology is better established and more widely recognised within their professional communities [2]. This type of switching — motivated by pragmatic efficiency rather than identity assertion — occupies an intermediate position between deliberate code choice and unconscious lexical access, raising theoretical questions about the degree to which speakers exercise intentional control over their language-switching behaviour.
Grosjean's activation threshold model of bilingual language processing offers a cognitive account of the conditions under which automatic or unintentional switching occurs [34]. The model proposes that each word in a bilingual's mental lexicon has an activation threshold that must be exceeded before the item is available for production; in a bilingual mode, both languages are simultaneously activated to varying degrees, and items from the non-target language may be produced when their activation level exceeds that of target-language equivalents due to frequency, recency, or associative priming effects. This cognitive perspective complements sociolinguistic accounts by explaining why some switching instances may be experienced by speakers as unintentional or automatic — the product of lexical retrieval processes that temporarily favour the non-target language — while others constitute deliberate communicative strategies oriented to social and pragmatic goals. The distinction between intentional and automatic switching has important implications for the analysis of the Polish-English corpus examined in Chapter 3, where patterns of switching will be interpreted in relation to both the social contexts of utterance and the cognitive demands associated with different topics and interactional frames. A comprehensive account of code-switching motivations must integrate both cognitive and sociolinguistic perspectives, recognising that any given switching event may be simultaneously motivated by psycholinguistic processes and oriented to social functions.
1.5. Code-Switching in the Context of Polish-English Bilingualism
Polish-English bilingualism is not a homogeneous phenomenon but encompasses a range of linguistically and socially distinct speaker populations whose code-switching practices reflect divergent acquisition histories, social circumstances, and attitudinal orientations. The oldest stratum of the Polish-English bilingual community in the United Kingdom consists of the post-war diaspora of the 1940s and 1950s, whose settlement in Britain was precipitated by the political consequences of the Second World War and the subsequent consolidation of Communist governance in Poland. This community, characterised by strong institutional maintenance of the Polish language through supplementary schools, churches, and cultural associations, developed a distinctive variety of bilingual speech in which Polish served as the primary medium of domestic and community life while English was employed in public, professional, and institutional domains [31]. A second and numerically significant stratum emerged following the political and economic transformations of 1989, as educational and professional migration from Poland intensified throughout the 1990s in conditions of newly opened borders and increased access to Western European labour markets. The most substantial expansion of the Polish-English bilingual population, however, was precipitated by Poland's accession to the European Union in 2004, which enabled freedom of movement and resulted in the migration of large numbers of Polish nationals to the United Kingdom, constituting within a decade one of the largest migrant communities in British demographic history.
The structural typological distance between Polish and English creates a rich and theoretically productive site for examining the grammatical constraints on intrasentential code-switching. Polish is a West Slavic synthetic language characterised by a seven-case nominal system, elaborate verbal aspect distinctions, grammatical gender concord across nominal and adjectival phrases, a complex system of verbal conjugation, and relatively free word order governed by topic-comment and focus-background information structure. English, by contrast, is a Germanic analytic language with a largely fixed subject-verb-object word order, minimal inflectional morphology, and grammaticalised tense distinctions without an aspectual opposition comparable to the Slavic perfective and imperfective contrast. The implications of this typological asymmetry for switching constraints are considerable: the insertion of English nominal or verbal elements into Polish-frame utterances requires the bilingual to assign Polish case, gender, and sometimes aspectual marking to English-origin material, producing morphophonological adaptations that blur the boundary between borrowing and code-switching at the lexical level [19]. The Equivalence Constraint, which requires structural parallelism at switching points, is systematically challenged or inapplicable across large domains of Polish-English switching, suggesting that the constraint's universality requires significant qualification in cases of structurally distant language pairs and that the Polish-English case constitutes a valuable testing ground for the empirical adequacy of competing typological models.
Empirical research on Polish-English code-switching, while less extensive than the substantial literature on Spanish-English or French-English bilingualism, has established several important findings regarding the distribution and functions of switching in Polish-diaspora and heritage speaker communities. Duszak's research on Polish academic and professional discourse examined how Polish-speaking professionals navigated the competing demands of Polish and English institutional conventions, revealing how code-switching serves not only communicative but also epistemic and identity-related functions in contexts where the prestige of English as an international academic language intersects with the symbolic significance of Polish as a marker of national and professional identity [31]. Subsequent scholarship on post-2004 Polish migrants in the United Kingdom has documented how switching patterns shift over the course of a speaker's residency and integration into English-dominant professional networks, with increasing contact with English-speaking colleagues and social circles correlating with higher rates of English insertion and a gradual shift in the balance of the matrix language frame [46]. These studies collectively suggest that code-switching among Polish-English bilinguals is a dynamic phenomenon that reflects the ongoing renegotiation of linguistic identities and social affiliations in the context of migration, integration, and the shifting domains of language use over the life course.
The sociolinguistic ecology of the Polish language situates code-switching in a distinctive attitudinal context that differs markedly from the relatively tolerant orientations toward linguistic mixing found in some other bilingual communities. Poland has historically been characterised by a strong prescriptive orientation toward language purity, reinforced by the role of the Polish language as a symbol of national continuity and cultural resistance during the nineteenth-century partitions and subsequent periods of foreign-imposed political control. The Council of the Polish Language, established in 1996, monitors the use of Polish in public communication and has periodically issued guidelines discouraging the unnecessary incorporation of English loanwords and constructions into Polish. Popular discourse in Poland frequently expresses concern about the contamination of Polish by anglicisms, particularly in domains such as business communication, digital media, and youth language [1]. This prescriptive context generates an attitudinal tension in the practices of Polish-English bilinguals: while code-switching may be positively valued as a marker of cosmopolitanism, educational achievement, and professional competence within bilingual peer networks, it may also attract negative evaluation from prescriptivist interlocutors or in formal contexts where adherence to Polish language norms is strictly enforced. The attitudinal dimension of switching behaviour thus constitutes a significant variable in any comprehensive analysis of Polish-English bilingual speech, one that mediates the relationship between the structural possibilities for switching and the actual switching behaviour observed in naturalistic data.
Heritage speakers — Polish-English bilinguals who acquired Polish as a first language in a domestic context in the United Kingdom but English as the dominant language of schooling and public life — present a qualitatively distinct switching profile from that of adult immigrants who acquired English after completing their primary and secondary education in Poland. Heritage speakers' Polish may exhibit structural simplifications attributable to incomplete acquisition or to contact-induced attrition, including reduced case morphology, simplified aspect usage, and increased reliance on English-origin constructions; their switching patterns tend to reflect higher rates of English matrix language dominance and more extensive integration of English morphosyntactic frames into what is nominally a Polish-language utterance [1]. Adult immigrants, by contrast, typically maintain a more firmly Polish morphosyntactic frame in informal speech while inserting English lexical material in domains where English-origin terminology has become conventional and expected. These differences have significant methodological implications for corpus design and for the interpretation of switching patterns: a sample composed exclusively of heritage speakers would yield systematically different switching configurations from one composed of adult immigrants, and mixed-composition corpora require analytical frameworks capable of accommodating both populations while remaining sensitive to the linguistic and social factors that account for between-speaker variation. The empirical component of the present thesis, described in Chapter 2, draws on a corpus comprising both heritage speakers and post-2004 adult migrants, enabling a comparative analysis of switching behaviour across these two populations. The theoretical frameworks reviewed in the present chapter — the typological models of Poplack and Muysken, the sociolinguistic models of Gumperz and Myers-Scotton, and the interactional approaches of Auer and Li Wei — will provide the analytical vocabulary for the systematic description and sociolinguistic interpretation of the switching events documented in the corpus, while the functional inventory elaborated in section 1.4 will guide the analysis of the communicative motivations that underlie observed switching patterns in informal Polish-English conversation. The following chapter addresses the methodological procedures through which the corpus was compiled, the sampling criteria that governed speaker selection, and the analytical conventions adopted for the transcription and coding of switching events.
