You Know the Subject — But Can You Show What You Know in Another Language?
“A student can understand the subject and still fail to demonstrate that understanding when language becomes part of the task.”
— Tymur Levitin
Imagine two students solving the same mathematics problem.
Both understand the mathematics.
Both know which method should be used.
Both can perform the calculation.
But the task is written in a language that is not equally accessible to them.
One student immediately understands what is being asked.
The other has to decode the wording first.
Then comes another problem.
The answer must be explained.
The first student writes a clear justification.
The second knows why the solution works but cannot express the reasoning precisely enough.
The final papers may look very different.
Does that necessarily mean their mathematical knowledge is equally different?
No.
And this distinction becomes crucial whenever people learn, study, solve problems, take examinations or demonstrate academic knowledge through another language.
Subject knowledge and the language used to demonstrate it are not the same thing
Suppose a student knows that photosynthesis converts light energy into chemical energy.
Now ask the student:
Explain how light-dependent reactions contribute to the formation of glucose.
Knowing the biology is one requirement.
But answering the question also requires the learner to:
understand the command explain;
identify the expected causal relationship;
retrieve the relevant terminology;
construct an academically appropriate response;
connect processes linguistically;
and express the explanation with enough precision for another person to evaluate it.
The task therefore does not test biology in complete isolation.
It tests biology through language.
Language can become part of the academic task
This does not mean that every academic difficulty is a language problem.
It also does not mean that subject standards should simply be lowered for multilingual learners.
The important point is more precise:
When knowledge must be accessed, interpreted or demonstrated through language, linguistic processing becomes part of performance.
That creates a distinction between:
subject competence
and
observable subject performance through a particular language.
The two are related.
They are not identical.
The Subject-to-Demonstration Chain
We can represent this process as:
Subject Knowledge → Conceptual Access → Academic Language → Task Interpretation → Response Construction → Demonstration
I will call this the:
Subject-to-Demonstration Chain
It gives us a practical way to locate where performance breaks down.
1. Subject Knowledge
Does the learner actually know the subject?
This must remain our first question.
Sometimes the problem really is mathematical, biological, historical or physical.
The learner does not understand fractions.
The chemical mechanism is unclear.
The historical relationship is missing.
The physics model has not been understood.
Language support cannot replace missing subject knowledge.
If the conceptual foundation is absent, we need to teach the subject.
But we should not assume this is always the cause simply because the final answer is weak.
2. Conceptual Access
Can the learner activate the relevant knowledge?
Knowledge can exist without becoming available at the right moment.
A student may recognize a concept immediately when someone mentions it, but fail to retrieve it independently during a task.
This is familiar in language learning:
recognition is not the same as retrieval.
The same distinction exists in academic subjects.
A learner may understand a theorem when looking at a worked example but fail to identify it as relevant in a new problem.
Before language even becomes the main issue, there can therefore be an access problem.
3. Academic Language
Does the learner possess the linguistic resources required by the subject?
Academic language is not simply “advanced vocabulary.”
Different subjects require different ways of expressing relationships.
Mathematics may require:
greater than, proportional to, tends toward, therefore, assuming that...
Physics may require:
acts on, increases with, remains constant, is transferred, is conserved...
History may require:
contributed to, resulted from, accelerated, was preceded by, can be interpreted as...
Biology may require:
is regulated by, consists of, is transported through, triggers, inhibits...
The learner may understand the underlying relationship while lacking a sufficiently precise way to express it in the language of instruction.
Everyday fluency is not academic fluency
A student can speak English comfortably with friends and still struggle in an English-medium chemistry lesson.
That is not surprising.
Compare:
It gets hotter.
with:
The increase in temperature accelerates the reaction rate under these conditions.
Or:
They wanted more power.
with:
The reform can be interpreted partly as an attempt to centralize political authority.
Academic communication requires specific forms of:
definition;
classification;
comparison;
causation;
qualification;
argumentation;
evidence;
evaluation.
Conversational competence helps.
But it does not automatically provide all of these resources.
4. Task Interpretation
Does the learner understand what the task is asking them to do?
This is one of the most underestimated layers.
Consider these instructions:
Describe.
Explain.
Compare.
Evaluate.
Justify.
Discuss.
Derive.
Estimate.
Interpret.
They are not interchangeable.
A student may know the content and still answer the wrong intellectual question.
For example:
Describe the graph
is different from:
Explain the trend shown in the graph.
The first asks what is visible.
The second asks for relationships or causes.
If the learner does not distinguish the task verbs, a subject teacher may see an incomplete answer even though part of the underlying knowledge is present.
