What is the primary function of human language—thought or communication?
At first glance, the question may seem almost terminological. Language is obviously used for both. We speak to other people—and we speak to ourselves. We formulate messages—and we formulate thoughts.
But if we ask not what modern humans can use language for, but what problem might have led to its emergence, the question becomes much more interesting.
Noam Chomsky has long argued that the fundamental organization of human language is primarily connected with thought rather than communication. I would like to propose the opposite hypothesis.
Perhaps thinking in the form of T emerged precisely because communication became necessary.
If so, the possibility of subsequently interpreting and processing information would not be the original reason for the emergence of T, but a derivative consequence of it.
What Chomsky Actually Claims
It is important to represent Chomsky’s position accurately.
He does not simply say that people often use language for thinking. His claim is considerably stronger.
Chomsky distinguishes between an internal language system and its externalization—its expression through speech, sign language, and other sensorimotor means. The internal system generates hierarchically structured expressions that are interpreted as thoughts; externalization makes it possible, among other things, to use them for communication.
In his work on language evolution, Chomsky explicitly argues against identifying language with communication. He regards language as a specific internal cognitive system, while its externalization and communicative use are secondary to this internal core.
At the same time, Chomsky formulates his position more cautiously than it is sometimes presented. The statement that the “function of language” is thought is, by itself, too broad. The stronger substantive claim is that the design of language is a system of thought.
Thus, the popular formulation “language was created for thinking” would be too teleological and too crude.
But the direction of the hypothesis is indeed this:
internal system first—communication later.
Chomsky has even sketched a possible evolutionary scenario: a change in the brain might initially have given a single individual a new internal combinatorial capacity that provided advantages in planning, interpretation, and thought. Only after this capacity had spread through a population would connecting it to sensorimotor systems—and thus externalizing it—have become relevant.
I would like to consider the opposite possibility.
Why Should Knowledge Become Information at All?
In the model of the psyche proposed here, the functions S and N generate knowledge about objects.
S generates knowledge of invariants—of what remains unchanged.
N generates knowledge of changes—of what changes.
But knowledge existing in the psyche of one subject cannot be transmitted directly to another subject.
Suppose I see an animal approaching.
For myself, I do not need first to translate the knowledge that has arisen into the sentence:
“A large animal is approaching us.”
The knowledge already exists as content of my psyche.
But the situation changes fundamentally when this knowledge is significant for the non-subject—for example, for another person.
Now a new problem arises:
How can I make my knowledge accessible to someone who does not possess it?
Psychic content itself cannot be transmitted.
Knowledge must therefore acquire a represented form with which another organism can interact and in relation to which another psyche can generate its own content.
Perhaps this is where we should look for the origin of the T function.
T as a Function of Communication
In the proposed model, T—thinking—generates information about an object in accordance with its significance for the non-subject.
This definition acquires a particular meaning if T is considered in relation to communication.
First there is knowledge.
Then the significance of this knowledge for the non-subject arises.
But the second subject cannot directly acquire the knowledge that exists in the psyche of the first.
The knowledge must therefore be represented in a form suitable for transmission, interpretation, and processing.
This is how information arises.
The hypothetical sequence might therefore be:
knowledge about an object → significance of this knowledge for the non-subject → need to make the knowledge accessible to the non-subject → representation of knowledge in a transmissible form → information.
If this hypothesis is correct, communication is not a later application of an already existing T.
On the contrary:
The need for communication may have been an evolutionary condition for the emergence of the T function itself.
The possibility of subsequently using information for internal interpretation, combination, and further processing may have developed later.
Why Does F Not Need an Equivalent Form?
Here the comparison with the opposite rational function, F, becomes particularly interesting.
F generates the significance of an object for the subject.
This significance can be positive or negative and can vary in intensity.
I may like a piece of music a little.
I may love a person very deeply.
An activity may be mildly unpleasant.
Another object may provoke intense rejection.
Thus, F-content cannot be reduced to a binary alternative:
acceptance—rejection;
pleasure—displeasure.
There are degrees of intensity between the poles.