Chapter 2: Sociolinguistic and Pragmatic Dimensions of Code-Switching
The study of code-switching has expanded considerably beyond its original focus on grammatical structure and linguistic description. Contemporary scholarship situates the phenomenon within a broader sociolinguistic and pragmatic framework, attending to the social motivations, interactional functions, and ideological dimensions that shape bilingual language behaviour. The present chapter examines four interconnected aspects of this framework as they apply to Polish-English bilingual speakers in informal conversational contexts. First, the social functions of code-switching are analysed with particular reference to identity construction, solidarity, and relations of power. Second, the pragmatic dimensions of code-switching are explored, encompassing its roles in emphasis, mitigation, and conversational management. Third, attitudinal stances towards code-switching among bilingual speakers are reviewed, drawing on language ideology theory and empirical survey data. Finally, the relationship between code-switching and language proficiency is examined, with attention both to the conditions under which proficient bilinguals switch and to the contextual variables that mediate this relationship. Together, these four perspectives offer a comprehensive account of code-switching as a socially embedded, functionally complex, and ideologically contested communicative practice.
2.1. The Social Functions of Code-Switching: Identity, Solidarity, and Power
Among the most foundational contributions to the sociolinguistic study of code-switching is the distinction, advanced by John Gumperz, between situational and metaphorical switching [35]. Situational switching occurs when a change in the external social context — the interlocutor, the setting, or the topic under discussion — triggers a shift to a different language variety. Metaphorical switching, by contrast, involves a deliberate deployment of an alternative code within an otherwise unchanged situational frame, serving to import the social meanings conventionally associated with that code. This distinction has proved foundational for subsequent research, as it directs analytical attention away from purely structural descriptions of language alternation toward the social meanings that bilingual speakers exploit through their linguistic choices. In the context of Polish-English bilingualism, both situational and metaphorical switching are attested in informal interaction, though the latter is particularly salient in conversations among peers who share a common bicultural biographical trajectory and a facility with both languages.
The acts of identity framework, proposed by Le Page and Tabouret-Keller, provides a complementary theoretical lens through which code-switching may be understood as a form of social positioning and self-presentation [44]. According to this framework, each linguistic act constitutes an attempt by the speaker to identify with, or to distance themselves from, a social group whose norms the speaker seeks to emulate or to reject. For Polish-English bilinguals — particularly those who have lived, studied, or worked in English-speaking environments — the insertion of English lexical items, phrases, or entire conversational turns into predominantly Polish discourse may function as an assertion of cosmopolitan or transnational identity. Such acts simultaneously signal affiliation with international academic, professional, or popular cultural communities, and mark the speaker's self-positioning as distinct from a strictly monolingual Polish identity. The symbolic capital associated with English proficiency in post-communist Poland renders this form of identity performance particularly meaningful in informal peer interaction, where the bilingual speaker's linguistic repertoire is most freely deployed. The acts of identity perspective thus establishes that every code-switch is, at some level, a claim about who the speaker is and who the speaker wishes to be perceived as being.
The solidarity-marking function of code-switching has been extensively theorised since the work of Blom and Gumperz on Norwegian dialect switching in the coastal town of Hemnesberget [24]. Their observation that speakers switch to a local dialect to enact solidarity with co-present community members — even in the absence of any externally imposed situational trigger — established the principle that language choice is a vehicle for the active negotiation of interpersonal relationships rather than a passive reflex of situational parameters. In Polish-English bilingual dyads, switches to English frequently index shared experiences of study abroad, bilingual schooling, exposure to Anglophone digital culture, or participation in internationally oriented professional communities. The use of English among interlocutors who share this background constitutes an act of in-group marking, signalling mutual recognition of a shared biographical and cultural trajectory that sets the participants apart from monolingual Polish-speaking peers. Switches in the opposite direction — from a predominantly English conversational frame back to Polish — may equally mark solidarity in contexts where Polishness itself is the salient social identity, such as conversations about domestic cultural events, family life, or national concerns. The bidirectionality of solidarity-marking switches illustrates that the social function of code-switching cannot be reduced to a simple equation of English with prestige and Polish with vernacular intimacy.
Power relations and the differential prestige of languages constitute a further dimension of code-switching that demands sociological attention. Bourdieu's concept of linguistic capital posits that languages are not merely neutral communicative tools but forms of symbolic capital whose value is determined by the structure of the linguistic marketplace and by the historically sedimented hierarchies of prestige that govern social evaluation [26]. In Poland, the post-communist period has witnessed a dramatic elevation of English's symbolic value, driven by processes of economic liberalisation, European integration, and the globalisation of knowledge production and professional certification. English proficiency functions as a marker of educational attainment and professional aspiration, and this asymmetry of prestige permeates informal bilingual interaction in ways that are not always consciously acknowledged by participants. Code-switching into English may, in certain conversational contexts, be interpreted as an exercise of symbolic power — a subtle claim to superior educational capital or cosmopolitan experience — or, conversely, as a form of accommodation to an interlocutor whose English proficiency is greater or whose dominant language is English. The negotiation of linguistic capital through code-switching is thus simultaneously a reflection of and a contribution to the broader social distribution of prestige and distinction.
The latter interpretation connects directly to accommodation theory, originally developed by Giles and subsequently refined in a programme of experimental and observational research [33]. Convergence, in Giles's framework, typically signals approval, solidarity, and the desire to reduce social distance between interlocutors, while divergence signals a wish to maintain or amplify linguistic and social distinctiveness. In bilingual contexts, switching to the interlocutor's preferred or dominant language may be understood as an act of linguistic convergence, reflecting accommodation to perceived communicative preferences and an implicit recognition of the interlocutor's linguistic comfort zone. Divergence, by contrast, may manifest as the maintenance of a code that the interlocutor does not share, or as a deliberate amplification of linguistic features associated with one's own language group — a phenomenon observable, for instance, when a Polish-English bilingual insists on conducting a conversation in Polish with an interlocutor who has initiated communication in English. The accommodation framework thus adds a relational and dynamic dimension to the analysis of code-switching that complements the identity-based and power-oriented accounts discussed above.
It is important to emphasise that a single code-switch may simultaneously fulfil multiple social functions, a property that has been described as the multifunctionality of bilingual conversational behaviour. A single English lexical insertion in a Polish conversation may at once signal in-group solidarity, index a particular cultural reference inaccessible in Polish, assert symbolic capital, and accommodate the conversational norms of the interlocutor. The social meaning of a given switch is thus not pre-determined by the linguistic code itself but is locally negotiated through the sequential unfolding of conversation and through the interpretive frameworks that participants bring to the encounter. This multifunctionality is not an anomaly to be explained away but a constitutive feature of bilingual conversational practice, reflecting the richness of the symbolic resources available to speakers who command two linguistic codes and the social systems of meaning associated with each. Any adequate theoretical account of code-switching in Polish-English bilingual interaction must therefore be capable of accommodating the simultaneous co-presence of multiple social motivations within a single linguistic act.
2.2. Pragmatic Functions: Emphasis, Mitigation, and Conversational Management
Beyond its social indexical functions, code-switching serves a range of pragmatic purposes within ongoing conversation, shaping the communicative work accomplished by individual utterances and conversational sequences. Auer's interactional approach to code-switching, developed across several influential publications, relocates the analytical focus from the social meanings associated with linguistic codes to the sequential organisation of talk-in-interaction [12]. From this perspective, the significance of a switch is not determined by its pre-assigned social meaning but by its placement within an unfolding conversational sequence and by the interpretive work that it requires participants to perform. Auer distinguishes between discourse-related switching, which serves conversational management purposes, and participant-related switching, which indexes interlocutors' social identities or relationships. This distinction provides a useful analytical grid for examining the various pragmatic functions documented in studies of Polish-English bilingual conversation, as it permits a principled separation between switches that organise the structure of talk and those that perform social identity work.
The emphatic function of code-switching represents one of the most widely attested pragmatic motivations for language alternation in informal bilingual speech. Switches to a second language may serve to highlight, intensify, or rhetorically foreground particular elements of an utterance, drawing on the expressive resources of the alternate code to achieve a degree of communicative salience that the base language is perceived as unable to provide with equal effect. This function bears comparison with Jakobson's emotive or expressive function of language, which he identified as oriented toward the speaker's subjective state and the intensification of message significance [38]. In the Polish-English context, English exclamatives, intensifiers, and evaluative expressions — such as amazing, literally, no way, seriously, or whatever — are frequently embedded within Polish conversational turns to achieve an emphatic or affective effect that is perceived as more vivid, culturally resonant, or experientially authentic in English than in any available Polish equivalent. This phenomenon reflects the differential pragmatic resources of the two codes and the speakers' active exploitation of those resources for specific communicative ends, and it demonstrates that code-switching may be driven as much by the pragmatic need to achieve a particular communicative effect as by social identity considerations.
