Competence and Performance in Language Learning

Competence and performance are the two concepts that define what it means to "know" a language — and the tension between them runs through every serious argument about how languages should be taught and tested. Competence is the internal system a speaker carries in their head: the implicit knowledge of a language that lets them produce and understand sentences they have never met before. Performance is what actually comes out of the mouth or onto the page in real time — complete with hesitations, false starts, slips and the effects of tiredness. Noam Chomsky drew the line between them in 1965, and within a decade other linguists were arguing that the line was drawn in the wrong place. That argument is the reason modern teaching cares as much about using a language as about knowing its rules.

Chomsky's distinction: competence versus performance

Chomsky introduced the pair in Aspects of the Theory of Syntax (1965), in one of the most quoted passages in modern linguistics: "Linguistic theory is concerned primarily with an ideal speaker-listener, in a completely homogeneous speech-community, who knows its language perfectly." That idealisation was deliberate. Chomsky wanted to study the underlying competence — the system of rules that a speaker has internalised — stripped of everything that gets in its way when the system is put to use.

Competence, then, is knowledge: the finite, internalised grammar that allows a speaker to generate and judge an infinite number of new sentences. It is what explains the central fact of language for Chomsky — that speakers routinely produce and understand utterances they have never previously encountered, a property he called the creative aspect of language use. No stock of memorised sentences could account for this; only a generative system could. Performance is the opposite pole: "the actual use of language in concrete situations." Performance is where competence meets the real world of limited memory, divided attention, nerves and noise — and it never reflects competence perfectly. A fluent native speaker who stumbles over a long, deeply nested sentence has not lost their competence; their performance has simply hit a processing limit.

The distinction has an older ancestor in Ferdinand de Saussure's langue (the shared language system) and parole (individual speech acts), but Chomsky's version is sharper and psychological rather than social: competence sits in the mind of the individual speaker, not in the community. In his later work he renamed the pair to make this explicit — competence became I-language (internal, individual, intensional) and the externalised side became E-language — but the core intuition never changed. What a linguist wants to describe is the internal system, and performance is only the imperfect window through which that system can be observed. This is also why competence cannot be read off directly from a recording: linguists infer it from performance, using native-speaker judgements of what is and isn't a possible sentence to see past the noise.

The pushback: Hymes and communicative competence

Chomsky's idealisation bought analytical clarity at a price, and the sociolinguist Dell Hymes named the price in 1972. In his essay "On Communicative Competence" he argued that Chomsky's competence was too narrow to describe what a real speaker actually knows. A child, Hymes pointed out, does not just acquire the ability to build grammatical sentences; the child "acquires competence as to when to speak, when not, and as to what to talk about with whom, when, where, in what manner." Someone who could generate any grammatical English sentence but had no idea which one was appropriate — who greeted a stranger the way they'd tease a sibling — would not strike us as a competent speaker at all. As Hymes put it, "There are rules of use without which the rules of grammar would be useless."

Hymes replaced the two-way split with a broader question. For any utterance, he argued, a speaker judges it on four dimensions: whether it is formally possible (roughly, grammatical); whether it is feasible given human processing limits (a sentence can be grammatical yet too tangled to parse); whether it is appropriate to the context in which it is used; and whether it is in fact done — the thing people actually say. Chomsky's competence covers only the first of these. Hymes' term for the whole package was communicative competence, and crucially it folds appropriateness — a use-in-context notion Chomsky had exiled to "performance" — back into knowledge itself. Knowing a language, on this account, includes knowing its social rules, not just its syntax. This is the concept that the whole of communicative language teaching would later be built on.

Canale and Swain: breaking competence down

Hymes gave teaching a richer goal, but "communicative competence" was still a slogan, not something you could plan a syllabus or write a test around. The applied linguists Michael Canale and Merrill Swain made it operational. In their 1980 paper "Theoretical Bases of Communicative Approaches to Second Language Teaching and Testing" they broke communicative competence into components that a course could target and an exam could sample. Their original 1980 model had three:

• Grammatical competence — mastery of the linguistic code itself: vocabulary, word formation, pronunciation, spelling and sentence structure. This is roughly Chomsky's competence, now demoted from the whole of language knowledge to one component of it.
• Sociolinguistic competence — knowing the social rules: producing and understanding utterances that are appropriate to the setting, the topic and the relationship between speakers (formal vs informal, polite vs blunt). This is Hymes' appropriateness, made concrete.
• Strategic competence — the coping strategies speakers use when something breaks down: paraphrasing a word they can't retrieve, asking for repetition, using gesture or a filler to buy time. For a learner this component matters enormously, because it lets limited resources still get the message through.