The language of a question can change the difficulty of the problem
Consider a simple mathematical relationship.
The calculation itself may be easy.
But now embed it in a word problem containing:
irrelevant information;
a passive construction;
a comparison;
a condition;
a temporal sequence;
and an unfamiliar academic verb.
The mathematics has not necessarily become harder.
The access route to the mathematics has.
This distinction matters enormously in multilingual education.
5. Response Construction
Can the learner turn knowledge into an acceptable answer?
Knowing the answer internally is not always enough.
The learner may have to construct:
a proof;
an explanation;
a lab report;
a historical argument;
a written solution;
a definition;
an oral presentation;
a comparison.
Each has a structure.
For example, a scientific explanation may require:
claim → mechanism → evidence → conclusion.
A mathematical solution may require:
representation → operation → justification → result.
A historical response may require:
claim → evidence → contextualization → evaluation.
The learner needs both subject understanding and the ability to package that understanding into a form another person can evaluate.
6. Demonstration
What finally becomes visible to the teacher, examiner or institution?
At the end, someone sees:
a spoken answer;
a written paragraph;
a calculation;
a diagram;
an essay;
a presentation;
a test response.
That observable product becomes evidence of competence.
But it is the end of the chain.
If something failed earlier, the final product may underestimate what the learner actually understands.
This is why diagnosis matters.
One weak answer can have several completely different causes
Suppose a student gives a poor answer to a physics question.
Possible explanation 1:
The physics is not understood.
Possible explanation 2:
The physics is understood, but the student did not recognize which principle was relevant.
Possible explanation 3:
The principle was recognized, but the wording of the task was misunderstood.
Possible explanation 4:
The student knew what to say but lacked the academic language needed to explain the causal relationship.
Possible explanation 5:
The student could explain it orally but could not construct the required written response.
Same weak answer.
Different educational problem.
Therefore:
same performance ≠ same diagnosis.
Why “improve your English” is often too vague
A multilingual student receives feedback:
Your English needs improvement.
What exactly needs improvement?
Vocabulary?
Reading speed?
Understanding task verbs?
Sentence construction?
Technical terminology?
Causal explanation?
Argumentation?
Writing conventions?
Retrieval speed?
Listening under classroom conditions?
Without diagnosis, “improve your English” is almost as vague as telling a student:
“Get better at physics.”
Good teaching needs a smaller target.
Why “the problem is only language” can also be wrong
The opposite mistake is equally dangerous.
If a multilingual learner struggles, we may blame language automatically.
But perhaps the academic concept itself is missing.
A student may know all the relevant English vocabulary and still misunderstand Newton's laws.
Or understand every word in a mathematics problem but fail to construct the mathematical representation.
We therefore need to separate:
language difficulty
from
subject difficulty
without pretending that they never interact.
The 2 × 2 Language–Subject Diagnostic Matrix
A useful first diagnostic tool is a simple matrix.
| Language accessible | Language limiting | |
|---|---|---|
| Subject understood | Performance should be relatively strong | Knowledge may be hidden by language |
| Subject not understood | Primarily a subject-learning problem | Language and subject must both be developed |
This gives us four fundamentally different educational situations.
Quadrant 1: Subject strong, language accessible
The learner understands the subject and can work through the language.
Here the main task is normal progression:
more complex problems;
deeper understanding;
greater independence;
higher-level academic performance.
Quadrant 2: Subject strong, language limiting
This is the learner who often says:
“I know this — I just can't explain it in English.”
The wrong intervention would be to reteach the entire subject from zero.
The learner may instead need:
academic vocabulary;
task-language training;
explanation structures;
reading strategies;
oral rehearsal;
discipline-specific writing;
practice demonstrating existing knowledge through the target language.
This is one of the central spaces of Language + Subject learning.
Quadrant 3: Subject weak, language accessible
The learner understands the instructions perfectly.
The language is not the main obstacle.
The subject itself needs work.
Here adding more language exercises would distract from the actual bottleneck.
Teach the mathematics.
Teach the physics.
Teach the biology.
Teach the history.
Quadrant 4: Subject weak, language limiting
This is the most complex situation.
The learner is developing both:
the academic concept;
and
the language through which that concept is taught.
Now the two learning processes interact.
A teacher may need to:
simplify language without simplifying the concept unnecessarily;
make relationships visually explicit;
build terminology alongside concepts;
check understanding through multiple representations;
separate linguistic errors from conceptual errors;
gradually increase the linguistic demands of the task.
This requires integrated teaching rather than pretending we have two completely independent problems.