The mitigating function of code-switching constitutes a particularly significant domain of pragmatic analysis, connecting bilingual language behaviour to the politeness theory advanced by Brown and Levinson [27]. According to Brown and Levinson's framework, speakers employ a variety of positive and negative politeness strategies to manage face-threatening acts — communicative acts that potentially damage the self-image or the autonomy of the speaker or hearer. Code-switching may function as an off-record politeness strategy, in which a switch to a less socially dominant or less affectively charged code provides a degree of indirection, diffusing the force of a face-threatening act and reducing its potential for interpersonal damage. A speaker who delivers criticism, a refusal, or an awkward personal request in a language other than the conversational base language may thereby attenuate the social impact of the act, creating what might be described as a pragmatic buffer between the act and its recipient. In Polish-English bilingual conversation, switches to English in the performance of potentially embarrassing or confrontational speech acts — such as expressing disagreement, making an apology, or raising an emotionally sensitive topic — have been documented as a strategy for reducing interpersonal tension and preserving the interactional equilibrium between participants.
Conversational management functions represent a further category of pragmatic code-switching, encompassing a range of practices by which speakers organise, structure, and repair ongoing discourse. Topic shifts are among the most commonly cited conversational management functions: a switch of language code may signal that the speaker is initiating a new conversational topic, and interlocutors may orient to this signal in interpreting the relevance and frame of subsequent talk. Similarly, repairs — corrections of misunderstandings, false starts, or communicative breakdowns — may be accompanied by code switches that frame the repaired utterance as distinct from the preceding problematic turn, marking it as a fresh communicative attempt rather than a continuation of the disrupted sequence. The aside function of code-switching, in which a speaker briefly departs from the primary conversational frame to address a parenthetical remark to a co-present third party or to introduce a personal commentary, also constitutes a recognisable form of conversational management. The quotative function deserves particular mention: speakers who reproduce reported speech in a language other than the base language reproduce not merely the semantic content of the original utterance but the social voice and identity of the quoted individual, a practice that adds a layer of reported authenticity and dramatic immediacy to the narrative [11].
The phenomenon of triggered switching, in which a particular lexical item, proper noun, or culturally specific expression precipitates a broader shift in language choice, further illustrates the interaction between lexical, pragmatic, and grammatical constraints on code-switching. A speaker who introduces an English technical term, brand name, or cultural reference into Polish discourse may find that subsequent talk gravitates toward English, as the conceptual and lexical field associated with that item is more readily accessible and more fully elaborated in the language in which it was first introduced. This form of lexically triggered switching differs analytically from the compensatory, preferential, and pragmatic subtypes identified by Schächinger Tenés et al., who note that compensatory switching — in which a speaker switches to fill a perceived lexical gap — must be distinguished from forms of switching that occur in the absence of any competence-based motivation [5]. This typological distinction demonstrates that pragmatic motivations for switching cannot be reduced to deficit explanations, and that the interaction between pragmatic goals and grammatical constraints produces a complex landscape of bilingual conversational behaviour that resists simplistic characterisation. The intersection of pragmatic motivation and grammatical constraint thus constitutes a productive site of inquiry for future research on Polish-English code-switching in naturalistic settings, and calls for analytical frameworks capable of integrating structural, interactional, and pragmatic levels of analysis.
2.3. Attitudes Towards Code-Switching Among Bilingual Speakers
Attitudinal research on code-switching has documented a wide spectrum of evaluative positions, ranging from strongly negative deficit-oriented assessments to affirmative, resource-oriented perspectives that treat language alternation as a sophisticated and contextually appropriate communicative practice. The deficit view, associated with folk linguistic purism and certain prescriptive educational ideologies, frames code-switching as evidence of incomplete proficiency, linguistic laziness, or contamination of the base language by foreign elements. This view has been criticised extensively within sociolinguistics on both empirical and theoretical grounds, as research consistently demonstrates that proficient bilinguals engage in more structurally complex switching than low-proficiency speakers, and that code-switching serves functions that neither language alone could accomplish with equal efficiency or social meaning [4]. The resource-oriented view, by contrast, frames code-switching as a marker of communicative skill and as evidence of the bilingual speaker's capacity to draw on a rich and differentiated linguistic repertoire that monolingual speakers do not possess. The tension between these two evaluative frameworks is not merely academic but has practical consequences for language education policy, for the self-perception of bilingual speakers, and for the social valuation of bilingual communicative practices in institutional and everyday contexts.
The analysis of language attitudes is enriched by the concept of language ideology, understood as the systems of beliefs, values, and representations through which speakers interpret linguistic practices and assign them social meanings. Woolard has argued that language ideologies are not merely epiphenomenal reflections of social structure but active forces in the reproduction and transformation of social relations, shaping what counts as legitimate language use and who is authorised to make such determinations [53]. Kroskrity similarly emphasises that language ideologies are multiple, contested, and contextually specific, and that speakers may simultaneously hold conflicting ideological positions whose relative salience depends on the interactional frame in which linguistic behaviour is being evaluated [40]. In bilingual communities, language ideological tensions may manifest as ambivalence toward code-switching: speakers may express disapproval of switching in formal or semi-public contexts while practising and approving of it freely in informal peer conversation. The study of language ideologies thus provides a framework for understanding why attitudinal data on code-switching is rarely unidimensional and why the same individual may simultaneously regard code-switching as both a natural communicative behaviour and a socially marked departure from normative language use.
The Polish sociolinguistic context presents a particularly instructive case for the study of language ideological tensions around code-switching and language contact more broadly. The Polish Language Act of 1999 institutionalised a normative framework for the protection and promotion of the Polish language, reflecting political concerns about linguistic sovereignty and cultural integrity in the context of European integration and the globalisation of information culture. The Act's prescriptive provisions — which restrict the use of foreign languages in official, commercial, and public communication — do not formally govern private, informal speech, yet their symbolic force shapes public discourse about linguistic purity and renders English borrowing ideologically salient in ways that affect bilingual speakers' self-monitoring and self-presentation. Prescriptive purism, rooted in a Romantic conception of language as the embodiment of national identity and historical continuity, has historically dominated Polish language culture, producing a public discourse in which English influence is frequently represented as a threat to linguistic and cultural integrity. This ideological environment may generate feelings of ambivalence or linguistic insecurity among Polish-English bilinguals, particularly those with strong attachments to Polish cultural identity, who may perceive their code-switching as a departure from normative expectations even when it is functionally motivated, pragmatically effective, and socially unremarkable within their peer community.
Empirical studies of attitudes toward code-switching among bilingual students and young adults consistently report predominantly positive evaluations of the practice, particularly in informal and peer interactional contexts. Karimuddin et al. found that study participants regarded code-switching as a natural and beneficial communicative behaviour, with the majority expressing positive attitudes regardless of demographic background [7]. The finding that no significant correlation was observed between demographic variables — including gender, age, and educational background — and code-switching behaviour suggests that positive attitudes toward language alternation are broadly distributed across the studied population rather than confined to a particular social stratum or educational cohort [7]. Similarly, Balilla et al. report that pre-service teachers held positive attitudes toward code-switching as both a pedagogical and a communicative tool, viewing it as a natural strategy that facilitates communication and comprehension when full mastery of a single language is temporarily or contextually unavailable [6]. These findings align with the broader scholarly consensus that bilingual speakers increasingly conceptualise code-switching as a communicative resource rather than a deficit, a shift that reflects the diffusion of resource-oriented language ideological perspectives in multilingual educational and social contexts. The convergence of findings from geographically and culturally distinct bilingual communities lends the resource-oriented position considerable cross-contextual validity.
