I think anyone can deduce many things from and about how we use language (if I know you are a bachelor and the meaning of the word, then I know that you are not married), but can we (get outside our language practices sufficiently) infer inductively about how we are using language, from how we are using language?

the principal method of human communication, consisting of words used in a structured and conventional way and conveyed by speech, writing, or gesture.

Linguistics must not be a purely a priori science, so presumably yes, but are some such inductive inferences impossible? Can I inductively infer anything about the use of the word 'tree' from how 'tree' is used? Presumably: how tree will continue or was used.

Can meaning is use be the conclusion of an inductive argument?

  • A lot of NLP models (including LLMs) are based on contextual use of words and sentences. IMO this is not enough, there are a-linguistic features that need to be accounted for as well..
    – Nikos M.
    Commented May 3 at 17:07
  • I'm not sure what you are asking here. Are you asking whether some innate knowledge is necessary to understand language, or whether it can be inferred without any innate knowledge? Commented May 3 at 19:11
  • i am not asking about innateness @DavidGudeman i'm asking whether languag emeaning is use can be the conclusion of an inductive argument. there is no argument to the contrary here. i just find it strange that we might enumerate cases of meaningful use as if they supported the conclusion. it seems... off
    – andrós
    Commented May 3 at 19:21
  • 1
    Then the/my answer is simply no: "meaning is use" in any case is not the result of inductive reasoning. Commented May 3 at 19:40
  • 1
    Are you using "inductive" broadly, including abduction? If we can demonstrate that talk of meaning can be reduced to roundabout references to use by analyzing sufficiently many cases, and no plausible entity to embody "meaning" is forthcoming, then abduction gets us that meaning is just a manner of speaking about use.
    – Conifold
    Commented May 3 at 21:32

2 Answers 2


I think your question could be addressed in two ways:

  1. Inductivism, the epistemological view that science (and thus linguistics) should rely on induction, is quite obsolete. Karl Popper gave pretty strong arguments about this idea and proposed that science works "the other way around": first, scientists build theories/hypothesis that sound plausible to them; then they make observations and they check whether the observations match what the theory would have predicted. If the observations contradict the theory, the theory has then been falsified, which means that scientists should invent a better theory. Note that although you could sometimes come up with a theory by reasoning inductively, it is not what makes it scientific or not. What makes it scientific (according to Popper) is that the theory allows you to make predictions that could be wrong. Popper's view has been challenged since, but I don't think that inductivism ever made a come-back.

  2. It does not seem that "meaning is use" could be the result of inductive reasoning anyway. Inductive reasoning would require to have an pre-existing idea of what is "meaning" and what is "use", so that one can make a series of observations ("oh, in this case meaning and use were equivalent, oh in this other case also...") from which to infer that meaning is use. It seems more probable to me that "meaning is use" works as a definition of meaning than an empirical statement.

Remark: "meaning is use" is also a sentence meant to summarise late Wittgenstein's view on language, but it does not seem to fit your question.


Can meaning is use be the conclusion of an inductive argument?

(1) "Meaning is use" is a slogan that hints at a particular theory about what constitutes linguistic meaning, about how words and phrases acquire meaning: If you want to know what a word (phrase, sentence) means, then try to find contexts in which it is actually used (either in scientific discourse or in everyday, non-technical communication), and try to find out what the patterns are in the usage (similarities and differences).

This slogan also hints at an implicit criticism of a particular form of platonism were "meanings" are considered to be some kind of separately existing, non-linguistic, ethereal objects. The platonistic theory is misguided because it treats all words (phrases, sentences) as if they were proper names, as if a word could only acquire a meaning by referring to an object, equating all meanings with referents. This form of platonism is obviously nonsense, but it took a while to realize that, and it is still seductive and persistent. For instance, the ghost of it still persists in the misguided belief that the meaning of a word is (or should) always be something that is fully determined or determinable.

(2) Was the idea of "meaning is use" found by inductive reasoning? No. (Could it have been found by induction? No.) Does that imply it was found by deduction? No. By pure but non-deductive reasoning? No. Purely by observation? No, of course not.

In general, we don't acquire new ideas by something like inductive (or deductive) reasoning, but by trying to solve particular problems, looking again at what is in front of us or around us, and using our imagination. This basically involves pattern recognition, learning to see new patterns, learning to see new structures that pattern our observations, that somehow make us see the things we first saw as a new gestalt or in a new light.

This is neither induction, nor deduction, nor just summarizing observations, nor pure reasoning, introspection or intuition. It involves aspects of all of that.

The interesting question here is: What were the original problems that the initial "meaning is use" people were trying to solve? (For this you have to go back to Frege in the 19th century, but also to linguists like Roman Jacobson in the last century.)

You must log in to answer this question.

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