I have spent a lot of time thinking about language and how humans may have evolved language, and keep going further and further back into the depths of what may have been known at a certain time, and how they could have "evolved" or "figured out" the next step in the evolution to spoken language. I try and make some assumptions on what might be present, then see what I can do and discover from that point, then see if it goes from basic sound making, to sound sequencing, to word making, and to mapping sound words to mental concepts.

But up through now I have always assumed that proto-humans would have to have had the concepts all figured out before they created a mapping of the concepts to speech and eventually to written symbols like words (English/Greek/Latin) or logograms (Chinese) or others. In thinking about it some, it makes some sense that you would first have concepts of things like "tree" or "sun" or "sleep", and then after somehow speech capabilities evolve, map the sounds "s-u-n" to the concept of sun (using this for illustrative purposes, I'm aware English is a modern language derived out of many prior languages since the origin of civilization).

So that's the key, concepts needed to be distilled in the minds of the proto-humans, before they mapped them to sounds or speech/words. That is, they had in their minds concepts without words, and then mapped them to words.

That is hard to grok/grasp/conceptualize.

How could you store concepts clearly in your head enough to one day say "sun" and label your mental concept with the word sun?

At first you may just point to the sun:

And in your head you have some sort of clear visual of the sun being this bright roundish object in the sky. But again, you don't have any terms/words for this yet (so I currently imagine). You just somehow have a clear sense that what you are pointing at is a distinct concept.

And so before speech, you would repeat this process for let's say 4,000 different things, some things being objects like "tree" or "sun", others being actions like "sleep" or "eat", others being features like "red" or "far", all which (from what I feel in my psyche/being) are clearly distinguishable concepts which you can imagine in your mind, without words, just by relating entities in your head and doing something "before you give it a name".

My question is, what are the types of things that can be done before giving the concept a name?? How can you have a clear concept without a word for it? I don't see how this is possible currently.

What I imagine is like an Artifical Intelligence / Machine Learning sort of statistical neural network thing (like our brains), with billions of nodes and billions or trillions of connections. This can discover patterns (at what seems to be a blurry level), but yeah, they are not distilled into a concept. Instead, they are somewhat unclearly defined. The fact that we can recognize a piece of fruit doesn't necessarily mean we have a clear concept for it, it just means we might have a pattern recognizer for it.

I was reading recently about how Chimpanzees make "nests" every night for sleeping, and know which branches are the best for bed making. Do they have a mental concept for "nest", and for "branch", and for "branch strength" and "bendability" and such? Or are they just going off machine learning / pattern recognition? Many animals do complex things, but is it all just statistical pattern recognition, or do they have distilled concepts which are easily keyable and recallable as some sort of discrete symbol? That is, do they have each concept mapped to some sort of discrete symbol in their mind, which is somehow not a word?

I don't get how that would look or work....

How can you have a distilled, crystal clear sense of a concept, without words or speech?

If you could explain how this might work, that would be extremely helpful. For example, if many animals could have crystal clear concepts mapped to some sort of discrete, easily recallable thing like a word, then it would be relatively easy to figure out how to go from discrete concepts to mapping concepts to sounds and ultimately words! (Relatively easy, as opposed to somehow going from statistical machine learning to full-fledged speech in one jump).

What I would like to try and see is, if it's possible to store let's say 4,000 distinct concepts in your brain, without words. (4000 is an arbitrary number, which I got from making a conlang and trying to find a memorizable yet relatively large number of words which cover everything you might speak or think about, based off English and Chinese and a few other languages). I have read that Chimpanzees have been taught roughly 400 words. Maybe at first, proto-humans had a small vocabulary of say 400 (not 4k) concepts, which were all statistical probability distributions, not crystal clear symbols. Then they mapped those to sounds. Then by doing so, it expanded their capabilities to learn 4k words (which is like a fluent speaker today, roughly speaking, if you take words to mean "root" words). So all of a sudden, because of inventing the technology of "symbolic words", they expanded their mental capabilities, so to speak.

So in that sense, maybe it was only after inventing a way to easily navigate concepts through symbolic words (through speech/sound-sequences we roughly equate to words today) that we could build a bigger mental "concept base". Before having mental/speech words (and hence having clear, easily recallable, discrete symbols for concepts) maybe it was too hard to pinpoint your concepts. You just "felt a crazy feeling in your body like [recall some memory you had]" when you feel angry, or "felt a crazy feeling in your body like the last time you went near a cliff" when you felt scared of heights, etc... These sort of "statistical" feelings which are "vague concept maps". These vague concept maps seem like they would be hard to keep track of and mentally navigate and pluck out and select, and "think of", without words.

So to reiterate, how can you have clear, discrete, mental concepts, without words, so they are quickly callable and easy to remember, "think of", and visualize. What even is "thinking" without words? I can recall memories of my experiences by thinking, but can I isolate and collect 4,000 concepts without having words? Or would that be basically impossible without words, and hence I would be limited to just recalling, like Chimpanzees, a few hundred instead?

I ask partly because, I wonder did creation myths come before there was language? Like, could you have all the concepts in place, and possibly share them with others, without spoken language? Maybe it's only after spoken language that the mental capacities grew enough to be able to start modeling more and more things, and this would mean that most animals have a relatively small "discrete concept database" which they can easily use and "think about". I dunno.