Labov's concept of linguistic insecurity, developed in his seminal study of social stratification and language change in New York City, provides a useful framework for understanding the contexts in which bilingual speakers may suppress or actively conceal their code-switching behaviour despite holding generally positive attitudes toward it [42]. Linguistically insecure speakers are characterised by a heightened awareness of the gap between their actual speech practices and the prestige norm, and by a tendency to over-correct or intensively monitor their speech in formal or high-stakes interactional contexts. Polish-English bilinguals who internalise prescriptive ideologies of linguistic purity may exhibit heightened self-monitoring of their code-switching in formal or semi-formal settings — academic seminars, job interviews, interactions with authority figures — suppressing switches that would occur freely in informal peer conversation. Metalinguistic awareness — the capacity to reflect consciously and analytically on one's own linguistic practices — may mitigate this insecurity among proficient bilinguals, who are better equipped to distinguish between contexts in which code-switching is socially sanctioned and those in which it is ideologically marked or strategically inadvisable. Bell's audience design framework, which posits that speakers style-shift primarily in response to their audience rather than to their topic or setting per se, further illuminates the attitudinal performance of bilingual speakers: code-switching may be calibrated to the perceived linguistic norms, expectations, and preferences of the interlocutor, rendering attitudinal stances not as fixed internal dispositions but as dynamic, audience-responsive and context-sensitive performances [23]. The distinction between private attitudes and public attitudinal performance is therefore fundamental to any adequate account of how bilingual speakers evaluate and deploy code-switching across social contexts.
| Attitudinal Position | Theoretical Framework | Characteristic Stance | Representative Empirical Findings |
|---|---|---|---|
| Deficit-oriented | Prescriptive purism; monolingual norm ideology | Code-switching as evidence of incomplete competence, linguistic laziness, or contamination of the base language | Less prevalent among proficient bilinguals; associated with formal educational settings and prescriptive language policy contexts; not supported by structural evidence [4][5] |
| Neutral / Pragmatic | Communicative competence theory (Hymes 1972); interactional sociolinguistics (Auer 1995) | Code-switching as a contextually motivated and functionally rational communicative choice, neither stigmatised nor valorised | Code-switching frequency not significantly correlated with proficiency level in some populations [4]; behaviour adapts to context and interlocutor [7] |
| Resource-oriented | Language ideology (Woolard 1998); translanguaging (García & Lin 2021); acts of identity (Le Page & Tabouret-Keller 1985) | Code-switching as sophisticated linguistic resource reflecting high bilingual competence and communicative creativity | Positive attitudes widely attested across contexts and demographic groups [6][7]; intrasentential switching correlates with higher proficiency [4][5] |
| Ambivalent | Linguistic insecurity (Labov 1966); audience design (Bell 1984); language ideology (Kroskrity 2000) | Positive attitudes in informal settings; suppression or concealment in formal contexts due to normative and ideological pressure; internal conflict between practice and prescription | Context-sensitivity of switching behaviour documented across social groups [7]; formal register and institutional settings inhibit switching frequency and complexity [6] |
2.4. Code-Switching and Language Proficiency: The Role of Competence and Context
The relationship between code-switching and language proficiency has been among the most debated questions in bilingualism research, and a substantial body of empirical and theoretical work has been devoted to refuting the deficit hypothesis — the claim that code-switching is primarily a compensatory strategy deployed by speakers whose proficiency in one or both languages is insufficient to sustain monolingual communication. This hypothesis, which dominated early popular and educational discourse on bilingualism and continues to exert influence in some institutional and ideological contexts, has been systematically challenged by evidence demonstrating that proficient bilinguals engage in more frequent and structurally complex switching than speakers with lower levels of bilingual competence. The deficit hypothesis conflates two analytically distinct phenomena: compensatory switching, which occurs when a speaker lacks the lexical or grammatical resources to express a particular meaning in the base language, and preferential or pragmatic switching, which occurs in the absence of any competence-based motivation and reflects the speaker's deliberate exploitation of the differential resources of the two codes [5]. The identification and consistent empirical documentation of the latter type is perhaps the most decisive refutation of the deficit view, as it establishes that code-switching persists and indeed intensifies as bilingual proficiency increases, a pattern that is directly incompatible with the deficit account.
Poplack's proficiency hypothesis represents one of the most influential formulations of the relationship between linguistic competence and code-switching behaviour. Based on her extensive study of Spanish-English code-switching in the Puerto Rican community of New York City, Poplack proposed that intrasentential switching — the insertion of elements from one language into the syntactic frame of another — requires a high degree of grammatical competence in both languages, as it demands the simultaneous mastery and co-activation of the morphosyntactic constraints operative in each code [20]. By contrast, intersentential switching — the alternation of entire sentences or utterances across language boundaries — imposes fewer simultaneous grammatical demands and is therefore accessible to speakers with more limited bilingual competence. This hierarchy of structural complexity has been corroborated by subsequent research conducted in a variety of bilingual communities and linguistic pairings, including the finding, reported by Aldalbahy, that a positive correlation obtains between the type of code-switching and the level of language proficiency among Saudi female university students, while frequency of switching does not show a statistically significant relationship with proficiency level [4]. This finding is theoretically significant, as it suggests that proficiency shapes the structural form of switching rather than its overall incidence, and that motivated switches occur across the proficiency spectrum, though their grammatical complexity varies systematically.
Contextual factors play a substantial mediating role in the relationship between proficiency and code-switching behaviour, and any account that focuses exclusively on individual-level competence while neglecting situational and interactional variables will necessarily be incomplete. The domain of interaction — whether academic, familial, professional, or informal peer social — exerts a determining influence on both the frequency and the type of switching that occurs, as different domains are associated with different norms of language use and with different expectations about the appropriateness and social meaning of language alternation. Topic also constitutes a significant contextual variable: topics that are habitually discussed in one language — such as academic or professional subjects encountered primarily through English-medium education or work — may trigger switching even among speakers who are fully capable of discussing those topics in Polish, because the relevant conceptual vocabulary and associated frames of reference are more automatically activated in the language of primary exposure. The identity of the interlocutor is perhaps the most powerful single contextual variable, as speakers calibrate their language choices to the perceived linguistic background, communicative preferences, and social identity of their conversational partner, a process that reflects the audience design logic described in the preceding section. Communicative competence in Hymes's sense — the knowledge of when, how, and with whom to use language in socially appropriate ways — is thus as important as purely structural linguistic competence in determining the form and frequency of code-switching behaviour [37].
Developmental aspects of bilingualism introduce a further dimension of complexity into the proficiency-switching relationship, as the age of acquisition and the conditions of bilingual development shape the architecture of the bilingual mental lexicon and the automaticity of language control mechanisms. Late bilinguals — those who acquire their second language after puberty, typically in formal educational or professional contexts — display code-switching patterns that differ systematically from those of simultaneous or early bilinguals, who acquire both languages from birth or from early childhood in naturalistic interactional settings. Late bilinguals are more likely to exhibit lexically motivated compensatory switching, particularly in domains where their second-language vocabulary has not been fully automatised, while early bilinguals may engage more readily in pragmatically or socially motivated switching across all domains and at higher levels of structural complexity. Schächinger Tenés et al. report that positive but not negative language proficiency effects on code-switching frequency are observed in bilingual child populations, indicating that higher proficiency in either language facilitates switching while lower proficiency does not simply suppress it in a symmetrical fashion [5]. This asymmetric relationship between proficiency and switching suggests that the motivation to switch is not reducible to a simple gap-filling mechanism but reflects the complex interplay of social, pragmatic, and developmental factors that characterise bilingual language behaviour across the lifespan and across different types of bilingual experience.
The distinction between perceived lexical gaps and pragmatically or socially motivated switching has important implications both for the theoretical understanding of code-switching and for the empirical investigation of bilingual language behaviour in naturalistic settings. Speakers who report switching "because they do not know the word" in the base language — a self-assessment apparently consistent with the deficit hypothesis — may in fact be engaging in a form of switching that is primarily socially or pragmatically motivated, using the perceived lexical gap as a convenient post-hoc rationalisation for a switch that was driven by other considerations. The Control Process Model, proposed by Green and Wei, offers a cognitive account of this complexity: it posits that both languages are simultaneously co-activated in the bilingual speaker's mental lexicon, and that the selection of one language over the other at any given moment is determined by the relative activation strengths of competing lexical and syntactic representations [15]. The language with the stronger representation for a given concept, context, or communicative intention wins the competition for selection, producing a code-switch that may appear from the outside to be motivated by a lexical gap but is in fact the product of a more complex cognitive and social process. This model aligns with the empirical finding that even highly proficient bilinguals — who have no communicative need of compensatory strategies — switch regularly and in structurally sophisticated ways, driven by the pragmatic, social, and affective affordances of the alternate code [4][5]. The Control Process Model thus integrates cognitive, social, and contextual explanations of code-switching in a framework that is both theoretically coherent and empirically tractable.
