Lingua franca is probably one of the most widely circulating linguistic terms. Yet oddly, it lacks a commonly accepted, precise academic definition. It generally denotes a shared language used by people to communicate who speak different languages. In this short essay I wish to expand on that popular understanding and offer a theoretical foundation for it. I draw on my lecture ‘Towards a functional definition of lingua franca’ delivered at the Linguistics Colloquium at the University of Hamburg in April 2026 (Matras 2026).

The label lingua franca – ‘the Frankish Language’ – derives, apparently, from that given to a form of Latin disseminated by Christian crusaders, which later spread across the Mediterranean regions and was used exclusively in inter-ethnic encounters. Today’s use of the term does not, however, capture a particular language. Instead, it denotes a function that can be carried by a language alongside its other everyday functions. We therefore require a definition of lingua franca that addresses its distinctive communicative functionality rather than its ancestry, cultural roots or structural composition.

Theoretical anchoring

My point of departure is that a definition must be anchored within a general theoretical approach to the conditions, constellations and purpose of communicative interaction; and simultaneously within a dedicated, explanatory theoretical framework that accounts for communication in multilingual settings. For the first, I take inspiration from the action-based theory of language represented by the works of the Functional Pragmatics school in German linguistics (Rehbein 1977, Ehlich & Rehbein 1986). For the second, I rely on the functional approach to contact linguistics outlined in my work on the multilingual repertoire management model (Matras 2009/2020, forth.). I will briefly describe the basic tenets of both frameworks before explaining how they can help formulate a core definition of lingua franca.

An action-oriented theory of language views communicative interaction as a way of achieving goals by aligning knowledge between a speaker and a listener. Communicative interaction is understood as set within an action space, which determines the conditions under which transfer of knowledge is enabled. In the centre is the action constellation, which brings together actors for a particular purpose. Purpose is viewed as the driving motivation to change states-of-affairs, mediated through communicative interaction (Rehbein 2024). Interaction is shaped by established routines, which are specific to societies and cultures. It is often embedded into institutional settings, broadly understood as action constellations where actors have pre-defined roles from which they derive specialised knowledge and authority. These can be professional roles, as in the interaction between a vendor and a customer or that between a teacher and a pupil, or more casual roles, as in the communication between a parent and a child. In some cases (like court proceedings) institutional communication is tightly scripted; in others it tends to follow established sequencing patterns. Language structures are considered in this approach as procedures that activate distinctive fields of mental representations. Lexicon, for example, activates the field of symbolism of real-life entities while grammatical inflection is an operational procedure that processes links between propositional elements within the verbalised context of speech. I return to this below when discussing pidgins and creoles.

The multilingual repertoire management model (Matras 2009/2020) assumes that language resources are processed by users as an integrated whole – the repertoire of language resources – rather than as separate systems. Multilingual users constantly make active choices among counterpart, complementary or competing structural forms. The model recognises three pull-factors that users need to balance when managing their choices. The first is the need to accommodate to the expectations of the interaction setting. The second is the urge to exhaust the full expressive potential of the repertoire as a whole. That means being able to express the same nuance or identify the same concept or object in whatever setting one is interacting in at a given moment. The third pull-factor is the drive to reduce processing load. Making choices requires users to monitor the interaction and activate a selection-and-inhibition mechanism as part of their language processing and production apparatus. Functional categories that create potential tension between speaker and listener, such as expressions of contrast or those that have gesture-like features like exclamations and intonation, can put a stain on that mechanism. Balancing pull-factors leads to compromises of various kinds. Two principal balancing mechanisms are the replication of linguistic matter – the generalisation of a word-form across interaction settings and thus across sub-sets of the repertoire – and that of linguistic pattern – assigning a meaning that is associated with a form in one sub-set to a counterpart form in another, effectively generalising form-meaning relationship and semantic-pragmatic distribution across the repertoire. In structural linguistics, these phenomena are usually conceived of as interference, transfer, borrowing, or convergence. In our model they are viewed as a change in behaviour routines through which the notional demarcation boundaries that guide the deployment of sub-sets of resources within the repertoire are altered in various ways.

Core components

How can we define lingua franca drawing on these tenets?

First, lingua franca is a response to the emergence of new action constellations that bring together actors of different backgrounds around specific encounters. These new action constellations arise through a change of historical social and economic circumstances often involving some kind of mobility or displacement, and in some cases the emergence of new institutional routines. Distinctive of lingua franca is the targeted deployment of language resources within the repertoire to serve the communicative purposes around which actors come together.