But for this significance to exist for the subject itself, it does not first have to be represented in a form suitable for communication.
If I love someone, my love does not have to become the sentence “I love this person” in order to exist for me.
If I dislike a smell, I do not have to call it unpleasant in order to experience aversion.
If music gives me joy, that joy does not require a linguistic form in order to be significance for the subject.
The subject is already the pole of psychic reality for which this significance is generated.
F therefore does not need a communicative form of representation in order to exist as F.
But if I want to tell another person:
“I love him.”
“I find this smell unpleasant.”
“This music makes me happy.”
then I must represent my psychic content in a form with which another subject can interact.
And here T appears again.
Knowledge Cannot Be Transmitted Directly
The phrase “transmission of knowledge” is convenient, but in a strict sense it is misleading.
If one person says to another:
“It’s raining outside,”
their knowledge of the rain does not move from one psyche into another.
The first subject, through the organism, creates an object in independent reality—a spoken utterance.
This acts upon the organism of the second subject.
And the second subject’s psyche generates its own content.
Communication therefore does not have the structure:
psychic content A → psychic content A.
Rather:
psychic content of subject 1 → represented form in independent reality → interaction of subject 2’s organism with that form → generation of psychic content by subject 2.
This distinction is crucial.
The task of language, then, is not to copy psychic content into another psyche—something that is impossible in physical terms.
Its task is to create objects in independent reality with which another organism can interact, allowing another psyche, with sufficient success, to generate corresponding content for itself.
Why Information Is Structured Differently from Significance for the Subject
Here an interesting asymmetry between F and T becomes visible.
F-content differs primarily in the sign and intensity of significance.
Positive—negative.
Weak—strong.
T-content is structured differently.
The information
“The train departs at 8:17”
is neither positive nor negative and has no intensity in the sense in which love can be strong or aversion weak.
Instead, information requires extremely fine differentiation of meanings.
A train is not a bus.
8:17 is not 8:40.
Today is not tomorrow.
Departing is not arriving.
Berlin is not Hamburg.
If information is intended for the non-subject, such distinctions are crucial. Another subject must be given the possibility of generating knowledge about precisely the relevant objects and relations.
Perhaps this helps explain why T develops a very different kind of differentiation from F.
F differentiates significance for the subject by sign and intensity.
T produces a differentiated system of meanings required for representing knowledge for the non-subject.
And human language is an extraordinarily powerful means of such representation.
But Isn’t Human Language Far Too Ambiguous for Communication?
Here we encounter a strong Chomskyan objection.
If human language emerged primarily for communication, why does it fit the ideal of unambiguous transmission so poorly?
Natural languages are full of ambiguity.
Meaning depends on context.
The same word or sentence can be understood in different ways.
Certain linguistic constructions are difficult to interpret even though, from a purely communicative point of view, they might seemingly have been organized more simply.
Chomsky uses phenomena of this kind to argue that the internal organization of language is governed primarily by the requirements of internal computation rather than communicative efficiency. When the two requirements conflict, he argues, the evidence tends to favor internal computation.
This is a serious argument.
But it presupposes a particular conception of successful communication.
Communication Does Not Require Context-Free Unambiguity
Communication does not require every linguistic sign to have one and only one meaning independently of the situation.
What is required is something different:
In a particular context, the other subject must receive sufficient conditions to generate the corresponding psychic content.
Consider a very simple sentence:
“Put that over there.”
Outside a situation, it tells us almost nothing.
What is “that”?
Where is “there”?
But if two people are standing in the same room, one is holding a cup and pointing at the table, communication may work almost perfectly.
Context, then, is not necessarily a defect of language interfering with ideal transmission.
Perhaps context is part of the communication system itself.
And that would be entirely natural if communication takes place not between two devices copying symbols, but between two psyches, each of which generates its own content.
Morse Code Does Not Solve the Problem
For the same reason, the idea that Morse code represents an ideal form of communication seems peculiar.
Morse code can determine very unambiguously which sign has been transmitted.
It does not determine what the utterance means.