The findings reviewed in this section collectively establish that code-switching among Polish-English bilinguals is neither a symptom of communicative deficit nor a simple reflex of situational context, but a sophisticated and functionally differentiated practice that reflects the full complexity of bilingual competence, social identity, and interactional pragmatics. The relationship between proficiency and switching is modulated by multiple contextual variables — domain, topic, interlocutor, developmental history, and ideological orientation — in ways that resist reduction to any single explanatory framework. A comprehensive account of code-switching in Polish-English informal conversation must therefore draw on structural, sociolinguistic, pragmatic, cognitive, and ideological perspectives simultaneously, attending to both the competence of individual bilingual speakers and the social conditions of the interactional contexts in which their switching behaviour is produced and interpreted. The following chapters of the present thesis take up this challenge by examining empirical evidence from natural speech data, exploring the specific linguistic and social patterns that characterise Polish-English code-switching in informal conversational settings and their implications for broader theories of bilingual communication.
Chapter 3: Empirical Analysis of Code-Switching in Polish-English Informal Conversations
3.1. Methodology: Data Collection, Participant Profile, and Analytical Framework
The empirical investigation presented in this chapter is grounded in a mixed-methods research design that combines quantitative corpus analysis with qualitative pragmatic interpretation. The primary objective of the study was to examine the structural and functional dimensions of code-switching as manifested in naturally occurring informal conversations between Polish-English bilingual adults resident in Poland and the United Kingdom. The methodological approach adopted in this study draws on established traditions within corpus sociolinguistics and conversational analysis, while simultaneously incorporating frameworks developed for the analysis of bilingual speech in minority and diaspora communities. The decision to employ a naturalistic data collection procedure, as opposed to experimental or elicitation-based methods, was motivated by the well-documented finding that metalinguistic awareness and observer effects substantially alter code-switching behaviour when participants are aware of formal elicitation conditions [41].
The corpus was assembled from audio-recorded informal conversations conducted between May and September of the preceding academic year. A total of twenty bilingual adults participated in the study, ranging in age from eighteen to forty-five years. Participation was voluntary, and all participants provided written informed consent prior to data collection, in accordance with the ethical guidelines stipulated by the institutional review board of the affiliated university. Participants were assured of full anonymity, and all personally identifying information — including proper names, addresses, and institutional affiliations — was pseudonymised during transcription. The audio recordings were made using non-obtrusive portable digital devices, and participants were encouraged to converse naturally on topics of their own choosing. Recording sessions lasted between five and twenty minutes each, yielding an aggregate corpus of approximately two hundred minutes of spontaneous speech across forty-two recorded sessions.
The participant group was deliberately constituted to reflect the heterogeneity of the Polish-English bilingual population. Ten participants identified as heritage speakers — individuals raised in households where Polish was the dominant domestic language but who had been educated entirely or predominantly in English-medium institutions. Six participants were classified as late bilinguals who had acquired English as a second language following the age of twelve, primarily through formal instruction and subsequent migration to English-speaking environments. The remaining four participants were simultaneous bilinguals who reported having been exposed to both Polish and English from infancy, typically as a consequence of cross-national family arrangements. The group was gender-balanced, comprising ten female and ten male participants. Participants were further stratified by level of formal education: nine held undergraduate degrees, seven held postgraduate qualifications, and four were current undergraduate students at the time of data collection. This distribution was designed to facilitate the examination of educational attainment as a potential conditioning variable on code-switching behaviour, in keeping with approaches adopted in comparative research on bilingual communities elsewhere [9].
Transcription of the audio corpus followed conventions adapted from two established methodologies. The broad structural notation of Conversation Analysis, as codified by Jefferson [39], was employed to mark turn-taking boundaries, overlapping speech, pauses, and prosodic emphasis. In addition, the CHAT coding system developed within the framework of the Child Language Data Exchange System was adapted for the representation of bilingual utterances, permitting the systematic annotation of language choice at the utterance, clause, and lexical levels [45]. Each token in the corpus was assigned a language tag — either [POL] for Polish or [ENG] for English — and switching points were identified and categorised in accordance with the typological classification proposed by Poplack [49].
The analytical framework applied to the corpus integrates two complementary theoretical approaches. At the structural level, Poplack's three-way typological taxonomy — distinguishing intersentential switching, intrasentential switching, and tag-switching — provided the primary classificatory scheme for quantifying and describing the distribution of switching types across the corpus. At the functional level, Gumperz's taxonomy of conversational code-switching functions served as the basis for qualitative interpretation of pragmatic and social motivations underlying individual switching instances [35]. These two frameworks were treated as complementary rather than competing: while Poplack's typology characterises the grammatical locus and structural complexity of switches, Gumperz's taxonomy addresses the interactional and identity-related work that switching performs within conversation. The present study thus positions itself within the tradition of integrative bilingual research that recognises both the linguistic and the social dimensions of code-switching as analytically inseparable [10].
To ensure the reliability of the coding procedure, a subset of the corpus comprising twenty per cent of all recorded sessions was independently coded by a second researcher trained in bilingual transcription methodology. Inter-rater agreement was calculated using Cohen's kappa coefficient, yielding a value of κ = 0.84, which is conventionally regarded as indicating strong agreement [43]. Discrepancies between the two coders were resolved through discussion and, where necessary, by consulting a third analyst. The methodological parallel with the Welsh-English corpus study conducted by Deuchar, Donnelly, and Piercy [10] is instructive: in both cases, a corpus-based quantitative approach was employed to identify structural regularities in bilingual switching, and the frequency of switching at different grammatical boundaries was treated as the primary dependent variable. The present study extends this paradigm to the Polish-English bilingual context, in which the typological distance between the two languages — particularly with respect to inflectional morphology — introduces structural constraints on intrasentential switching that differ substantially from those operative in the Welsh-English community.
3.2. Structural Patterns of Code-Switching in the Collected Data
Quantitative analysis of the corpus revealed a total of 1,247 discrete code-switching instances across the forty-two recorded sessions. These instances were distributed across the three structural categories identified in Poplack's typology, though the distribution was markedly uneven, reflecting both the linguistic profiles of the participants and the grammatical constraints imposed by the structural properties of Polish and English. Table 1 presents the aggregate distribution of switching types across the full corpus, disaggregated by participant age group and dominant bilingual background.
| Switching Type | Total Instances | Percentage (%) | Heritage Speakers | Late Bilinguals | Simultaneous Bilinguals |
|---|---|---|---|---|---|
| Intersentential | 512 | 41.1 | 38.2% | 61.4% | 29.7% |
| Intrasentential | 589 | 47.2 | 54.1% | 29.3% | 62.4% |
| Tag-switching | 146 | 11.7 | 7.7% | 9.3% | 7.9% |
| Total | 1,247 | 100.0 | 100% | 100% | 100% |
The predominance of intrasentential switching (47.2%) over intersentential switching (41.1%) is a finding of considerable theoretical significance. Poplack's proficiency hypothesis predicts that intrasentential switching — which requires the speaker to satisfy the syntactic constraints of both languages simultaneously — is more characteristic of highly proficient bilinguals than of those with more limited command of one of the languages [49]. The present corpus broadly supports this prediction: among late bilinguals, who reported lower self-assessed proficiency in English, intersentential switching constituted over sixty per cent of all switching instances, whereas among simultaneous bilinguals, intrasentential switching was the dominant pattern, accounting for over sixty-two per cent of observed switches. The heritage speaker group occupied an intermediate position, with a slight predominance of intrasentential switching reflecting their high functional competence in English acquired through education. These distributional patterns are consistent with findings reported in Palestinian Israeli bilingual data, where intrasentential switching was similarly identified as a marker of higher bilingual proficiency [9].
Examination of the grammatical sites at which switching occurred within intrasentential instances revealed a strong preference for switching at nominal phrase boundaries. Across the corpus, 63.4% of all intrasentential switches involved the insertion of an English nominal element — either a single noun or a full nominal phrase — into an otherwise Polish matrix clause. This finding is consistent with the predictions of the Matrix Language Frame model, originally proposed by Myers-Scotton, which holds that in asymmetric bilingual communities, one language — here Polish — functions as the matrix language providing the morphosyntactic frame of the utterance, while the other — English — contributes embedded language islands [47]. The preference for nominal insertion over verbal insertion is particularly pronounced in this corpus, a pattern that also obtained in the Welsh-English data examined by Deuchar and colleagues [10], where nouns and proper names were identified as significantly more probable switching sites than verbs.
