• Dutch
  • Frisian
  • Saterfrisian
  • Afrikaans
Show all
Allomorphy
quickinfo

Allomorphy is the name of the phenomenon that a morpheme has different variants, between which there is a complementary distribution, without this variation affecting its meaning. This section gives a short general introduction to the subject.

readmore

Allomorphy is the name of the phenomenon that a morpheme has different variants, between which there is a complementary distribution, without this variation affecting its meaning. It is very common in language. Take the word goed /ɡuəd/ good, which has the variants [ɡuət] and [ɡuəd], as in it komt wol wer goed [ɡuət] it will turn out all right (lit. it comes well again good) and it goede [ɡuədə]antwurd the correct answer (lit. the good (inflected form) answer). Obstruents are voiceless in word-final position in Frisian (see Final Devoicing), so the variation between [ɡuət] and [ɡuədə] fits in a general pattern of alternation. In cases like these allomorphy lends itself to an account in general phonological terms, without reference to the morpheme goed itself.

There are also cases of allomorphy which do not lend themselves to such an account, i.e. which do not involve phonological variation. Take the personal pronoun hy /hi/ he, which has the clitic variant er /ər/, as in hy komt út Fryslân he is from Friesland (lit. he comes out of Friesland) and komt er út Fryslân? is he from Friesland? (lit. comes he out of Friesland?). There is a complementary distribution between hy and er, and they also have a semantic relation, but not a formal one, so both forms must be assumed to be part of the Frisian lexicon.

The clitic er, in turn, has the variant der, with which it is in free variation, although only in specific contexts. Because they share the part /ər/, there is a formal relation between er and der, which should be accounted for. This might be done in a 'dynamic' way, by means of a procedure which augments er with initial /d/ in certain contexts. This procedure, however, would have a very small scope − one morpheme, to be precise −, so it would not express a valid and significant phonological generalization. The relation between er and der therefore is better expressed in a 'static' way, either by a disjunctive lexical representation or a relational statement of some sort.

References
    printreport errorcite