Next, lingua franca is distinctively the process of drawing on shared language resources in an action constellation where the actors’ language repertoires are non-congruent or asymmetrical. The notion of non-congruence or asymmetry refers here not to competence or sequence of acquisition (i.e. whether a language is ‘native’, that is, acquired from birth, or not) but to the degree to which some partial overlap exists between language resources in the respective repertoires of participating actors. When actors’ repertoires are non-congruent, they must, in order to accommodate to the new action constellation that brings them together, select those language resources that are accessible to all participants in the interaction.

While the function of lingua franca is inherently linked to multilingual settings, not every communication in a multilingual setting constitutes a lingua franca.

Consider a group of Turkish-Kurdish bilinguals who alternate between their two languages in casual conversation in the home or beyond. They might choose Turkish for a conversation about a particular topic in order to connect to wider societal discourse, or Kurdish to demonstrate their shared ethnic identity. Neither choice is a lingua franca. That is because neither is the single shared resource on which actors can rely in order to enable communication.

By contrast, an encounter between persons of Armenian and Uzbek backgrounds in the Caucasus or Central Asia is likely to take place in Russian because that constitutes the only set of language resources that the two have in common. By that definition, it doesn’t matter if the shared language is a so-called L1 or L2. If a Russian person were to join that same encounter, and their use of Russian might be their default choice, for the encounter as a whole Russian would still be a lingua franca: It is the only set of language resources that is available to serve in that particular action constellation in which participants’ repertoires are non-congruent.

I now turn to three examples that show that lingua franca can appear under different socio-historical circumstances and have various social, communicative and structural manifestations.

Pidgins and creoles

Pidgins and creoles are often considered to be the quintessential lingua franca. Whether in trade encounters or as a result of colonial enslavement and subsequent emancipation and urbanisation, pidgins and creoles emerge in settings where individuals of different backgrounds come together around a particular set of transactions. These are new, purpose-bound action constellations that have not been part of previous action routines. They require a new form of communication. Given their different backgrounds, participants’ language repertoires are non-congruent. But they gain access to a pool of shared language resources: In plantation creoles it is usually the basic lexicon of the colonial language combined with some exposure to structural features drawn from the various heritage languages. In trade settings, repeated encounters allow actors to familiarise themselves over time with portions of each other’s repertoires.

Viewed from the combined perspective of action theory and the repertoire management model, pidgins serve an emerging action routine as users select and deploy language resources that are likely to be familiar to all participants. This approach offers a function-oriented alternative to the controversial view of pidgins as simplification or broken transmission: Pidgins show core lexicon but not bound morphology because in casual encounters, lexicon (as a procedure that activates the symbolism field of language) is more easily accessible and replicable than inflection (an operation that relies on processing connected propositional content and requires prolonged and more intense exposure in order to acquire). The repertoire management model allows us to dislodge the view of pidgins from the traditional historical linguistic paradigm that focuses on language ancestry (as the perceived continuation or discontinuation of structural forms) and instead to shift our attention to the functionality of choices within individual actors’ repertoires of language resources that draw on the local linguistic environment or ‘ecology’ (cf. Mufwene 2001).

The structural profile of pidgins and creoles is thus attributed to the need to create initially a makeshift mode of communication, which is then conventionalised and expanded. Tok Pisin is a lingua franca in Papua New Guinea but it is also the all-purpose language of entire villages and neighbourhoods. Formations that began as improvisation drawing on available resources, like bin ‘been’ to express past tense, the emulation of the object pronoun ‑im ‘him’ to express transitivity or the metonomy wanpla ‘one fellow’ to denote singularity, have become a part of canonical grammar in constructions like i bin raitim wanpla stori ‘he wrote a story’. That is part of a historical process of the social appropriation of the language, and in this case its standardisation and adoption as a national language. As we will see below, appropriation is a characteristic feature of lingua franca, one that accompanies and shapes the pathway from one-off encounters to established routines.

English

English is widely regarded as today’s global lingua franca, but it did not emerge specifically to serve multiethnic encounters and it does not display the feature profile that is said to be typical of pidgins and creoles. Nevertheless, its role as lingua franca in specific action constellations among actors with non-congruent repertoires, especially around institutional discourse, does on occasion give rise to some structural manifestations. I will give two real-life examples.