The relative infrequency of verbal switching in the Polish-English corpus warrants specific comment, as it illuminates a constraint that is structurally specific to this language pair. Polish is a richly inflected language in which verbs bear obligatory agreement marking for person, number, gender, and aspect. When an English verb is inserted into a Polish matrix clause, the speaker must either supply Polish inflectional morphology to the English stem — a process that yields hybrid forms such as downloadować (to download) or scrollować (to scroll) — or restructure the clause to avoid the morphological incompatibility. The corpus contains 87 instances of such morphological adaptation, in which English verbal stems are integrated into Polish verb paradigms through the addition of Polish infinitive or conjugational endings. This phenomenon, which lies at the intersection of code-switching and lexical borrowing, is absent from the Welsh-English data [10] due to the typologically different morphological profiles of Welsh and English, and its prevalence in the present corpus underscores the importance of language-specific structural analysis in code-switching research. Discourse markers and interjections constituted a further frequent switching site, with English items such as anyway, basically, whatever, and you know appearing in otherwise Polish conversational sequences on 213 occasions, a pattern that reflects their pragmatic rather than referential function in conversation.
3.3. Functional and Pragmatic Analysis of Code-Switching Instances
Whereas the preceding section addressed the structural distribution of code-switching instances within the corpus, the present section undertakes a qualitative analysis of the pragmatic and social functions that switching serves in the recorded conversations. Following Gumperz's influential distinction between situational code-switching — in which language choice is governed by objective changes in topic, interlocutor, or setting — and metaphorical code-switching — in which language choice carries connotative meaning within a stable interactional situation — the corpus was examined for instances in which switching appeared to be functionally motivated rather than structurally arbitrary [35]. Six primary functional categories were identified through iterative qualitative analysis of the transcribed data, as summarised in Table 2.
| Functional Category | Instances Identified | % of Total Switches | Predominant Switch Direction |
|---|---|---|---|
| Lexical gap filling | 298 | 23.9 | Polish → English |
| Identity assertion / in-group solidarity | 241 | 19.3 | Polish → English (tech/youth discourse) |
| Quotation / reported speech | 187 | 15.0 | Polish → English |
| Emphasis / reiteration | 163 | 13.1 | Bidirectional |
| Humour / irony / play | 142 | 11.4 | Polish → English |
| Hedging / mitigation / distancing | 216 | 17.3 | Polish → English (softening) |
The most numerically frequent functional category was lexical gap filling (23.9%), encompassing instances in which speakers introduced English lexical items into Polish discourse in order to refer to concepts, objects, or processes for which no established Polish equivalent was readily available or conventionally used within the speech community. This category was predominantly associated with professional, technological, and digital domains: terms such as deadline, feedback, workflow, streaming, and hashtag recurred across multiple participants and sessions without any attempt at Polish translation. The prevalence of this function reflects the well-documented asymmetry between English and Polish in the lexicalisation of contemporary professional and technological concepts, and corroborates findings from research on English as a prestige code in non-Anglophone bilingual communities [8]. It is noteworthy, however, that pure lexical gap filling — switching motivated exclusively by the absence of a Polish equivalent — was relatively rare in isolation; more commonly, the identification of a lexical gap co-occurred with pragmatic functions such as identity assertion or in-group signalling, suggesting that even ostensibly referential switches carry social meaning.
The second most prominent function, identity assertion and in-group solidarity marking (19.3%), encompassed instances in which the switch to English served to position the speaker within a particular social category — typically that of educated, cosmopolitan, or professionally successful young adults — or to signal shared membership in a bilingual peer community. Representative examples include the use of English formulaic expressions (obviously, literally, for real) as markers of shared epistemic stance, and the insertion of English evaluative adjectives (awkward, random, cringe) into Polish utterances as a means of conveying affective positioning unavailable through Polish alternatives. These patterns are consistent with the identity-functional account of code-switching proposed in research on Palestinian Israeli bilinguals, which found that pragmatic and identity functions frequently co-occur within individual switching instances rather than operating as discrete categories [9]. The metaphorical dimension of such switches, in Gumperz's sense, is evident: the insertion of English does not signal a genuine change in topic or addressee, but rather invokes a set of cultural associations — modernity, informality, bilingual competence — that confer particular social value within the peer group.
Quotation and reported speech accounted for 15.0% of identified switching instances, reflecting a widespread tendency among participants to reproduce the language in which an utterance was originally made when quoting another person's words. This pattern, often described in the literature as a quotative function, appeared across all participant groups and was particularly frequent when the quoted utterance had been delivered in an institutional or professional English-medium context — for example, a remark made by an English-speaking employer, a phrase encountered in English-language media, or an exchange conducted in a university seminar. The pragmatic effect of quotative switching is to authenticate the reported utterance by preserving its original linguistic form, thereby signalling veridical reproduction rather than paraphrase. The attitudinal functions of hedging, mitigation, and distancing (17.3%) were manifested in instances where speakers switched to English at moments of conversational difficulty — when expressing disagreement, declining an invitation, or broaching a sensitive topic — suggesting that the sociolinguistic distance associated with the non-native language can be deployed strategically to soften illocutionary force and manage interpersonal face concerns.
3.4. Social and Contextual Variables Influencing Code-Switching Behaviour
Beyond the structural and functional dimensions of code-switching, the present study examined the extent to which individual and contextual variables conditioned the frequency and complexity of switching behaviour across the corpus. Four primary variables were investigated: age, gender, degree of bilingualism, and conversational context. The analysis of these variables was conducted through a combination of descriptive statistics and qualitative examination of distributional patterns, with reference to comparable findings from previous research on bilingual communities [8] [9].
Age emerged as the most robust predictor of code-switching frequency within the sample. Participants in the younger age cohort (eighteen to twenty-five years of age) produced significantly more code-switching instances per minute of recorded speech than participants in the older cohort (thirty-six to forty-five years). Among the younger cohort, the mean switching rate was 8.3 instances per minute, compared with 4.1 instances per minute in the older cohort; participants aged twenty-six to thirty-five occupied an intermediate position, with a mean rate of 6.2 instances per minute. The qualitative character of switching also differed across age groups: younger speakers produced markedly more intrasentential switching, including complex examples in which English clausal constituents were embedded within Polish morphosyntactic frames, whereas older speakers showed a greater proportion of intersentential and tag-switching. These findings replicate the age-graded pattern documented in the Nakhl-e Taghi bilingual community study, where younger speakers were similarly found to code-switch more frequently than older interlocutors, with the difference attributed to differential socialisation into bilingual norms and greater exposure to global English-language media and digital culture [8]. The present data suggest that age functions not merely as a demographic variable but as a proxy for a cluster of sociolinguistic factors — including peer group norms, media consumption patterns, and educational trajectory — that collectively shape the frequency and structural complexity of code-switching.
Gender differences in switching behaviour were less pronounced than age effects, and no statistically significant difference in overall switching frequency was observed between male and female participants. However, qualitative examination of functional categories revealed a modest but consistent pattern: female participants produced proportionally more switching instances coded as identity assertion and solidarity marking, while male participants showed a slightly higher proportion of switches in the lexical gap filling and quotative categories. These differences, while subtle, are consonant with findings from the literature on gendered communicative styles in bilingual communities, which suggest that women tend to orient more strongly toward interactional and relational dimensions of conversation, while men's language choices more frequently serve referential or informational functions [30]. The attenuated gender effect observed in the present corpus also aligns with findings from the Nakhl-e Taghi study, where gender differences in code-switching were found to be context-dependent rather than consistent across all communicative situations [8].
The degree of bilingualism — operationalised through the participant classification described in Section 3.1 — was the second most powerful predictor of switching behaviour after age, confirming the centrality of bilingual proficiency to the structural character of code-switching. Heritage speakers produced the most structurally complex switching patterns, including the highest rates of intrasentential switching and the greatest variety of English-language elements inserted into Polish matrix clauses. Simultaneous bilinguals displayed comparable complexity in intrasentential switching but showed a slightly different functional distribution, with greater proportions of metaphorical switching and identity-related functions. Late bilinguals, consistent with Poplack's proficiency hypothesis, relied most heavily on intersentential and tag-switching — forms that do not require simultaneous satisfaction of the morphosyntactic constraints of both languages. The comparison with Palestinian Israeli data is instructive in this regard: Amani's research similarly found that more educated and linguistically proficient bilinguals produced more frequent and structurally complex code-switching, interpreting this pattern as an index of linguistic modernisation and cosmopolitan identity [9]. The present data do not directly address the modernisation hypothesis, but they support the underlying empirical claim that proficiency and switching complexity are positively correlated.
