I don't know whether this is the right place to ask this question. I have always wondered whether a language is needed to think. When I actively think, I automatically use my mother tongue to do so, and this will probably be the same for most people.

I know that people who cannot hear or speak are able to think, but how exactly?

Do we really need a language to think?

  • 2
    $\begingroup$ If you could edit your question and frame it in a world building context it could be a very good question. E.G: In designing a population for my world, do they need a language to form complex thoughts? What if the population was deaf, or mute? $\endgroup$ Oct 4, 2016 at 1:55
  • 2
    $\begingroup$ Have you done any research on the topic? Putting your title, "Do we need a language to think?" into Google gives quite a lot of content on the question. Most agree that the question is heavily under specified if asked the way to asked it. $\endgroup$
    – Cort Ammon
    Oct 4, 2016 at 3:24
  • 2
    $\begingroup$ As is, this feels more like a philosophy or cognitive science question, rather than a worldbuilding one. $\endgroup$
    – Frostfyre
    Oct 4, 2016 at 3:35
  • 3
    $\begingroup$ Have you ever had a concept you had to struggle to put into words? I know I have. So, no. (it's not terribly world-buildy, as questions go, though I've read multiple stories that play with the concept) $\endgroup$
    – The Nate
    Oct 4, 2016 at 5:18
  • 1
    $\begingroup$ Absolutely not! this is similar to can a car drive with a map? The car requires a driver just like thinking is associated with memories and experience since fetus and map helps to negotiate turns etc just like a language helps you to convey message to others. However your level of proficiency or being multilingual depends on the type of car and the condition of the road. $\endgroup$
    – user6760
    Oct 4, 2016 at 5:33

3 Answers 3


This is your friendly neighborhood cognitive science fanatic - this question is poorly-worded, because it is asking about thought in general, when it is fairly obvious to most observers that animals think. What is meant, more likely, is to ask whether human thought is possible without language. This is a variation of the Sapir-Whorf Hypothesis, which posits our language affects how we think about things - e.g., if we have more words for shades of blue, especially ones thay conceptualize them as totally different colors, then we will be more apt to differentiate between shades of blue.

In this sense, it is difficult to deny that human language influences thought. Look up Japanese street lights; color naming differences between Japanese and English have led to a fun postwar phenomenon in that the color of their "Go" light on traffic lights - normally green as an American would think of it - is in fact rather more blue...

But all that aside, if one has no language, can one have concepts equivalent to those of a linguistically-gifted individual? This is a more difficult and freighted question. As others have intimated, Pinker believed an underlying "mentalese" was built in to the human brain, providing a structure - a computational architecture, if we feel like pleasing the Functionalists - on which other languages were built and to which a mind fell back if no natural language was known. Problematically, we are left to determine what exactly the benefit of layering natural language over mentalese has for us, but perhaps it is an extension of the need to express thought externally which gives rise to the otherwise superfluous languages.

The alternative is to suppose that there is no language or protolanguage at all in the human mind, and that our language is fully a synthetic creation, which would mean thoughts prior to language learning would be by definition without concept - simply the apprehension of raw stimulus by a conscious mind, as perhaps a mouse might. This has its own problems, one of which is it doesn't explain how humans are so universally good at language.

I tend to fall into the prior camp. Part of this has to do with how difficult it ends up being to rigorously define "language" - much as we may conceive of nigh any set as a system, we may conceive of any system which we derive information from to be a language. It therefore seems likely that on some level, we are always thinking using some sort of "language."

I will close, however, on a quote from Helen Keller that has always fascinated me. She said of her preliterate life that it was:

"...a conscious time of nothingness. I did not know that I knew aught or that I lived or acted. I had neither will nor intellect. . . I had no power of thought."