In the first, a central European academic enquired with a university in one of the Baltic countries about the possibilities of carrying out contractual work as an external paid collaborator (not a member of staff). The university responded by sending a draft contract template. In the text, the prospective contractor was referred to as ‘the Employee’. The academic objected, claiming that the term Employee was reserved for staff (with implications for tax, pension, residence requirements and more). The university, however, was adamant that the term simply meant that the person would deliver the work and be paid for it. The discrepancy arose because one of the actors in this purpose-specific constellation, the university, appropriated the shared language resource, insisting that a particular interpretation was universally attributable to an English term.

In another example, a project manager at a UK university was corresponding with an administrator of French background working at the European Commission’s Research Directorate in Brussels. The EC administrator asked the UK-based manager what her project’s actual financial balance was. The UK-based manager interpreted that as an insinuation that the figures that she had previously provided were deemed inaccurate or incorrect. In fact, the Brussels-based administrator was using the English term as it is commonly used in the corridors of the EC Directorate, mapping onto it the meaning associated with its counterpart cognate words in French or German, which mean ‘at this given point in time’. It is the process to which I have referred as ‘pattern replication’. The EC-administrator’s confidence in her usage and and the fact that she rejected its questioning by the monolingual, native English speaker revealed a sense of appropriation of English within the action constellation of her EC institutional setting.

Israeli Hebrew

Israeli Hebrew is a revitalised language introduced through a top-down nation-building process and shaped over several generations by a founder population with diverse background languages. Compared to pidgins and creoles and English, it is a kind of hybrid: It possesses a target form with a strong normative, literary tradition, and yet it was adopted rapidly without input from native speakers (speakers of the language from birth). That resulted in an accelerated pace of internal structural change as well as the incorporation of items from a feature pool derived from the many background languages.

Analogical change across morpho-phonological paradigms is observed regularly in language but in Israeli Hebrew it is especially common. For example, formal Hebrew makir ‘knows’ is usually pronounced mekir by analogy to verbs like mexin ‘prepares’, mevin ‘understands’. Formal Hebrew kibasti ‘I did the laundry’ is pronounced xibasti by analogy to verbs like xipasti ‘I searched’ or xišavti ‘I calculated’. The rapid spread of change encompassing the entire first-generation of native speakers can be attributed to the absence of clear guidance imposed by a parental generation, who used Hebrew as lingua franca. It is also facilitated by the distinctive features of the Hebrew writing system, where contrasts between the vowels a/e and the consonant pair k/x are marked by diacritic symbols that are normally omitted from both handwriting and print (and digital writing). That results in permanent ambiguity, but it also allows users to reconcile formal written language with colloquial pronunciation without the risk of blatant non-compliance.

From core to scope

Appropriation and structural modification are typical properties that accompany the use of a lingua franca. Their precise direction is shaped by the historical and sociolinguistic circumstances. As we have seen, these can differ widely. We must, therefore, be open to consider a broad scope of historical and sociolinguistic phenomena if we are to adopt a consistent core definition of lingua franca as a communicative function. Cultural historians tend to associate lingua franca with the languages of past empires – Persian, Latin, Arabic, and Ottoman Turkish – or with regional trade languages like Swahili, Malay, Low German, and Urdu. All these are associated with historical circumstances that enabled a class of actors with socio-economic and political power to introduce or impose their languages on diverse populations. And yet these are accompanying factors, not the distinctive features that define lingua franca as a functional instrument of communication.

Imagine a nineteenth-century German PhD student who was expected to submit their thesis to their supervisor in Latin. Does their use of Latin constitute a lingua franca? That depends. If the thesis is to be read only by the student, the supervisor, and perhaps their immediate peers and colleagues, then the choice of Latin does not aim to enable interaction by selecting shared resources among actors with non-congruent repertoires. Rather, it is aesthetic and performative, symbolising belonging to an educated social class. If, however, the choice of Latin is to enable distribution of the thesis to a wider audience of readers with non-congruent language repertoires, then a case can be made for its use as lingua franca in this instance. Our decision whether to consider a choice as lingua franca or not does not rely exclusively on which language is chosen (‘Latin is a lingua franca’ is not the cue) or on that language’s historical sociolinguistic profile but on the function of the choice with regard to the management of language repertoire resources in the particular action constellation.

Nor is the choice of a language that is not the native language of the actors involved in the interaction constitutive as such of lingua franca. When a French schoolchild speaks English to their teacher during an English class at school, then the choice of English is the object of the action constellation itself, not the single most effective means to enable interaction by bridging non-congruent language repertoires.