Translation can help — but it does not solve everything
Suppose a student understands a physics concept perfectly in Ukrainian but studies it in German.
Translation can establish a bridge:
Ah, this German term refers to the concept I already know.
That is useful.
But the student may eventually need to:
read German tasks directly;
follow German explanations;
use German terminology;
write solutions;
participate in class;
take an examination.
At that point, translation alone is insufficient.
Existing knowledge has to become operational through the new language.
Learning the terminology is not the same as learning the subject through the language
A student can memorize:
force — Kraft
acceleration — Beschleunigung
mass — Masse
and still be unable to follow a German physics problem.
Why?
Because academic language also includes relationships:
wirkt auf
ist proportional zu
nimmt zu
bleibt konstant
lässt sich daraus ableiten
The real unit of academic competence is often not an isolated term.
It is the ability to construct relationships between concepts.
Mathematics is not language-free
Mathematics is sometimes treated as if language barely mattered because numbers are universal.
But mathematical learning contains enormous amounts of language.
Students must interpret:
conditions;
quantifiers;
comparisons;
definitions;
logical relationships;
word problems;
proof instructions;
exceptions;
constraints.
Consider the difference between:
at least
and
at most.
Or:
increases by
and
increases to.
A tiny linguistic distinction can completely change the mathematics.
Physics adds another translation layer
Physics is particularly interesting because the learner may have to perform several translations:
natural-language situation → physical model → mathematical representation → physical interpretation
In our German reference page Warum du die Physikformeln kennst, aber trotzdem keine Aufgaben lösen kannst, we describe this through the Physics Translation Chain:
Situation → System → Quantities → Relations → Model → Equations → Solution → Physical Check
Now imagine that the initial situation itself is written in a language the student is still developing.
A linguistic bottleneck appears before the physical modeling even begins.
That is precisely why Language + Subject deserves its own educational architecture.
Biology requires language for systems and processes
A biology student may know all the nouns:
cell;
membrane;
protein;
enzyme;
DNA.
But academic understanding becomes visible through relationships:
passes through
binds to
is encoded by
regulates
causes
inhibits
is converted into
A list of terminology is not yet a biological explanation.
The student needs language for process architecture.
History requires language for interpretation
History presents a different challenge.
It is not enough to know:
dates;
names;
events.
Students need to express:
causation;
continuity;
change;
perspective;
uncertainty;
relative importance;
evidence;
interpretation.
Compare:
X caused Y
with:
X contributed to Y
and:
X created conditions that made Y more likely.
These are not stylistic decorations.
They express different historical claims.
Academic language is part of disciplinary precision.
A student may appear less intelligent in another language
This is one of the most important consequences.
A person can sound simpler in a second or third language than the complexity of their thinking actually is.
Their available language may force them to express:
a nuanced idea as a basic statement;
a qualified claim as an absolute one;
a complex causal relationship as two simple sentences.
An observer who sees only the output may underestimate the underlying knowledge.
This does not mean we should ignore the weakness in output.
It means we should diagnose it correctly.
The goal is not to excuse weak performance
This distinction must be clear.
Saying:
“Language is limiting the demonstration of knowledge”
does not mean:
“The student does not need to learn the language.”
Often it means exactly the opposite.
If the student needs to study, work or take examinations through that language, then academic language is a real competence that must be developed.
But now we know what to develop.
We are no longer confusing it with missing subject knowledge.
How can we test whether the bottleneck is language or subject?
Use multiple representations.
Ask the learner to explain the concept:
in the target language;
in their stronger language;
with a diagram;
through equations;
through an example;
by selecting between alternatives;
by demonstrating a process.
If conceptual performance changes dramatically when the language demand changes, that gives us diagnostic information.
It does not automatically prove that language is the only problem.
But it helps us locate the bottleneck.
The Same-Concept / Different-Language Test
A particularly useful diagnostic method is:
Step 1
Give the learner a subject problem in the language of instruction.
Step 2
Observe where the difficulty appears.
Step 3
Present the same underlying concept through a linguistically more accessible form.
Step 4
Compare performance.
Step 5
Ask the learner to reconstruct the reasoning in the target language.
This allows us to separate three questions:
Does the learner understand the concept?
Can the learner access it through the target language?
Can the learner demonstrate it through the target language?
Those are not the same question.
Assessment should know what it is measuring
Suppose an examination is intended to measure physics.
But success depends heavily on difficult linguistic structures unrelated to the intended physics construct.
Then language may introduce additional variance into the result.
Sometimes this is unavoidable.
Sometimes language is legitimately part of the competence being assessed.