  • 1
    $\begingroup$ Asking someone who learned language later in life is probably the definitive answer! $\endgroup$
    – JDługosz
    Oct 4, 2016 at 7:35
  • $\begingroup$ @JDługosz, at first glance, I would agree. However, because language is poorly defined, and because Helen Keller does in a sense contradict herself within that quote alone, I'm not 100% convinced by her claim. Still, a fascinating insight into the quality of preliterate consciousness nonetheless. $\endgroup$
    – Adam Wykes
    Oct 4, 2016 at 13:33
  • $\begingroup$ Also, just being able to communicate with other human beings offered many thousands of times more things to think about for Helen Keller. Thoughts go from "I'm hungry, I'm tired", because those are the only thoughts that all people conceive of independently to complicated thoughts about ideas and the outside world. $\endgroup$ Oct 4, 2016 at 15:00
  • $\begingroup$ Eh, I don't know about that. I'm sure she felt the full range of emotions. $\endgroup$
    – Adam Wykes
    Oct 4, 2016 at 15:02
  • $\begingroup$ Testosterone and estrogen both influence thought, yet neither are required for it. (beyond the minima to produce a functioning body, I suppose) The correlation is backwards. We have language centers that, when we think, attempt to translate ideas into language. (What's that model called, oh, cog.sci. wonk?) $\endgroup$
    – The Nate
    Oct 4, 2016 at 19:43

It depends on whether you're talking about basic cognition or full-on human-level thought. Animals for instance, can think without language, although their thoughts are of course much less complicated than human thought. On the other hand, children who are neglected and never learn to speak are severely mentally handicapped for the rest of their lives, not just in speaking, but in all areas of intelligence. This could indicate that language is important for cognition.

Now, neither of these are really very convincing. But my opinion is that the answer to your question is, no. Anecdotally, athletes don't think to themselves in words when they make split-second decisions on the field. Nor do great musicians or mathematicians.

Again, that's not a fully convincing example, so let me try a philosophical tack. Can language really be the fundamental unit of thought? That is, when a thought first pops into your mind, does it come pre-formed with words assembled into sentences and paragraphs? If this is the case, how would people be able to think of things that they do not have words to express, such as complicated 3-dimensional reasoning, or music? And isn't human language just a fundamentally clumsy way to express thought? How often do we struggle to say what we really mean?

As an analogy, can you imagine building a computer that "thinks" in English, or even a programming language? Sure, a computer can transcribe its thoughts into a high-level language. But the thoughts, when they are first thought, come as a string of 0s and 1s, and are only later turned into something that can be expressed. Noam Chomsky believes that humans do a similar thing, that humans think in a universal "language of thought," which is only later transcribed into spoken languages like English.

Steven Pinker goes in depth into this in The Stuff of Thought, and also The Language Instinct, of which I highly recommend the latter.


Nope. Absolutely not.

Proof - Answer the following:

Do babies think?

Answer: Yes.

Do babies know language?

Answer: Nope.

Q.E.D, language is not needed to think.

  • 3
    $\begingroup$ So without language one could only think as well as a baby. $\endgroup$
    – JDługosz
    Oct 4, 2016 at 2:49
  • $\begingroup$ That becomes much, much harder to test, simply because we do not spend time developing ways to learn high-level knowledge without language. Do you know how to learn math without language? I don't... is it possible? I have no idea! $\endgroup$
    – Nelson
    Oct 4, 2016 at 2:51
  • 1
    $\begingroup$ You're assuming the answer to the question of whether babies think. Of course this depends on exactly what you mean by 'think' - after all, by some definitions C. elegans thinks - but in human context it would seem that thinking includes forming memories, and I don't think anyone has been shown to have real memories of things that happened to them before the age of two or so, by which time they've acquired the rudiments of language. $\endgroup$
    – jamesqf
    Oct 4, 2016 at 4:33
  • 1
    $\begingroup$ I’m pointing out that your QED only shows that babies think as well as babies without language and you might actually use this as an argument for the opposite, that cognition is very limited until language is acquired. $\endgroup$
    – JDługosz
    Oct 4, 2016 at 7:22
  • 1
    $\begingroup$ @JDługosz I see what you're pointing out, and my counter argument is that it doesn't matter with regards to the question because all the question asks is "can they think" - at what level they think makes no difference. $\endgroup$
    – Aify
    Oct 4, 2016 at 7:34

Not the answer you're looking for? Browse other questions tagged .