  • Maybe useful Harald Haarmann, Weltgeschichte der Sprachen. Von der Frühzeit des Menschen bis zur Gegenwart [World history of languages. From the early period of humans up to the present] (2006). Mar 13, 2023 at 12:23
  • 1
    Suggested improvement to your writing style. Don't use the "thing-A/thing-B/thing-C" idiom. Pick the right word and use it by itself. Making your reader follow you through struggling to get the right word is annoying. The reader must stop and consider each word, try to grasp if you really are using them the same way, then try to guess what meaning you intended to be gleaned from the intersection.
    – Boba Fit
    Mar 13, 2023 at 13:28
  • The central assumption of this question, that having a word for a concept makes it clearer or easier to conceptualize strikes me as unmotivated and implausible. A word is just a tag you attach to a concept. In fact I would posit that most of the concepts you use in everyday thought don't have names. Mar 13, 2023 at 15:25

1 Answer 1


Have you examined your presumptions about 'concepts' (SEP)? For instance, when you speak about them developing and then being mapped to words, you seem to take the view they are discrete, abstract objects that can be neatly mapped. In fact, I'm not sure how that can be shown to be true. What exactly do you believe a concept to be? What exactly are your beliefs about mental representations (SEP)? Do you subscribe to LoT (SEP)?

Personally, I'm skeptical that 'concept' is anything more than a short hand to understand word use. You certainly CANNOT put a concept under a microscope and examine it. There are no ways of doing neuroimaging to prove they exist as processes. The word 'concept' seems to be a dummy word we use to insist that somehow words are processed in nice discrete units of the mind. I approach philosophy of language from a naturalized epistemology, and when I reach for my psychology of language textbook (GB), 'concept' doesn't have an entry in the index. How are you so sure 'concepts' exist? I can hear a word, write it on paper, and observe their use. But that can't be said about 'concepts'. (For the record I find the LoT hypothesis ridiculous given what science knows about neural architecture.) From SEP's Concepts:

Can a creature have a belief if it does not have the concept of belief? It seems to me it cannot, and for this reason. Someone cannot have a belief unless he understands the possibility of being mistaken, and this requires grasping the contrast between truth and error—true belief and false belief. But this contrast, I have argued, can emerge only in the context of interpretation, which alone forces us to the idea of an objective, public truth. (Davidson 1975, p. 170).

That is, language determines concepts, not the other way around according to Davidson and Dummett. I hope I've cast some doubt on the metaphysical speculation that concepts are necessarily for mental representations that entail thought. Personally, I see 'concepts' as being in the same category as 'elan vital'. After all, we once thought Aristotle's categories of causality were "real", and have since abandoned them. Perhaps we should be doing the same with classical notions of 'concept'. (I could very well be wrong but am encouraged that I reached my conclusions independently of Dummett and Davidson.)

As for the Origins of Human Communications (GEB) I wouldn't reinvent the wheel. Tomasello's account includes ontogenetic, phylogenetic, and grammatical analyses after covering shared intentionality as an evolutionary strategy, and unless you have his experience and empirical background, you're just engaging in metaphysical speculation. As you have noted, Great Apes can learn signs, but not grammars, and that sort of science informs Tomasello's views. It would be a shame not to at least consider his theory before trying to invent one of your own.

I ask partly because, I wonder did creation myths come before there was language?... this would mean that most animals have a relatively small "discrete concept database" which they can easily use and "think about".

So, let's make clear here and now that both language of thought and the information processing model of thought are METAPHORS, not literal truths about how the brain functions. The literal operation of the brain, including explanations of how Wernicke's and Broca's areas are indubitably fleshed out in connectionist models of natural language generation and don't remotely resemble anything like Montague grammars or large language models. There is no actual language of thought that can be mapped one-one to utterances or databases or abstract datatypes of the mind.

So it bears reiteration from SEP's Concepts:

Some philosophers maintain that possession of natural language is necessary for having any concepts (Brandom 1994, Davidson 1975, Dummett 1993) and that the tight connection between the two can be established on a priori grounds.

  • This is great, thank you for debunking "concept". However, what is in place of concept? What is beyond machine learning so to speak? Is that in that Origins of Human Communications (GEB) book?
    – Lance
    Mar 13, 2023 at 16:50
  • On the physicalist side, it's all about -emes as units of meaning. Morphemes and lexemes. I'm not rejecting 'concepts' as a concept, because I don't think we can have meaningful discourse in semantics without concept. What I'm arguing is that we can talk concretely about semantic units and grounding in terms of public language first to explain the concept of 'concept'. That's a lot to wrap a mind around because our mind intuits the existence of the concept, and that has to be philosophically deconstructed...
    – J D
    Mar 13, 2023 at 16:57
  • Ostensive definition, which is the act of using gestures and sounds to build references, for instance is how we go about acquiring meaning as children. I can utter /cat/ and point in the direction of a cat, and there's no great mystery. But when a German utters /kotseh/ (G. Katze), then all of the sudden, we have a philosophical problem. There clearly must be something in common psychologically? A 'concept' is the easy way out.
    – J D
    Mar 13, 2023 at 17:05
  • What I'm suggesting is that propositions are fundamental units of mental representation, not brain operation, and they're an ad hoc explanatory device to show how brains can have different language for the same experience. In reality, brains map two different utterances to the same units of experience which arise from neural computations.
    – J D
    Mar 13, 2023 at 17:06
  • 1
    @Lance I just saw in your bio you took a class with Lakoff. That's awesome. Major influence on my thinking. I grok his embodied realism. I defend conceptual metaphor vigorously.
    – J D
    Mar 14, 2023 at 8:23

You must log in to answer this question.

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