For example, if a student must study physics professionally in German, explaining physics in German is itself relevant.
The key is not to pretend language never matters.
The key is to know whether language is part of the target competence or an unintended barrier to measuring another competence.
Teaching through another language can deepen understanding
So far we have discussed difficulty.
But Language + Subject learning can also create advantages.
When learners meet the same concept through another language, they may be forced to notice distinctions they previously took for granted.
Different terminology can expose:
hidden assumptions;
different conceptual boundaries;
alternative metaphors;
different ways of organizing information.
Explaining the same concept in two languages can therefore become a tool for conceptual comparison, not merely translation.
A second language can reveal the structure of a concept
Suppose you know a concept so well in your first language that its terminology feels transparent.
You no longer notice how the concept is packaged.
Another language makes the packaging visible again.
You ask:
Why is this relation expressed this way?
Does the other language divide the concept differently?
Is the technical term historically constructed from different components?
What distinction becomes explicit here?
This is one reason multilingual academic learning can become intellectually powerful when taught deliberately.
Language + Subject is not “a language lesson with some physics vocabulary”
Nor is it:
a physics lesson translated word for word.
A genuine integrated approach asks:
What does the learner need to understand in the subject?
What linguistic operations are required to access that understanding?
What linguistic operations are required to demonstrate it?
Where should language support be explicit?
Where should the learner struggle productively with the subject?
Which difficulty belongs to which layer?
This is a diagnostic architecture.
Three different lesson goals
A Language + Subject lesson may have three very different primary goals.
Goal A — Learn the subject
The language is mainly a medium.
Goal B — Learn the academic language of the subject
The concepts may already be known.
Goal C — Develop both simultaneously
Subject and language are both active learning targets.
Confusing these goals makes lessons inefficient.
A good teacher knows which one is dominant at a particular moment.
Why individualization matters here
Two students can attend the same German-medium mathematics class.
Student A understands the mathematics but struggles with German task language.
Student B speaks German well but has gaps in algebra.
Giving both “more German mathematics practice” sounds reasonable.
But they need different things.
A needs linguistic access to existing mathematical competence.
B needs mathematical development through language that is already sufficiently accessible.
Same class.
Different bottleneck.
The Language–Subject Bottleneck Principle
This leads to a broader principle:
Performance in Language + Subject learning is constrained by the weakest necessary link between subject understanding and the ability to access and demonstrate that understanding through the required language.
I will call this the:
Language–Subject Bottleneck Principle
It explains why adding more subject content may fail when language is the limiting link — and why adding more language lessons may fail when the missing link is conceptual understanding.
More language is not always the answer
Suppose the student understands every word in a mathematics problem but does not know how to represent the situation.
Another vocabulary worksheet will not solve it.
This is a mathematical reasoning problem.
Likewise, if the student can solve the same problem immediately in Ukrainian but cannot interpret the German wording, another page of algebra may not address the actual obstacle.
Good diagnosis prevents us from training the wrong system.
More subject practice is not always the answer either
A student repeatedly solves chemistry exercises but fails when questions use unfamiliar academic wording.
The teacher gives twenty more chemistry exercises with the same linguistic structure.
Performance improves.
Then the examination phrases the same concept differently.
The difficulty returns.
The student may have learned the surface pattern of the task, not developed flexible access to the chemistry through academic language.
Variation matters.
Independence requires both systems to work together
At first, a teacher may translate.
Explain terminology.
Rephrase questions.
Highlight task verbs.
Provide sentence frames.
Help structure an answer.
These supports can be useful.
But if the learner's real goal is independent study or work through the target language, the support must gradually change.
The progression becomes:
supported access → guided performance → reduced support → independent academic action
The goal is not permanent simplification.
It is increasing independence.
From “I know it” to “I can demonstrate it”
Our broader Four-Level Learning Model distinguishes:
Knowledge → Understanding → Ability → Independence
Language + Subject adds another question at every level:
Through which language can the learner perform this competence?
A person may have subject understanding.
Now that understanding must become operational under the linguistic conditions of real life.
School.
University.
Examination.
Work.
Migration.
Professional communication.
That is not merely language learning.
And it is not merely subject tutoring.
It is the development of integrated competence for action.
Errors can reveal which layer is failing
A weak answer is data.
Ask:
Was the concept wrong?
Was the task misunderstood?
Was the terminology inaccessible?
Was the relationship understood but expressed imprecisely?
Was the response structure inappropriate?
Was the learner unable to retrieve the required language quickly enough?