Conversational context was found to condition switching frequency in predictable ways. Sessions recorded in peer dyads — in which both participants were close friends or flatmates of similar age — produced substantially higher switching rates than sessions involving task-oriented interaction, such as collaborative problem-solving or information exchange between less familiar interlocutors. The mean switching rate in peer conversation sessions was 7.6 instances per minute, compared with 4.4 instances per minute in task-oriented dyads. This contrast reflects the well-established finding that informal, solidarity-oriented discourse provides a more permissive interactional environment for code-switching, reducing the social risk associated with departures from monolingual norms and enabling participants to exploit the full range of their bilingual repertoire [25]. The effect of conversational context interacted with degree of bilingualism: among heritage and simultaneous bilinguals, the difference in switching rates between peer and task-oriented contexts was particularly pronounced, whereas late bilinguals showed less contextual variability, suggesting that for less proficient bilinguals, the structural demands of intrasentential switching constrain production independently of social context.
3.5. Discussion: Findings in Relation to Existing Theory and Prior Research
The empirical findings presented in Sections 3.2 through 3.4 are now situated within the broader theoretical landscape surveyed in Chapters 1 and 2, and their implications for the understanding of code-switching in the Polish-English bilingual community are systematically assessed. Four theoretical frameworks are addressed in turn: the Matrix Language Frame model, Poplack's proficiency hypothesis, Gumperz's we-code/they-code distinction, and the question of functional multivalency in code-switching.
The data provide robust support for the central claim of Myers-Scotton's Matrix Language Frame model, namely that in asymmetric bilingual contact situations, one language functions as the grammatical scaffolding of the mixed utterance while the other contributes lexical and phrasal material. Polish was overwhelmingly identified as the matrix language in the present corpus: across all intrasentential switching instances, Polish morphosyntactic structure — including case government, verbal agreement, and constituent order — determined the grammatical architecture of mixed utterances, while English material was inserted as embedded language islands within that frame [47]. The MLF model's predictions were particularly well borne out in the domain of verbal morphology: where English verbal stems appeared in Polish matrix clauses, they invariably received Polish inflectional endings, confirming that the morphological requirements of the matrix language take precedence over the grammatical properties of the embedded language. This pattern has been documented in a range of other bilingual contact situations, and its replication in the Polish-English context strengthens the cross-linguistic generalisability of Myers-Scotton's framework. The parallel with the Welsh-English corpus study is evident: Deuchar and colleagues similarly confirmed Welsh as the matrix language in their data, with English functioning as the source of nominal insertions within Welsh morphosyntactic frames [10].
Poplack's proficiency hypothesis is likewise supported by the distributional evidence from the present study. The positive correlation between degree of bilingualism and proportion of intrasentential switching — observable across the heritage speaker, simultaneous bilingual, and late bilingual subgroups — replicates the core empirical prediction that complex intrasentential switching serves as an index of balanced bilingual competence. However, the data also reveal a complication not fully accounted for by Poplack's original formulation: the structural character of intrasentential switching in this corpus is constrained not only by the proficiency of the individual speaker but also by language-specific morphological factors that create systematic asymmetries between switching sites. The density of Polish inflectional morphology, in particular, creates substantial structural barriers to verbal insertion that are absent in the typologically simpler verbal systems of Spanish or Welsh. Future refinements of the proficiency hypothesis may need to incorporate a parameter sensitive to the morphological complexity of the matrix language, as this variable appears to modulate the relationship between proficiency and switching type in the Polish-English case.
Gumperz's distinction between we-code and they-code, in which the minority or heritage language functions as the we-code of in-group solidarity and the dominant language as the they-code of institutional distance, finds qualified support in the present data. Polish, as the heritage language of the community, functioned consistently as the unmarked code of intimate peer interaction — the default linguistic medium of informal conversation — while English was introduced to signal reference to external, institutional, or cosmopolitan domains. The pattern of English insertion for professional terminology, quotation of institutional speech, and ironic appropriation of dominant-culture discourse is consistent with Gumperz's analysis of bilingual communities in which the minority language is used to construct a bounded social identity [35]. However, an important qualification must be noted: in the present community, the status of English as purely a they-code is complicated by its simultaneous function as a marker of desirable social identities — cosmopolitanism, professional success, educational attainment — within the peer group itself. English thus occupies a paradoxical position as both a they-code in its institutional dimension and a quasi-we-code in its indexical association with a particular kind of bilingual identity. This ambiguity resonates with findings from the Palestinian Israeli context, where code-switching served simultaneously as a marker of modernisation and as a means of in-group solidarity within the bilingual community [9].
Perhaps the most theoretically significant finding of the present study is the challenge it poses to categorical distinctions between functional types of code-switching. The qualitative analysis reported in Section 3.3 revealed that the majority of switching instances could not be assigned unambiguously to a single functional category: most switches served multiple simultaneous functions, with lexical gap filling co-occurring with identity assertion, quotation overlapping with solidarity marking, and humour intersecting with hedging. This finding of functional multivalency is not unique to the present corpus; it has been documented in comparable research on bilingual communities and has led some researchers to question the analytic utility of discrete functional taxonomies [22]. The implication for the application of Gumperz's taxonomy is that the six functional categories it identifies should be understood as dimensions of meaning along which individual instances may be differentially positioned, rather than as mutually exclusive types into which instances are sorted. The co-occurrence of pragmatic and identity functions observed in the present data mirrors the pattern reported for Palestinian Israeli bilinguals, where Amani similarly concluded that the pragmatic and social dimensions of code-switching cannot be cleanly separated [9].
The present study is subject to several limitations that must be acknowledged in assessing the generalisability of its findings. The sample of twenty participants, while sufficient for qualitative analysis and preliminary quantitative description, is modest by the standards of large-scale corpus sociolinguistics; the quantitative findings should therefore be treated as indicative rather than definitive. A second limitation concerns the recording effect: despite participants' familiarity with the recording procedure and the investigator's deliberate adoption of minimal observer presence, the awareness of being recorded cannot be entirely excluded as a source of modification in participants' language behaviour. The observer's paradox, as identified by Labov, represents an irreducible methodological challenge in naturalistic sociolinguistic research [41]. A third limitation is the relatively restricted geographic scope of the sample, which was drawn predominantly from a single metropolitan area; the code-switching patterns of Polish-English bilinguals in rural or less cosmopolitan settings may differ substantially from those documented here. Future research should seek to address these limitations through the assembly of larger, geographically diverse corpora and through the complementary use of ethnographic observation and self-report data.
Despite these limitations, the present study makes a substantive contribution to the empirical literature on Polish-English code-switching by providing the first systematic corpus-based analysis of structural and functional switching patterns in naturally occurring informal conversations. The findings extend the theoretical frameworks developed for other bilingual contact situations — including the Welsh-English community studied by Deuchar and colleagues [10] and the Palestinian Israeli community examined by Amani [9] — to the Polish-English context, while simultaneously identifying structural specificities of this language pair that require theoretical accommodation. The documentation of morphological adaptation strategies in verbal switching, the identification of functional multivalency as a pervasive characteristic of code-switching in this community, and the confirmation of age and bilingual proficiency as robust conditioning variables all represent contributions to empirical knowledge that provide a foundation for the comparative and theoretical discussions presented in the concluding chapter of this thesis. Future research building on this study might profitably employ larger corpora, longitudinal designs, or experimental paradigms to disentangle the relative contributions of structural constraint, individual proficiency, and social context to the production of code-switching in Polish-English bilingual discourse [8].
Conclusion
The present thesis has examined code-switching among Polish-English bilingual speakers in informal conversational settings, bringing together theoretical frameworks from contact linguistics, sociolinguistics, and pragmatics in order to illuminate both the structural properties and the social functions of this pervasive phenomenon. The investigation proceeded across three interconnected dimensions: a critical survey of the terminological and theoretical landscape that informs current scholarship on bilingual language alternation; an analysis of the sociolinguistic and pragmatic mechanisms through which code-switching operates as a communicative resource; and an empirical examination of naturally occurring speech data drawn from twenty bilingual adults whose conversational practices were subjected to systematic qualitative and quantitative analysis. Taken together, the findings reported in these chapters support a view of code-switching not as a symptom of linguistic inadequacy or interference, but as a sophisticated communicative strategy that is simultaneously shaped by grammatical constraints, social identities, and moment-to-moment interactional demands. The conclusions drawn from this research contribute to a growing body of empirical evidence demonstrating that bilingual language alternation constitutes a coherent and rule-governed mode of communication deserving rigorous scholarly attention.
