A Diagram Can Be Correct and Still Explain Nothing
Why accurate representations are not the same as understanding
A diagram can contain every correct label.
Every arrow can point in the expected direction.
Every term can be spelled correctly.
Every element can be placed where the textbook says it belongs.
And the diagram can still explain almost nothing.
This sounds contradictory only if we assume that representing knowledge and explaining knowledge are the same intellectual act.
They are not.
A representation shows us something about a system.
An explanation tells us why the system behaves as it does.
That difference matters in biology, physics, mathematics, history, language — and in education itself.
Start With a Perfectly Correct Diagram
Consider:
CELL → TISSUE → ORGAN → SYSTEM
Nothing is wrong with it.
It represents levels of biological organization in a useful simplified form.
A student may memorize it.
They may reproduce it.
They may correctly identify what comes before or after organ.
But now ask:
What does the first arrow mean?
Does a cell simply become a tissue?
Do identical cells accumulate until a tissue exists?
Does organization matter?
Do interactions matter?
Does specialization matter?
What properties appear at the tissue level that cannot be understood merely by naming individual cells?
The original diagram does not answer those questions.
It was correct.
But correctness was not the same as explanation.
Representation Compresses
This is not a defect of diagrams.
It is what makes them useful.
Representations compress complexity.
A map leaves out most of the physical world.
A formula leaves out many details of the situation it models.
A timeline reduces years of interacting processes to selected events.
A grammatical table removes conversation, intention, tone and context so that a structural pattern becomes visible.
Compression allows us to think.
The problem begins when we forget that compression has occurred.
A useful representation can then be mistaken for the phenomenon itself.
An Arrow Is Not a Mechanism
Suppose we write:
A → B
The arrow looks explanatory.
But what exactly does it mean?
Perhaps:
A causes B.
Or:
A changes into B.
Or:
A enables B.
Or:
B depends on A.
Or:
A usually precedes B.
Or:
A increases the probability of B.
Or simply:
After discussing A, we will discuss B.
The visual symbol is identical.
The underlying relationships are radically different.
An arrow therefore does not automatically contain an explanation.
It can merely mark the place where an explanation is needed.
Biology: Structure Does Not Explain Function by Itself
Biology contains enormous amounts of visual representation.
Cells are drawn.
Organs are labelled.
Pathways are mapped.
Systems are divided into components.
These representations are indispensable.
But imagine that a learner can identify every part of a biological structure while being unable to answer:
Why is it shaped this way?
How does this structure support its function?
What interacts with what?
What changes if this component is damaged?
How might the effect propagate through the larger system?
The learner has access to the representation.
The mechanism remains weak.
This is one reason biology can feel like a subject of memorization even though biological understanding is deeply relational.
Physics: The Equation Can Be Correct Too
Write:
F = ma
The equation is correct within its appropriate model.
A student knows the symbols.
They substitute numbers correctly.
They obtain the correct answer.
Now ask:
Why does this equation apply to this situation?
That is a different question.
Or change the representation.
Remove the familiar diagram.
Introduce an irrelevant quantity.
Describe the situation verbally.
Ask what would happen if one variable changed.
The equation did not change.
The learner's ability to connect reality to the equation is what is now being tested.
A formula represents a relationship.
Understanding requires knowing how that relationship maps onto a physical situation.
Mathematics: A Procedure Can Hide the Problem
Mathematics produces another form of representation:
the worked example.
A textbook shows:
Step 1 → Step 2 → Step 3 → Answer
Everything may be mathematically valid.
But a learner can imitate the sequence without understanding why those transformations are licensed.
This becomes visible when the surface changes.
The numbers are different.
The unknown moves.
An unnecessary piece of information appears.
The familiar visual structure disappears.
Suddenly the procedure cannot be copied.
The question becomes:
What mathematical relationship made those steps legitimate?
A correct procedure can demonstrate how a problem was solved.
It does not guarantee that the learner understands why the solution works.
History: A Timeline Is Not a Causal Model
A timeline may show:
Event A → Event B → Event C
Every date may be correct.
But chronological order does not establish causal structure.
Why did B follow A?
Did A make B possible?
Did it make B more likely?
Was B primarily a reaction to A?
Were both consequences of another process?
Would B probably have occurred without A?
A timeline represents temporal order beautifully.
It may represent causation poorly unless the relationships are explicitly investigated.
Again:
correct representation ≠ sufficient explanation
Language: The Grammar Table Problem
Language teaching contains its own highly compressed representations.
A tense table may show:
I work
I worked
I will work
The forms are correct.
But what has actually been explained?
Perhaps morphology.
Perhaps a simplified time contrast.
But real tense use involves relationships among events, reference points, discourse context, speaker perspective and communicative intention.
A learner can therefore reproduce an entire grammatical table and still hesitate when deciding what to say in an actual conversation.
The table was not useless.
It was incomplete by design.
The problem appears only when the representation is mistaken for the whole linguistic system.
