Is there a term for e.g. the lexical symbol "duck"? It is both a verb and a noun, in contemporary use having no apparent connection, and so would appear to be represent two words.

Then, is the a better term than 'lexical symbol' to describe any grouping of letters having some meaning? I just made up lexical symbol because it seemed convenient.

  • 4
    On the contrary, I would myself appreciate a name for a word that can be only a single POS. Get the idea?
    – Kris
    Feb 19, 2012 at 14:19
  • @Kris, yes, your point will probably keep me awake tonight.
    – ProfK
    Feb 19, 2012 at 14:30
  • "a group of letters having some meaning" is what we now call a "word" -- why another name for that? There's a difference between the title and the implication of the body.
    – Kris
    Feb 19, 2012 at 14:35
  • 2
    Because words don't have letters. Words have sounds. Letters are only for writing, but illiterate people still have words. And even if it were true, it begs the question of "having some meaning". Meaning is not quantifiable like gasoline. Feb 19, 2012 at 17:42
  • 1
    Kris would like a term for a word that can be only a single part of speech - unlike "like," "term," "word," "can," or "single." Hmmm... I try to pen a sentence / Free from homonyms / But when I count, it's just too wild / How many I've left in... (I'd like to know that term for "unhomophonic" as well)
    – J.R.
    Feb 19, 2012 at 22:50

4 Answers 4


At http://www.thefreedictionary.com/duck, both the AHDEL and Collins claim that there are four English words 'duck'. They distinguish them by right-superscripts (which device is restricted to such analyses, of course).

The one they both choose to label duck³ (a heavy cotton fabric) is obviously not related to the other three – this is a case of convergent evolution, from different sources. Different etymologies.

Although words 1, 2 and 4 are obviously etymologically related to each other (word 4 in a punning way), they are still considered to have diverged sufficiently to be classed as three further isoformal words (same spelling; these also have the same pronunciation) - homonyms (strict definition).

However, if we delve deeper, and look, for instance, at AHDEL's treatment of duck³, we see that it lists two (closely related) senses (admittedly one existing only in the plural form and the other almost invariably in the singular) – ONE the material and TWO clothing (usually trousers) made from that material. These senses are classed as not being separate words. Different senses of the same word are known as polysemes. An obvious example: to play football, you need a football.

Sadly, I have not come across a consensus on whether polysemy is a term that is allowed to be applied to isoformal / homographic orthographic words of different word-classes (eg house (n) and house (v); round (preposition) and round (adj) etc).

Oh, and in answer to the second question – an orthographic word is 'a meaningful (within the parameters of the language being used) string of letters bounded by spaces'. The term lexeme covers 'families' of 'the same word' , so man = man & men; go = go, goes ...

Edwin Ashworth

  • Nice answer. Some grammars regard polysemy as necessarily a feature of the same lexeme. So CaGEL, for example and even Aarts (if my memory serves me correctly) would regard lexemes of different categories (as opposed to ones which may or may not be of different categories) as not being polysemes but homonyms. So we get descriptions such as we would regard this as a case of homonymy not polysemy, becasue X1 is a determiner and X2 is an adjective, for example. Mar 14, 2016 at 23:10
  • @Araucaria I added an answer (and got told off for giving two 'answers') below when I found the term 'intercategorial polysemy' months later. But as you say, it depends on how 'word' and 'lexeme' (which Crystal introduced to try to clear up the confusion!) are defined. Mar 14, 2016 at 23:15
  • Can I cite you on the Crystal? I never knew that. UCL, seems to have been a hotbed for stuff like that. It never, ever ceases to amaze me. Mar 14, 2016 at 23:20
  • Am about to go AFTK (newly learned newspeak), but if if you have any nice refs for either the Crystal or the intercategorial polysemes, I'd be very grateful. (can't find much doing cursory IP search on google) Mar 14, 2016 at 23:24
  • Quote me? Certainly not. Andrew Moore's teaching resource site has: << Words and lexemes As a lexical unit may contain more than one word, David Crystal has coined the term lexeme. This is usually a single word, but may be a phrase in which the meaning belongs to the whole rather than its parts, as in verb phrases tune in, turn on, drop out or noun phrase[s] (a) cock up, [ship of the desert].>> You can find references for "intercategorial polysemy" in a Google search. Mar 14, 2016 at 23:27

I've come across the term "intercategorial polysemy" used in lexicography and cognitive linguistics for the form of polysemy where the same orthographic word (and with the same etymology - not a homonym) is used in different word-class usages.

Thus bank (your money) and bank (where you bank it) ... but not bank (a steep natural incline).

Or bank (a steep natural incline) and bank (to border or protect with a ridge or embankment) ... but not bank (where you put your money).

This would make lecture (n) and lecture (v) say intercategorial polysemes.

  • Did you forget that you already answered almost exactly this?
    – Mitch
    Jan 8, 2013 at 0:23
  • Intercategorial polysemy is a compound. Are you saying that wolf and werewolf don't warrant different entries in a dictionary? And it does look like it is the correct answer, after all. Jan 8, 2013 at 14:09
  • You have multiple much more distinct words in your first answer. Adding a modifier to one of them doesn't warrant a full separate entry. Unless of course you want to do the work to separate your other answer. Just move this to the 'polysemy' paragraph in your other answer.
    – Mitch
    Jan 8, 2013 at 15:49
  • @Mitch I couldn't disagree more. Mar 14, 2016 at 23:16
  • Do you have a source I could use in relation to this? Mar 14, 2016 at 23:16

Homophony - words sound the same but have different meanings Homography - words are written the same but have different meanings Polysemy - more than one sense to a word or sentence

"I'll not be the first president of the United States to lose the war" Is polysemic, has 2 meanings, even though there's no homophony of homography

"Fan" Homography - a ventilator or an admirer?

Sorry I can't give you and example of homophony, I only have exemples in Portuguese.

Whenever a signifier has more than one possible sense, it is polysemic (so it would apply to homophony and homography), but polysemy doesn't require homophony and homography to happen

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.