Content and Language Integrated Learning (CLIL)

Content and language integrated learning (CLIL) means teaching school subjects — geography, biology, history — through a language that is not the students’ first, with a deliberately double aim: learn the subject and learn the language in the same lessons. The term is an umbrella, coined in 1994 by David Marsh and colleagues, and it was stretched wide on purpose: it covers everything from a full science course taught in English to a few content-flavored “language showers” a week. But the typical European case has a recognizable shape: the teaching language is a foreign language — overwhelmingly English — rather than a language of the local community; the setting is an ordinary state school; and the foreign language carries a slice of the curriculum, not half of it.

This entry is about that European branch of a larger family: where the term came from, the framework that gives it pedagogical substance, its hard and soft varieties, and what the research — and its critics — have found. For the family as a whole — the school models that teach content in two languages, from transitional programs to Canadian immersion, and the long-term evidence on them — see the entry on bilingual education.

A term coined for a European problem

CLIL has an unusually precise birth certificate. The acronym was launched in 1994 by David Marsh, then working at the University of Jyväskylä in Finland, in work connected to European Commission language initiatives. The practice it named was not new: Canadian French immersion had been running since 1965, and North American schools had a parallel tradition of content-based instruction. The new name was strategic. “Immersion” implied a second language of the surrounding community and half the timetable or more; “bilingual education” dragged along decades of American political controversy. Europe needed a neutral word for its own arrangement — a foreign language, an ordinary school, a couple of subjects — and CLIL, defined as a dual-focused approach in which an additional language is used for the learning and teaching of both content and language, was cut to fit exactly that.

The problem it answered was part arithmetic, part politics. The European Commission’s 1995 White Paper on education set the goal that every EU citizen should master two Community languages beyond the mother tongue — a target the Barcelona European Council of 2002 sharpened into the “mother tongue plus two” formula. School timetables, however, were full: no country could triple its foreign-language lesson hours. CLIL was the elegant way out — convert existing subject hours into language contact hours. The CLIL/EMILE report Marsh compiled for the Commission in 2002 made the case explicitly, and the Commission’s Eurydice network surveyed the result in 2006: some form of CLIL existed in nearly every European country, though usually in a minority of schools and for a minority of pupils, with English as the vehicular language in the great majority of programs — alongside a quieter parallel current of CLIL in regional and minority languages, from Welsh to Basque. CLIL managed to be everywhere and rare at the same time.

The 4Cs: Coyle’s framework

An umbrella term has a weakness: anything can shelter under it, including a history lesson “translated” into stilted English, or an English lesson wearing a thin costume of biology. The framework that gave CLIL pedagogical substance is Do Coyle’s 4Cs, sketched in 1999 and canonized in the field’s standard reference, Coyle, Hood and Marsh’s CLIL (2010). A CLIL lesson is planned in four interlocking dimensions: content — the subject matter, which drives the lesson; communication — the language learned and used to handle that content; cognition — the thinking skills the content demands, from classifying to hypothesizing; and culture — the intercultural perspective that a second language naturally opens on any topic. The point of the geometry is balance: drop cognition and you get vocabulary drill; drop communication and you get a silent worksheet; drop content and you are back to an ordinary language lesson.

The framework’s most practical piece is the language triptych, which replaces the grammar syllabus with three questions. What is the language of learning — the terminology and text types the subject itself requires (“photosynthesis”, “the graph shows…”)? What is the language for learning — the classroom language pupils need in order to operate (asking for clarification, disagreeing, reporting group work)? And what language will emerge through learning — needs nobody could have predicted, met as they arise? This is CLIL’s deepest break with conventional teaching: language is sequenced by the needs of the content, not by a grammatical gradient from the present simple upward. That is at once its strength — the language is real and immediately used — and, as the research below shows, its structural weakness: what the content never happens to demand may never get learned.

Hard CLIL, soft CLIL and the classroom reality

Practice spans a continuum that teachers call hard and soft CLIL. In hard, content-driven CLIL, a subject teacher teaches geography or chemistry through the foreign language, follows the subject curriculum and grades subject knowledge; language learning is deliberate but secondary. In soft, language-driven CLIL, a language teacher builds language lessons around subject content — a unit on volcanoes inside the English class — and the content serves the language. Between the poles lie mixed forms: bilingual modules, project weeks, short “showers”. The distinction matters because each end has its own failure mode: hard CLIL risks pupils drowning in language they cannot yet lift; soft CLIL risks content so thin it motivates nobody.

What makes either end work is scaffolding — the machinery of support that separates a CLIL lesson from mere exposure: visuals and diagrams, glossaries, graphic organizers, sentence frames (“The higher the temperature, the…”), key terms pre-taught, and judicious use of the first language to check that a concept — not just its label — has landed. Two chronic dilemmas remain. Assessment: if a pupil explains erosion brilliantly in broken English, what grade does she get, and in which subject? And staffing, the bottleneck everywhere: hard CLIL needs subject specialists confident in a foreign language — a rare species that teacher-training systems across Europe have struggled to breed.

In Poland the model exists under its own name: oddziały dwujęzyczne, bilingual streams within ordinary schools that teach at least two subjects partly or wholly in a foreign language — since the 2017 reform available from the seventh grade of primary school, and crowned by the matura dwujęzyczna, a school-leaving language exam pitched explicitly above the standard advanced level, with optional subject papers taken in the teaching language. Structurally this is hard CLIL de facto: foreign language, ordinary school, a slice of the curriculum. One detail deserves attention before reading any research: entry to these streams runs through a language-aptitude test, so the pupils inside are selected. Keep that in mind for the next section.

