For example, in the Spanish sentence "Yo era chico y ella era vieja" [I was little and she was old], era appears twice, each time as the same part of speech (a verb) but with different morphology (first person singular vs third person singular).

Similarly, in the English sentence "The glass is made of glass", glass appears twice, each time as the same part of speech (a noun, albeit countable then uncountable) however the two uses have different meanings.

What is the term in linguistics to describe the "hash" or "fingerprint" of each use -- to borrow from computer science -- such that any difference of parameter (POS, morphology, semantics) renders the word distinct?

(In a dictionary, any such difference would result in a separate definition.)


Your second case is covered by the term "sense" (i.e. the two instances of the word "glass" are different senses)

You could make an argument for using the same terminology for the first case too, but it seems a stretch to me, as this syncretism between the 1sg & 3sg imperfect is systematic, meaning one could argue that there is a single sense here that leaves it unspecified whether it is 1sg or 3sg


I'm afraid there is no such term. Different entries in dictionaries are usually named lemma but this leaves out the details of your first example with era (1p sg) vs era (3p sg). A lemma is also given in a conventionalised form. And the level of semantic difference that is sufficient to justify different lemmas may vary from dictionary to dictionary and according to traditions in a speech community—English dictionaries tend to be more fine-grained than German ones in this respect.

  • If so, this is a very surprising omission. The need for such a concept seems obvious to me. In its absence, a frequency dictionary, for example, conflates forms and meanings. – zadrozny Feb 28 at 18:15
  • 2
    @zadrozny not really, because they can just list forms as "era (1p sg)" and "era (3p sg)". The glass example can be handled with two separate lemmas, glass-1 and glass-2. – Keelan Mar 1 at 11:04

Another option is wordform, as opposed to lemma. For example, the CELEX corpus allows you to download lists of lemmas, syllables, or wordforms; if you choose lemmas, write and writes would be grouped together in a single entry, but if you choose wordforms, they would be listed separately.

However, this usually doesn't include distinctions that are never made in the language's morphology. For example, English has no morphological distinction between verbs with first person plural subject and with third person plural subject (we write, they write, we are, they are). So those generally would not be considered separate wordforms, since the language never distinguishes them.


Re: your English example. It really depends on what theory you subscribe to.

In addition to sense (or meaning), another term I've seen used mostly in linguistic research coming from Eastern Europe is a lexico-semantic (or lexico-grammatical) variant - see a screenshot of page 181 from Karpova and Kartashkova (eds.) 2010 below:

enter image description here

The idea is that the difference between glass1 (a container) and glass 2 (a substance) is not only semantic - they also have different paradigms (i.e. they have different word forms, e.g. glass2 is a singulare tantum) and collocate differently.

Juri Apresjan would most likely say we have one word here (in his theory, slovo or vokabula) GLASS that has at least two different lexemes (leksemy), glass1 and glass 2. So, another option is lexeme.

NB: lexeme and word are understood and used differently by different linguists, so it depends on your theory of morphology and semantics. For instance, I myself use the term lexeme very differently from Apresjan.

Re: your Spanish example.

It is very different from your English example.

Here, the same word form era is used to express different grammatical meanings, in this case 1 and 3 person singular in the past. It's usually called syncretism in linguistic morphology. The most important thing about your Spanish example is that the word form era 1sg.PAST and era 3sg.PAST are members of the same paradigm, unlike glass1 and glass2 in your English example. They belong to the same lexico-semantic variant and obviously they do not differ syntagmatically.

Summary: what your Spanish and English examples illustrate are two completely different phenomena, i.e. grammatical variants (or word forms) vs. lexico-grammatical variants (different senses of the word).

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

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