For example, these multi-word expressions or lexical units can be listed in their very basic forms on thefreedictionary.com:

leave one's mark
leave mark on
develop from
develop from someone
fight shy of
fight shy of something
be fighting shy of doing something
be out of one's shell

instead of any of their variations like the following:

leave his mark
left mark on
developing from
developed from a painfully shy girl 
fighting shy of
fight shy of controversy
was fighting shy of cheating
is out of your shell

I know that they are not the idioms(as idioms may include possessive pronouns), stems(as stems include subwords), lemmas(as lemmas also include pronouns), or lexemes(also include pronouns; and 'is' and 'be' are two different lexemes) of their inflections or variations.

I wonder if such a very basic form has a terminology? I want to normalize the variations into their most basic units. Possessive pronouns should become forms like "one's" which cannot be further normalized.

This question was originally posted on ell.stackexchange.com here.

  • Someone removed my comments; First, they are not basic forms. There is a main dictionary entry word and maybe an idiom. When an idiom is given, it is usually given in its most basic form, pronoun-wise. And some are outright inaccurate. "leave mark on" is not really grammatical. "developing from" is just some piece of something larger., None of these are really kosher. The examples you provide don't always make semantic sense.
    – Lambie
    Mar 14, 2023 at 15:33
  • What you want has already been done for each of these expressions. And some are wrong, like: leave mark on was fighting shy of cheating, Some are listed as idioms. You have mixed up idioms with pieces of text with no meaning on their own.
    – Lambie
    Mar 14, 2023 at 15:41

1 Answer 1


You could call this the canonical form or the citation form, or perhaps even extend the sense of lemma to include MWEs. I’m not aware of an equivalent of “lemma” which is specifically inclusive of multi-word expressions.

  • No, this is most definitely not correct. And these are words like develop + idioms associated with it. Some of them are even non-grammatical: leave mark on, for example. We'd use: leave a mark on something.
    – Lambie
    Mar 14, 2023 at 15:35
  • I’m not going to nitpick OP’s typos — I made an assumption about their intent and followed that. I agree certainly that “developed from a painfully shy girl” is not an MWE, in canonical form or otherwise.
    – jogloran
    Mar 14, 2023 at 15:37
  • canonical form is math or computers, not linguistics. The question is not accurate at all. So, any answer has to put that to rights.
    – Lambie
    Mar 14, 2023 at 15:39
  • Check out the Wikipedia page on the linguistics sense of “lemma”. It lists “canonical form” as a synonym of lemma. And linguistics is an umbrella that includes computational linguistics — perhaps modern linguistics is more interdisciplinary than you imagine.
    – jogloran
    Mar 14, 2023 at 15:47
  • Yes, canonical form of a word; not a group of words, unless they are same lexeme. Groups of words that go together are idioms, not lemmas. But as I said, the OP is mistaken about those lists which contains idioms and gibberish. I don't get how it's possible to think the question is accurate when it is not.
    – Lambie
    Mar 14, 2023 at 15:53

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.