What the research shows — and the selection problem

The most cited synthesis is Christiane Dalton-Puffer’s 2011 review, and its pattern is remarkably consistent across countries. CLIL students show clear advantages in vocabulary (especially receptive and subject-specific), listening comprehension, fluency and the affective outcomes — confidence, willingness to speak, lower anxiety. The advantages shrink or vanish for accuracy: syntax, writing and pronunciation improve little or not at all. Subject knowledge, meanwhile, is generally not harmed — most studies find CLIL pupils matching conventionally taught peers on the content itself. Readers of the bilingual education entry will recognize the shape: it is the Canadian immersion result in miniature — massive meaning-focused input builds comprehension and fluency, while grammatical precision does not come free.

Dalton-Puffer’s own classroom-discourse research (2007) points at one mechanism. Recordings of CLIL lessons across Austrian schools showed classrooms dominated by teacher talk, with pupils mostly producing short factual answers: the students understood a great deal and said very little — precisely the imbalance that Merrill Swain’s output hypothesis predicts will leave production lagging behind comprehension. Academic language functions — defining, hypothesizing, justifying — did not develop by osmosis; they appeared only where teachers taught them deliberately.

The sharpest critique targets the comparisons themselves. Anthony Bruton (2011) re-examined celebrated CLIL studies and argued that much of the measured advantage may be selection: CLIL streams recruit by aptitude test or attract motivated pupils who already have better English and more out-of-school contact with it, and they often receive extra English hours on top. Comparing such a class with a regular one and crediting the difference to CLIL flatters the method. The German longitudinal evidence bears the worry out: Dominik Rumlich’s large study of schools in North Rhine-Westphalia (2016) found that future CLIL pupils were already substantially ahead in English before their CLIL instruction began, and the additional program-attributable gains over two years were modest. None of this makes CLIL worthless — the input, the authentic need and the motivation are real — but the honest reading is narrower than the enthusiasm: CLIL multiplies contact with the language for pupils likely to thrive anyway, and it does not, by itself, deliver accuracy.

What this means for language learning

Strip away the school scaffolding and CLIL’s logic travels well — three lessons in particular. First, content is the cheapest source of massive input: learning something you actually care about through the language — a course, a craft channel, a history podcast — supplies hours of comprehensible input with motivation built in, and that is exactly the ingredient the research keeps validating: vocabulary, listening and fluency grow. Second, what the content never demands, exposure never teaches: the CLIL and immersion evidence agree that accuracy — grammar, writing, pronunciation — needs deliberate work, being pushed to produce and corrected, not just more input. Pair content with systematic production practice: active recall and spaced repetition of full sentences supplies precisely the output-and-form component that content-driven learning lacks. Third, read the selection critique optimistically: what actually powered those CLIL classes — motivation, extra hours, prior knowledge — is not a school privilege. An adult learner can grant all three to herself.

Frequently asked questions

Is CLIL the same as immersion?

Same logic, different scale and language status. Immersion schools children largely or entirely in a second language, usually one spoken in the surrounding community — French in Canada — and can occupy most of the timetable. CLIL typically means a few subjects taught through a foreign language, most often English, inside an otherwise ordinary school. Both teach content through language; CLIL is immersion’s machinery scaled down and pointed at a lingua franca.

Does learning a subject in a foreign language harm knowledge of the subject?

The research consensus is: generally not — CLIL pupils usually match conventionally taught peers on content — but the result is conditional on good practice. Where scaffolding is strong and the first language stays available for checking difficult concepts, content survives the language barrier; where a linguistically overloaded class merely copies notes, it does not. The assessment question — grading content versus grading language — remains one of CLIL’s genuinely unsolved problems.

Do CLIL students end up better at the language?

On average yes — in vocabulary, listening, fluency and confidence — with two caveats. Part of the measured advantage reflects selection and extra hours rather than the method itself, as Bruton’s critique and Rumlich’s longitudinal data show. And the advantage is uneven: grammatical accuracy, writing and pronunciation profit little unless the program adds explicit language work. Content teaching multiplies input; it does not replace focused practice.

Sources

David Marsh (ed.), CLIL/EMILE — The European Dimension: Actions, Trends and Foresight Potential, European Commission, 2002.
Eurydice, Content and Language Integrated Learning (CLIL) at School in Europe, European Commission, 2006.
Christiane Dalton-Puffer, Discourse in Content and Language Integrated Learning (CLIL) Classrooms, John Benjamins, 2007.
Do Coyle, Philip Hood, David Marsh, CLIL: Content and Language Integrated Learning, Cambridge University Press, 2010.
Kay Bentley, The TKT Course: CLIL Module, Cambridge University Press, 2010.
Christiane Dalton-Puffer, “Content-and-Language Integrated Learning: From Practice to Principles?”, Annual Review of Applied Linguistics 31, 2011.
Anthony Bruton, “Is CLIL so beneficial, or just selective? Re-evaluating some of the research”, System 39(4), 2011.
Dominik Rumlich, Evaluating Bilingual Education in Germany: CLIL Students’ General English Proficiency, EFL Self-Concept and Interest, Peter Lang, 2016.