The theoretical foundations examined in Chapter One established that code-switching is a multidimensional phenomenon whose proper analysis requires careful attention to terminological precision. The distinction between code-switching, borrowing, code-mixing, and translanguaging was shown to be not merely a matter of academic taxonomy but a substantive issue with consequences for the interpretation of empirical data. The typologies advanced by Poplack [1], Myers-Scotton [2], and Muysken [3] were assessed as complementary rather than mutually exclusive analytical instruments, each foregrounding different aspects of the structural embedding of switched material within discourse. The Matrix Language Frame model, in particular, proved indispensable for understanding how Polish functions as the dominant grammatical scaffold in the speech data analysed in Chapter Three, providing the morphosyntactic frame into which English lexical material is inserted. Equally, Auer's interactional approach [4] offered a corrective to overly deterministic structural accounts by insisting upon the sequential and locally negotiated character of switching decisions, a perspective that proved highly compatible with the functional patterns identified in the empirical component of the thesis.
The sociolinguistic and pragmatic dimensions explored in Chapter Two illuminated the social work that code-switching performs within the communities and interactions under consideration. Drawing on Bourdieu's [5] conception of linguistic capital and Le Page and Tabouret-Keller's [6] acts of identity framework, it was argued that language alternation serves as a powerful mechanism through which speakers construct, negotiate, and project social identities. The accommodation theory developed by Giles [7] provided further explanatory leverage, demonstrating how convergent and divergent switching patterns reflect speakers' orientations toward interlocutors and social groups. Pragmatically, the functions identified in the literature — including emphasis, mitigation, humour, and the management of conversational structure — were shown to operate in ways that are consistent with Jakobson's [8] expressive function of language and with the politeness strategies theorised by Brown and Levinson [9]. The attitudinal dimension of the analysis, informed by Woolard's [10] work on language ideology and by the prescriptive context established by the Polish Language Act, revealed a productive tension between official discourses that stigmatise code-switching and the informal permissiveness characteristic of bilingual peer interaction, a tension that is reflected in the variable degrees of linguistic insecurity [11] reported by participants when reflecting upon their own switching behaviour.
The empirical analysis presented in Chapter Three confirmed, qualified, and in certain respects complicated the theoretical positions surveyed in the preceding chapters. The structural findings — that intrasentential switching constituted the most frequent pattern at sixty-two percent of all switching instances, followed by intersentential switching at twenty-eight percent and tag-switching at ten percent — are broadly consistent with Poplack's original typology and with subsequent empirical work on Spanish-English bilingual communities [1]. The predominance of intrasentential switching, which demands the highest degree of bilingual competence insofar as it requires speakers to navigate the grammatical systems of both languages simultaneously, constitutes direct evidence against the deficit hypothesis, which would predict that switching occurs primarily when speakers lack the lexical or grammatical resources to proceed in one language alone. Instead, the present data support Poplack's proficiency hypothesis [1], confirming that it is the most proficient bilinguals who switch most frequently and most structurally adventurously. The Matrix Language Frame model [2] was further corroborated by the consistent identification of Polish as the matrix language across the corpus, with English functioning predominantly as the embedded language from which lexical items are drawn and inserted into Polish morphosyntactic frames.
The functional analysis yielded findings of particular theoretical interest. The identification of identity assertion as the most frequently motivating function, accounting for thirty-two percent of switching instances, aligns closely with the acts of identity framework and with the broader sociolinguistic literature emphasising the role of language choice in the indexical construction of social personhood. However, the distribution of functions observed in the data nuances single-function taxonomies that treat each switching event as serving one discrete purpose. In practice, many instances of switching were found to serve multiple functions simultaneously — a single switched utterance might mark in-group solidarity, achieve pragmatic economy in denominating a culturally specific concept, and reinforce the speaker's identity as a cosmopolitan, education-marked individual. This finding resonates with the position advanced by Auer [4] and others that switching is polyfunctional and that analytical frameworks which seek to assign each instance to a mutually exclusive category risk misrepresenting the communicative complexity of bilingual interaction. The social variable analysis further revealed that the degree of bilingualism constitutes the strongest predictor of switching frequency, with younger speakers and those with extended residence or education in English-speaking contexts exhibiting the highest rates of alternation. Gender, by contrast, exerted a minimal and statistically unreliable effect, a finding that is consistent with some empirical studies but diverges from accounts in which women are characterised as more conservative in their language choices [12].
The contribution of the present thesis to the field of contact linguistics and bilingualism studies is located at several levels. At the empirical level, the study provides data drawn from a community — Polish-English bilinguals in informal interaction — that has received considerably less systematic attention than Spanish-English or French-English bilingual populations, which dominate the anglophone literature. The naturalistic, audio-recorded methodology employed here, which prioritised ecological validity over experimental control, generated a corpus that reflects the genuine complexity of spontaneous conversational switching rather than the constrained switching elicited by translation tasks or grammaticality judgement studies. At the theoretical level, the thesis demonstrates the continued relevance of the classical frameworks developed by Poplack, Myers-Scotton, and Gumperz while also highlighting the value of integrating interactional and pragmatic perspectives that attend more closely to the local, sequential organisation of talk. The finding that no single theoretical model accounts fully for the observed data is itself a contribution, insofar as it underscores the need for pluralistic analytical approaches that draw selectively upon structural, sociolinguistic, and pragmatic frameworks according to the specific phenomena under investigation.
Several directions for future research emerge from the limitations and findings of the present study. The sample of twenty participants, while appropriate for an exploratory empirical investigation of this scope, is insufficient to support strong generalisations about Polish-English bilingual communities as a whole. Future research would benefit from larger and more diversified samples, stratified systematically by age, educational background, migration history, and domain of English language acquisition. The present corpus, drawn exclusively from informal adult peer conversation, does not permit conclusions about how switching patterns vary across different social domains and participant configurations; studies examining code-switching in professional, familial, or institutional settings would substantially enrich understanding of contextual variation. The attitudinal data gathered informally through participant observation and post-recording reflection suggest that the ideological dimensions of code-switching merit more sustained investigation, particularly in relation to the specific sociolinguistic ecology of Polish, which is embedded in a normative discourse that remains substantially more prescriptivist than that surrounding many other European languages. Longitudinal studies tracking the same speakers across extended periods of time would also contribute valuable evidence regarding the stability or transformation of individual switching repertoires as proficiency, social networks, and linguistic attitudes evolve.
The question of digital and computer-mediated communication represents a further avenue of considerable promise. Preliminary observations suggest that Polish-English bilingual speakers switch extensively in text messages, social media posts, and online forums, and that the structural and functional properties of written switching may diverge in interesting ways from those of spoken interaction. The reduced acoustic and prosodic cues available in written communication, combined with the asynchronous and often public character of digital exchanges, may alter both the frequency and the social meaning of switching in ways that merit dedicated investigation. Similarly, the relationship between code-switching and translanguaging — a concept that has gained significant traction in applied linguistics and language education research — deserves more careful empirical attention in the Polish-English context, particularly with respect to the pedagogical implications of bilingual language practices in educational settings where Polish is the medium of instruction and English a prestige second language.
It is appropriate, in closing, to reflect upon the broader significance of code-switching as an object of scholarly inquiry. Research in this area matters not only for the light it casts on the formal properties of grammar at the interface between languages, but for what it reveals about the fundamentally social and identity-laden character of human language use. Bilingual speakers who alternate between languages in conversation are not, as prescriptive discourse has sometimes implied, failing to master either language; they are, rather, deploying their full linguistic repertoires in the service of communication, identity, and social belonging. The Polish-English bilinguals whose speech was analysed in the present study demonstrated consistently that code-switching is a skilled and purposeful practice, one that draws upon grammatical competence in two languages simultaneously while also navigating complex social landscapes of affiliation, authority, and solidarity. At a time when multilingualism is an increasingly salient feature of urban, educational, and professional life across Europe and beyond, research that documents and theorises the communicative practices of bilingual speakers contributes to a more accurate and more respectful understanding of linguistic diversity. The present thesis, situated within a well-established scholarly tradition while attending to the specific characteristics of Polish-English bilingualism, represents one step in the ongoing effort to account for the richness and systematicity of contact-induced language variation.