Is the reality of zero morphs controversial among linguists? I haven't been able to find a wealth of information online about zero morphs, but did find a definition of them at the SIL Glossary of Linguistic Terms, which read as follows:

"zero morph is a morph, consisting of no phonetic form, that is proposed in some analyses as an allomorph of a morpheme that is ordinarily realized by a morph having some phonetic form. Example (English)"

example: "The plural form that is realized in two sheep is Ø, in contrast with the plural -s in two goats."

Is this analysis of singular "sheep" vs. plural "sheep" accepted by most linguists, or do linguists debate other possibilities, e.g. that the singular and plural forms of sheep happen to be identical, no zero morpheme needed?

  • I guess you could come up with someone proposing a different analysis, but I think it's pretty much accepted.
    – robert
    Commented Aug 5, 2013 at 9:17
  • Pace @robert, linguists are generally wary of positing zero morphemes without good reason. Commented Aug 5, 2013 at 13:11
  • It's too easy to add imaginary affixes or nodes to hang your favorite stuff on, so linguists treat it as a version of Occam's Razor. (The two morphology problems referred to are Amharic and Lamba)
    – jlawler
    Commented Aug 5, 2013 at 18:22
  • 1
    As far as "reality" is concerned, we're talking theories here, so just about everything we're talking about is deficient in some sense of reality. I mean, we can't even talk sensibly about "the English phoneme /e/" without stopping to define terms and look at the whole system as presented by this or that analyst.
    – jlawler
    Commented Aug 5, 2013 at 18:25
  • I just remembered that I discussed varieties of zero morphs in an IJAL review. There were five zero morphs posited in the grammar of Tojolabal, but four of them turned out to be ponenda sine necessitate.
    – jlawler
    Commented Aug 5, 2013 at 18:41

2 Answers 2


It's probably worth making the more general point here that not everyone agrees with a Bloomfieldian 'constructive' conception of morphology, in which words are built-up from sub-word units. There are a few morphologists working within an 'abstractive' word & paradigm based framework (Jim Blevins' terminology), where words, the basic units, are associated with paradigms, e.g. 'dogs' is a basic unit associated with the 'plural' cell of a paradigm. Formatives and inflection etc. are treated as epiphenomenal abstractions over the lexicon. See for example the work of Jim Blevins at Cambridge , and his 2006 paper 'word based morphology' (can find here: http://www.jpblevins.net/) for an intro to this kind of approach.

So to answer your question: The reality of 'zero morphs' is controversial, in the sense that word and paradigm models (which do seem to be in the minority) do not even commit to the reality of any sub-word unit (other than as an abstraction), and this includes zero morphs.

  • What about triple paradigms like 'marked as this/marked as that/unmarked? E.g. Definite/Indefinite/Zero articles in most European languages, or Accusative 2 / Accusative 2 / Partitive in Finnish?
    – Manjusri
    Commented Aug 7, 2013 at 18:42
  • I'm afraid i'm not intimately familiar with this kind of approach personally, aside from having seen a couple of talks, so i'm not really sure what proponents would say. I think that Blevins handles this type of thing by enriching the information in the paradigm such that there are implicational relationships between cells. It's a good question though.
    – P Elliott
    Commented Aug 7, 2013 at 22:25

The idea of zero endings is quite useful in PIE studies.

Consider for example, two words:

nebhos has no ending

nepōt has zero ending

This is because the first one is of Neuter gender while the second one is Masculine.

The role of ending is played by the longer vowel in final syllable.

Also a zero ending can be a case marker. For instance, in Nomenative Russian word папа "father" has ending -a, but in Vocative it has zero ending: пап. This is the only singular case where the word has no ending.

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.