A fourth component, discourse competence — the ability to knit sentences into a coherent, cohesive whole across a paragraph or a conversation, rather than producing correct sentences in isolation — was added by Canale in 1983. The expanded four-part model is the one most textbooks now cite. Its importance for our purposes is structural: it takes Chomsky's single, indivisible "competence" and shows it to be just one quarter of what a proficient user actually commands. Grammatical knowledge is necessary, but on this model a learner can be strong on grammar and still communicatively incompetent — accurate and unusable at the same time.

Usage vs use, and what it means for teaching and testing

Henry Widdowson sharpened the same point into a distinction every teacher recognises. In Teaching Language as Communication (1978) he separated usage — a person's knowledge of linguistic rules, their ability to produce correct forms — from use, their ability to deploy those forms to achieve real communication. A learner can demonstrate perfect usage by conjugating a verb flawlessly in a drill and yet be helpless in use, unable to order a coffee or follow a phone call. Usage maps onto Chomsky's competence; use is where communicative competence lives. Widdowson's argument was that classrooms had been teaching usage and hoping use would follow by itself — and it doesn't reliably follow.

This reframing changed both what "knowing a language" means and how you measure it. If competence in the narrow sense were the goal, the natural test is one that samples grammatical knowledge in isolation: gap-fills, multiple-choice grammar items, translation of decontextualised sentences. Such tests are reliable and easy to mark, but they measure usage and infer use — and the inference often fails, which is the whole reason a person can pass a grammar exam and freeze in conversation. Take communicative competence as the goal and the test has to sample performance under realistic conditions: role-plays, interviews, writing a real email, listening to authentic speech. This is exactly the shift language assessment made across the late twentieth century, and it is a direct descendant of the competence/performance debate. The irony is worth noting: Chomsky idealised performance away as noise, and applied linguistics responded by putting performance at the very centre of how we teach and assess.

The same axis explains the everyday tension between accuracy and fluency. Accuracy is performance that faithfully tracks the grammatical system — few errors, correct forms. Fluency is performance that flows in real time — smooth, appropriate, communicatively effective, even if it isn't error-free. An accuracy-first classroom polishes competence and stops learners the moment they slip; a fluency-first classroom lets them run, tolerating error as the by-product of actually using the language. Neither is wholly right: a learner who is only accurate can't communicate under time pressure, and a learner who is only fluent fossilises mistakes. The competence/performance frame explains why the two can be traded against each other — they sit on opposite sides of the same divide — and why a good course deliberately develops both.

What this means for language learning

Chomsky's distinction gave us the modern definition of the goal: to learn a language is to build an internal generative system, not to memorise a stock of sentences — which is why he matters far beyond his own theory, as our piece on Chomsky's influence on language learning lays out. But the half-century of pushback that followed added the decisive rider: that internal system is worth nothing until it drives real performance, and the knowledge worth having includes knowing when and how to say things, not only what is grammatical. The practical lesson for a learner is to refuse the false choice. Grammar study builds competence; it is necessary and it is not enough. Only repeated, meaningful use — producing and understanding whole utterances in context — turns that competence into performance you can rely on, and that is the ground on which communicative language teaching was built. Learning systems that drill whole sentences, ask you to recall and produce them rather than merely recognise a rule, and space that practice over time — such as the Taalhammer method — are aimed squarely at closing the gap between what you know and what you can actually do.

Frequently asked questions

What is the difference between competence and performance in simple terms?

Competence is the language knowledge in your head — the internal system of rules that lets you build and understand sentences you've never heard before. Performance is what you actually produce when you speak or write, including the mistakes, hesitations and gaps that come from doing it in real time. You can have solid competence and still perform badly under stress; the two are related but never identical.

Who introduced the terms competence and performance?

Noam Chomsky, in Aspects of the Theory of Syntax (1965). He later renamed the pair "I-language" (internal) and "E-language" (external), but the idea is the same. The distinction has an older parallel in Saussure's langue and parole, though Chomsky's version locates the system in the individual mind rather than in the community.

Why does the competence/performance distinction matter for learners?

Because it explains why studying grammar isn't the same as being able to use a language, and why accuracy and fluency can pull in different directions. Dell Hymes, and later Canale and Swain, argued that real knowledge of a language also includes knowing what is socially appropriate — "communicative competence." That is why modern teaching and testing focus on performance in realistic situations, not just on grammar in isolation.

Sources

• Noam Chomsky, Aspects of the Theory of Syntax, MIT Press, 1965.
• Dell Hymes, "On Communicative Competence", in J. B. Pride, J. Holmes (eds.), Sociolinguistics, Penguin, 1972.
• Michael Canale, Merrill Swain, "Theoretical Bases of Communicative Approaches to Second Language Teaching and Testing", Applied Linguistics 1(1), 1980.
• Michael Canale, "From Communicative Competence to Communicative Language Pedagogy", in J. C. Richards, R. W. Schmidt (eds.), Language and Communication, Longman, 1983.
• H. G. Widdowson, Teaching Language as Communication, Oxford University Press, 1978.