By contrast, consider a grocery shop in a suburb of Hamburg that is owned and operated by people of Turkish background. They have customers of Arab and Afghani background and they even translate some of their signage into Arabic and Persian in order to accommodate them. Yet the spoken language used between the shopkeepers and their Arab and Afghan customers is German. It is a form of German that differs from the general norm, characterised by gaps in lexicon and grammatical constructions. Quite likely, some of the participants show a form of German that has fossilised. Perhaps that fossilisation is reciprocal, giving rise to a kind of migrant variant of German. A notable feature is the generalisation of the adjective ending ‑e for all cases, genders and numbers, as in willst du eine große Apfel? ‘do you want a large apple?’. There is also a carry-over from the cross-linguistic conventions of Muslim communities, as interlocutors are addressed in German as ‘brother’ and ‘sister’ respectively. There is thus some degree of appropriation, typical of the social conventionalisation of a language resource as lingua franca. But irrespective of any conventionalisation, the grocery example shows an established routine where in a particular action constellation participants bridge non-congruent language repertoires by drawing on shared language resources. That meets our functional definition of lingua franca.

Life is not simple, and there are borderline cases. In England I observed encounters between a Kurmanji-Kurdish speaking student from Syria and a Sorani-Kurdish speaking student from Iraq who always conversed in English. They found speaking their respective varieties of Kurdish too cumbersome to allow smooth and convenient communication. Both know Arabic from childhood but avoided the language as they associated it with cultural repression. In their subjective perception, English was the only shared language resource that could effectively serve their interaction. Arguably, some aspects of our definition are met in this case, but only if we include the emotive aspects that actors take into consideration even when they do have a wider choice. In that regard, this example deviates from the prototype.

In defining lingua franca, our interest is in the social phenomenon of an established communicative instrument for regular action routines that involve a reasonably significant population of actors. While it can sometimes be fruitful to brainstorm around single individual encounters, our definition seeks to generalise about wider forms of social practice.

Conclusion

Pennycook (2012) has suggested to view lingua franca as an emergent collection of local language practices built from bottom-up in a multilingual environment. I agree. But I believe we can go a step further and describe the mechanism by which a lingua franca emerges within a more general framework of linguistic action and multilingual communication.

Multilingual encounters can give rise to action constellations that bring together participants with non-congruent (asymmetrical) repertoires of language resources. These action constellations are usually construed around particular communicative purposes, which are often transactional.

To manage interaction in such constellations, actors avail themselves of a subset of their respective linguistic resources that they share. Its conventionalisation in an action routine can lead to an appropriation (an emerging sense of ownership) of that subset. That in turn can license various processes of structural change, either contact-induced (following the usual mechanisms of matter- or pattern replication) or internal (such as inter-paradigm analogy and grammaticalisation).

The term lingua franca thus captures the reliance on a shared subset of the repertoire in a particular action constellation and the potential processes of change that can accompany it.

References

Ehlich, K., & Rehbein, J. 1986. Muster und Institution: Untersuchungen zur schulischen Kommunikation. Tübingen: Gunter Narr.

Matras, Y. 2009 [second edition 2020]. Language Contact. Cambridge: Cambridge University Press.

Matras, Y. 2026. Towards a functional definition of lingua franca. Public guest lecture. Zentrum für Sprachwissenschaft, Universität Hamburg. 8 April 2026.

Matras, Y. forth. From category essentialism to process dynamism. Reconsidering theory in contact linguistics. In: Drinka, Bridget, Terttu Nevalainen & Gijsbert Rujen. eds. Handbook of Historical Sociolinguistics. Berlin: De Gruyter. [pre-print]

Mufwene, S. 2001. The ecology of language evolution. Cambridge: Cambridge University Press.

Pennycook, A. 2012. Lingua Francas as Language Ideologies. In: English as an International Language in Asia: Implications for Language Education, edited by Andy Kirkpatrick and Roland Sussex. Dordrecht: Springer. 137–154.

Rehbein, J. 1977. Komplexes Handeln: Elemente zur Handlungstheorie der Sprache. Stuttgart: Metzler.

Rehbein, J. 2024. ‘Zweck’ als Interpretament funktional-pragmatischer Analyse ‘authentischer’ Daten – eine Annäherung. In: Sprache – Denken – Wirklichkeit: 14. Internationale Konferenz zur Funktionalen Pragmatik, 23. bis 25. September 2021, Chemnitz, edited by Matthias Meiler and Winfried Thielmann. Chemnitz: Universitätsverlag Chemnitz. 13–76.