Our German framework Nicht jeder Fehler bedeutet dasselbe: Wie man Lernfehler richtig diagnostiziert develops this general principle:
different errors can require different interventions.
Language + Subject makes that distinction even more important because errors can originate in two interacting systems.
The final grade is the end of the process, not the explanation
A student receives 62%.
That number tells us something about performance under the assessment conditions.
It does not automatically tell us:
how much of the subject was understood;
how much language interfered;
whether the problem was retrieval;
whether the instructions were misinterpreted;
whether the learner could demonstrate the same concept another way;
whether the student could perform better with different linguistic access.
Grades are outcomes.
Teaching requires diagnosis.
A practical diagnostic sequence for Language + Subject
When a learner struggles with a subject through another language, ask in this order:
1. Can the learner explain the underlying concept in a stronger language?
If not, investigate the subject knowledge.
2. Can the learner recognize the relevant terminology in the target language?
If not, build conceptual-language mappings.
3. Can the learner interpret the task independently?
If not, work with academic instructions and task structures.
4. Can the learner solve the subject problem once the task is understood?
If not, investigate subject reasoning.
5. Can the learner explain or write the solution in the target language?
If not, develop discipline-specific academic production.
6. Can the learner do all of this without scaffolding?
If not, reduce support gradually and develop independence.
This sequence prevents us from treating every difficulty as the same difficulty.
The purpose is real-world competence
A learner may need to:
study mathematics in German;
take biology in English;
attend university in Polish;
work with technical documentation in Spanish;
take an entrance examination in another language;
move into a new school system;
communicate professionally about a subject.
The educational goal is therefore not:
learn the subject for the subject's sake
or
learn the language for the language's sake.
The goal is:
be able to perform the required real-world academic or professional action through the required language.
When the language stops hiding the knowledge
The strongest outcome is not merely that the learner knows more terminology.
It is that the gap between:
what I understand
and
what I can demonstrate
becomes smaller.
The learner can read the task.
Recognize the concept.
Select the method.
Use the terminology.
Construct the explanation.
Check the result.
And do it with progressively less external support.
At that point, language is no longer primarily a barrier between knowledge and performance.
It has become one of the tools through which knowledge can operate.
“The goal of Language + Subject learning is not to make the learner choose between knowing the subject and knowing the language. It is to make subject knowledge increasingly usable through the language in which real life requires it.”
— Tymur Levitin
Continue Learning
For the general distinction between knowledge, understanding, ability and independence, read Knowing vs Understanding: The Four Levels of Real Learning.
For a concrete example of how subject reasoning works before formulas can be applied, see Warum du die Physikformeln kennst, aber trotzdem keine Aufgaben lösen kannst.
For mathematical thinking and the difference between obtaining an answer and understanding the mathematics, see Understanding Mathematics: How Mathematical Thinking Develops.
For the broader architecture of learning a subject through another language, continue with Learn a Subject Through Another Language: How Language and Knowledge Develop Together.
Individual Language + Subject Learning Online
Levitin Language School provides individual online learning for children, teenagers, university students and adults internationally.
Our educational system works across three connected but distinct layers:
Languages · School and Academic Subjects · Language + Subject
The third layer is particularly important for learners who need not merely to study another language, but to use that language to learn, solve problems, take examinations, study at school or university, or work with academic and professional knowledge.
Depending on the learner's real task, we can therefore diagnose and develop the language component, the subject component, or the interaction between both rather than treating every difficulty as a generic “language problem” or “subject problem.”
International and U.S.-focused educational resources are also available through Language Learnings.
Contact — Levitin Language School
Email: notification@levitintymur.com
Phone / WhatsApp: +380 93 291 34 29
WhatsApp: https://wa.me/380932913429
Telegram: https://t.me/START_SCHOOL_TYMUR_LEVITIN
Telegram: @START_SCHOOL_TYMUR_LEVITIN
Website: https://levitintymur.com/
About the Author
Tymur Levitin
Founder & Director, Levitin Language School
Educator, language teacher and author working across language learning, academic education, learning diagnosis, problem solving and integrated Language + Subject education.
His work focuses on a practical distinction that is central to multilingual education: what a learner knows, what a learner understands, and what that learner can actually access, communicate and demonstrate through the language required by a real academic or professional environment.
Levitin Language School: https://levitintymur.com/
Language Learnings — USA: https://languagelearnings.com/
Language Thinking Laboratory: https://languagethinkinglab.blogspot.com/
Author contact: tymurlevitin@levitintymur.com
© Tymur Levitin — Founder & Director, Levitin Language School. All rights reserved.
Comments
Post a Comment