If an ambiguous English, German, or Russian sentence is transmitted in Morse code, it remains ambiguous.
Morse code encodes signs.
It does not eliminate semantics, context, metaphor, irony, implication, or the dependence of meaning on a situation.
Two different questions are being conflated:
Was the signal transmitted correctly?
and
What content does the subject generate in interaction with this signal?
The first problem can be solved technically.
The second belongs to the psyche.
Perhaps Ambiguity Is Not a Defect of Communication
If psychic content cannot be transmitted directly, then ideal communication cannot consist in producing an exact copy of that content in another psyche.
The other subject necessarily generates its own content.
The context-dependence of language may therefore be not evidence of its poor suitability for communication, but a property of an extraordinarily flexible communication system.
The same word can participate in the generation of very different contents because its meaning is specified by other words, by the situation, by the interlocutors’ shared knowledge, by intonation, gesture, and the preceding conversation.
Language does not have to encode independent reality anew in its entirety in every sentence.
It can rely on psychic content that already exists in the participants in communication.
The question can therefore be reversed:
Would it not be more surprising if a communication system between generative psyches were not context-dependent at all?
Chomsky and an Alternative Hypothesis
We thus have two different hypotheses.
In simplified form, Chomsky’s hypothesis can be represented as:
internal combinatorial system → structured thought → externalization → possibility of communication.
The hypothesis proposed here points in the opposite direction:
knowledge → its significance for the non-subject → need for communication → need to represent knowledge in a form accessible to the non-subject → information → development of T.
This does not prove that Chomsky is wrong.
The evolution of language is an extraordinarily difficult research problem, and specific scenarios of its origin inevitably remain, to a considerable degree, hypothetical.
But the proposed model allows us to ask a question that appears differently in the Chomskyan framework:
Why did knowledge need a form of representation at all?
For the subject itself, knowledge already exists.
The new problem arises when that knowledge becomes significant for the non-subject.
Perhaps the Inner Monologue Did Not Come First
Imagine a distant human ancestor who perceives a change in the environment.
This individual knows something that another does not know.
And this knowledge is significant for the other.
At that moment, a fundamentally new problem arises.
Not simply to know.
To make knowledge accessible to the non-subject.
The knowledge itself cannot be transmitted.
The subject can only perform an action in independent reality—a sound, a gesture, a movement, and later a sign, a word, a sentence, a text—which becomes an object of interaction for another organism.
The other subject must generate the knowledge for itself.
If this problem was one of the conditions for the emergence of T, then communication is not a by-product of thinking.
On the contrary:
Thinking in the form of T may have emerged as a psychic function required for communication.
And then something extraordinarily important may have happened.
The form that had emerged to make knowledge accessible to another became available to the subject itself as well.
Information can do more than be communicated.
It can be interpreted.
Compared.
Combined.
Transformed.
Used to generate new information.
What emerged for communication could become a new instrument of the subject’s own psychic activity.
In that case, the internal use of language does not demonstrate its evolutionary primacy.
It may instead be one of the greatest secondary achievements of communication.
Which Came First—the Thought or the Message?
At present, this question cannot be answered conclusively.
Chomsky proposes that the core of language should be understood as an internal system structuring thought, while communication becomes possible through subsequent externalization.
The model proposed here allows the opposite hypothesis.
Perhaps the original problem was not:
“How can I organize my own thought more effectively?”
but:
“How can I make my knowledge accessible to someone who is not me?”
If psychic content cannot be transmitted directly, solving this problem required not a copy of knowledge but a form in which knowledge could be represented.
It is precisely such a form that makes information possible.
Language can therefore be understood not merely as a means of expressing already completed thought, but as the most highly developed way in which knowledge can exist for the non-subject.
And the human ability to use this form for one’s own thinking may have developed later—as a consequence of the fact that knowledge had once acquired a way of existing not only for the subject who generated it.
The famous question about language and thought can then be turned around.
Perhaps we do not speak because we first learned to think.
Perhaps we learned to think in a particular way because, at some point, we needed to tell another person something.