"What" is defined grammatically as an interrogative pronoun

... used interrogatively in asking for the specification of an identity, quantity, quality, etc. (Wiktionary)

In dictionaries, however, what is never defined semantically.

If I meditate on a word like car, I inevitably associate images, feelings and experience with the word. Even if I think of nothingness I somehow associate sensations and a mental image of emptiness (even if it may be wrong). When I contemplate the word who, I associate the word with a "person whose identity is unknown". When I think of what however, in terms of pragmatics, it seems to suggest a request for more information to the interlocutor, but semantically, I cannot find any meaning to it.

Hence this odd question:

What is the semantic meaning of what (or its equivalent in any language)? Could the meaning of what be defined as "a thing unknown to me"?

  • 1
    What do you mean by 'semantic' (and maybe also 'pragmatic')? Sure, you can't point at it. It has lots of semantics, but mostly as a place-holder, a variable, more grammatical function (-in- the language rather than in the world).
    – Mitch
    Commented Jan 18, 2012 at 14:18
  • 1
    In the sentence "what is your name?" the verb "to be" suggests your name = what. What relates to that as a pronoun and its semantic is abstract, standing for some entity the interlocutor and I both know. Yet at the same time what denies the premise of a pronoun, because it expresses the absence of knowledge about that. Can it be said then that what means that which is unknown to me? In which case it would be an pragmatic request for more information. Commented Jan 18, 2012 at 14:39
  • 1
    I don't see the difficulty. Sure, 'what' means something unknown. Why does requiring a verb cause a problem? Do you expect 'What are you called?' really to be 'What you?' ? Is the problem 'synsemantic'? (if so then all verbs are a problem for you). You say "what is never defined semantically". I don't agree; I find the dictionary definition semantic. What's your idea of 'semantic'?
    – Mitch
    Commented Jan 18, 2012 at 14:58
  • 1
    It's still not clear what you mean by 'semantic'. My conception of it includes the definition of 'what'. What is the semantics of any purely grammatical function word? It has them, they're just not so comparable to those you get for nouns or verbs. You say "I cannot find any meaning to it", what do you mean by 'meaning"? Do you need a referent? A sensation? What's the meaning of 'two'? or 'idea'? or 'among'? or 'you'? or 'not'? They all have meaning/semantics, just not as obvious as 'apple' or 'running'.
    – Mitch
    Commented Jan 18, 2012 at 16:26
  • 1
    In the same vein as "who," perhaps "An object or concept the nature of which is unknown to me." Commented Jan 19, 2012 at 3:06

2 Answers 2


I'd say that "what", like other interrogatives, is generally a function word -- synsemantic -- and has little to no intrinsic lexical meaning. Its meaning is always determined by context. So, the semantic meaning you're looking for probably isn't there!

  • I thought that too, on the other hand though, if the common sentence were "What your name?" I'd think what purely is synsemantic, turning the sentence into an interrogative. But sentences are "What is your name?" suggesting what is something. See my point? Commented Jan 18, 2012 at 10:46
  • For instance in Hindi, the equivalent of what, kyā, can be used in two ways: 1. as with English tumhārā nām kyā hai? (what's your name?) and 2. purely synsemantic kya tumhārā nām John hai?" (*is your name John?). In the second case, the interrogative pronoun is functioning as a pronoun, it merely indicates the sentence is a question as with Spanish ¿. Commented Jan 18, 2012 at 10:51
  • 1
    Some function words are deictic, that is their meaning is determined by context. Also, you made a wrong assumption about "What is your name." I speak several languages where the verb to be is absent/not required in that sentence. Not all words of natural language have clearly defined referents. There are words which underwent grammaticalization, that is they almost lost their lexical meaning and now they merely express some grammatical concept.
    – Alex B.
    Commented Jan 18, 2012 at 15:08
  • 4
    What is a variable, like the X in an equation; it can mean anything that isn't an individual human (which would require who), depending on who's saying it to whom in what sentence in which context. It has no fixed meaning, just a set of parameters (non-human) and a set of usages (pseudo-cleft constructions, wh-question constructions, etc.). It's not that it has no meaning, it's that it doesn't have much, and what it means in a given context isn't likely to be the same in a different context.
    – jlawler
    Commented Jan 18, 2012 at 17:03
  • 1
    @Benjamin It may be helpful to consider that in English (many other languages) the interrogative and relative pronouns have identical forms. Informally -- in a sentence like "Who is the person who did that?" both who 's have the same unknown referent but different contextual 'meanings'. Its function determines its meaning from the context. Commented Jan 18, 2012 at 22:24

After all the comments above, there's not much left to say. I didn't really want to post it as an answer but the quotes from Bhat 2004 (since you wanted references) are too long for a comment.

"Interrogative proforms are used primarily for obtaining information from the addressee regarding an unknown entity, and hence the speaker can provide only a general indication of its identity. He is therefore forced to use a general term for referring to it. Even when interrogative pronouns have some of their extended uses like testing the knowledge of the addressee, there would still be a need to use a general term."


"Proforms [...] are general terms that are used for carrying out different functions like (i) identifying the participants of an event by locating them with reference to the spatio-temporal location of the speech act participants (or indicating the location, time, manner, etc. of the event itself in a similar fashion), (ii) referring back (or forward) to other expressions that occur in the utterance or in previous utterances, or (iii) indicating the scope of a question, negation, or exclamation. They are generally made up of two different elements of which one indicates the function of the preform (demonstrative, interrogative, indefinite, or anaphoric (relative) ) and the other one denotes their category or scope (person, thing, place, manner, quality, quantity, etc.)."

"All proforms share a set of general concepts and hence expressing a general concept can be regarded as the first set of their main functions."

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.