Even Words Are Representations
This principle goes deeper.
A technical term is itself a compressed representation.
Consider the word:
photosynthesis
Knowing the term is useful.
Knowing its definition is better.
But the label does not itself contain the full mechanism.
The same happens with:
democracy
acceleration
metaphor
inflation
ecosystem
case
probability
A learner can recognize the label while possessing very different depths of knowledge underneath it.
This is why vocabulary can sometimes disguise conceptual gaps.
The name is available.
The structure behind the name may not be.
So What Turns Representation Into Explanation?
Not more labels.
We need relationships.
Then mechanisms.
Then conditions.
Then consequences.
A useful progression is:
REPRESENTATION → RELATIONSHIP → MECHANISM → CONDITIONS → CONSEQUENCES
Suppose we have:
A → B
First ask:
What relationship does the arrow represent?
Then:
By what mechanism does A affect B?
Then:
Under what conditions does this relationship hold?
Then:
What should we expect if A changes?
Now the diagram is beginning to support explanation rather than merely recognition.
Explanation Creates Predictions
One of the strongest tests of an explanation is whether it allows us to reason beyond the original example.
If I understand why a biological structure performs a function, I should be able to reason about what may happen when the structure changes.
If I understand a physical relationship, I should be able to anticipate the direction of change before calculating it.
If I understand a historical mechanism, I should be able to discuss what evidence would strengthen or weaken a causal interpretation.
If I understand a grammatical distinction, I should be able to reason about a new communicative situation rather than waiting for an exercise to name the required form.
Explanation makes knowledge productive.
It allows the learner to generate something that was not explicitly memorized.
This Changes How We Ask Questions
Instead of asking only:
What is this?
we can add:
How is it connected to that?
Instead of:
Which formula should you use?
ask:
Why does this relationship apply here?
Instead of:
What happened next?
ask:
What mechanism connects these events?
Instead of:
Which tense is correct?
ask:
What relationship between the events are you trying to express?
The first questions are not bad.
Identification matters.
But the second questions reveal whether the representation has acquired explanatory depth.
Language + Subject Makes the Difference Visible
When learners work with an academic subject through another language, a revealing distinction appears.
A student may know the English names of:
cell
tissue
organ
system
But explaining biology requires language for relationships:
consists of
depends on
contributes to
regulates
causes
results in
increases
decreases
is affected by
differs from
This is where Language + Subject becomes intellectually interesting.
Language is no longer simply attached to the labels.
It becomes a tool for constructing the explanation.
The learner must make the arrows explicit.
A Better Question for Teachers
When a student reproduces a correct diagram, formula, table or sequence, we can ask:
What is this representation leaving out?
That question is powerful because every representation leaves something out.
Then ask:
Does the omitted information matter for the problem we are trying to understand?
Sometimes it does not.
Good models deliberately ignore irrelevant complexity.
Sometimes it matters enormously.
The educational task is not to demand maximum complexity everywhere.
It is to help learners distinguish between:
useful simplification
and
missing explanation.
Correctness Has Layers
A representation can therefore be correct at one level and insufficient at another.
A biological diagram can be anatomically correct but mechanistically weak.
A historical timeline can be chronologically correct but causally empty.
A grammatical table can be morphologically correct but pragmatically inadequate.
A mathematical procedure can be operationally correct but conceptually opaque.
These are not contradictions.
They are different dimensions of knowledge.
Recognizing them helps us avoid a common educational mistake:
treating visible correctness as proof of invisible understanding.
The Diagram Is the Beginning
We need diagrams.
We need formulas.
We need timelines.
We need tables.
We need terminology.
We need simplified models.
Education would become impossible without compression.
But a representation should open the door to reasoning rather than close it.
The question after seeing a diagram should not always be:
Can you remember this?
Sometimes the better question is:
Can you explain why it looks this way?
And then:
What would happen if something changed?
Because a diagram can contain every correct box and every correct arrow and still leave the most important intellectual work unfinished.
Representation shows us the structure.
Explanation tells us why the structure matters.
And understanding begins when the learner can move between the two.
Continue Exploring
Knowing vs Understanding: The Four Levels of Real Learning
https://languagethinkinglab.blogspot.com/p/knowing-vs-understanding-four-levels-of.html
Biology Is More Than Facts: How Systems Turn Information Into Understanding
https://levitintymur.com/biology-is-more-than-facts-how-systems-turn-information-into-understanding/
The Question After the Correct Answer: How to Test Whether a Student Really Understands
https://timurlevitin.blogspot.com/2026/09/the-question-after-correct-answer-how.html
The Arrows Matter More Than the Boxes
https://medium.com/@timurlevitin/the-arrows-matter-more-than-the-boxes-3a37d770b91a
Tymur Levitin
Founder & Director, Levitin Language School / Language Learnings
Languages • Academic Subjects • Language + Subject
© 2026 Tymur Levitin. All rights reserved.
Comments
Post a Comment