We can have two words which describe a similar concept but have different parts of speech, for example live V, life N.

I live well.

Semantically similar or arguably equivalent construction:

My life is good.

Are those two words describing two different concepts or not? Can words with different parts of speech constitute one concept?

This question came to my mind during reading the book Cognitive Exploration of Language and Linguistics by Rene Dirven, Marjolijn Verspoor.

  • Continuation of linguistics.stackexchange.com/questions/28903/…. Sep 21, 2018 at 10:14
  • Original sample from the mentioned book (page 15). "a. Look at that rain ! b. It's raining again. c. And the rain, it raineth every day. In all three sentences we have chosen the same lexical category rain, but it is construed as two different word classes, as a noun in (15a) , as a verb in (15b) and both as a noun and a verb according to Shakespeare in (15c)". For me is a little unclear what authors mean by term lexical category.
    – Adrian
    Sep 27, 2018 at 13:33

1 Answer 1


These words are related by derivation. In each case, one of the parts of speech was first. Then others were created from one or more of its senses, but can then take on a semantic life of their own.

(In English it is very subtle because the surface forms are identical or, in your life N / live V / live ADJ example, very similar and the derived by a process that is no longer productive.)

Because of constant semantic shift, the senses are hardly ever 1:1 even on day one, so it is dangerous to define concepts as if they are logical and orderly, not fuzzy and tangential.

live ADJ is in the same semantic field as television and concert, but in some other languages live in the sense of television is literally direct ether, and literally alive or living is only used for in-person performance as opposed to something remove.

(In English it's a bit hard to make this distinction, so for example in a café full of screens showing the Eurovision final in real time, it's live but not live.)

This hints at us that this sense of live is not very tightly bound to live V, and, in fact, upon thought it's true, show and concert have nothing special to do with live V or life N, nor with lifer or leave, which are also derived from the same root.

Another proof of this is that with enough derivations, we can get back to the same part of speech, yet they are not equivalent. lively is not the same as alive, enliven is not the same as live, liveliness is not the same as life.

So dictionaries like Wiktionary take the safe route by listing a flat representation. (A lemma can have multiple senses, a sense can only have one lemma.)

  • Can you mention 'lexeme' vs 'lemma'?
    – amI
    Sep 21, 2018 at 16:58
  • @aml Out of scope here, ask a separate question if wiki/Lexeme is not sufficient. Sep 21, 2018 at 20:15
  • Scope was that dictionary lemma is narrower than lexeme, but psycholinguistic lemma is broader than lexeme. Probably is worth a question.
    – amI
    Sep 22, 2018 at 1:23
  • I'm confused on how morphological derivation applies in cases where there are no affixes.
    – user21043
    Sep 22, 2018 at 2:07
  • 1
    If no visible affixes exist, sometimes "zero-affixes" (affix "Ø") are used to mark different types of the same surface form as morphologically derived from a base form.
    – peschü
    Sep 23, 2018 at 17:31

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.