Analysis explores morpheme identification and distributional contrast among similar S morphs, suggesting nuances in meaning.
13.1. D istributional contrast and morpheme identification . There are S morphs of similar meaning which differ in shape in such a way that no alternant is predictable from another, except by a specific list of the words in which each morph occurs. If the morphs are in non-contrastive distribution, they may be classed into the same morpheme. Frequently, however, the distribution contrasts for one of the following reasons: (1) Each morph occurs by itself as a word. The two words may be synonyms, but we assume some difference of meaning. Our data need not show the words in exactly the same environment; existence as words is enough to prove the possibility of a minimal contrast. For example, čí, jí, cučí, and tó are all words meaning ‘earth’; presumably there is some semantic distinction, corresponding to the morphemic difference. (2) There is a word AB contrasting (as above) with a word A′B′, in which A and A′ on the one hand, B and B′ on the other, are morphs semantically similar and in non-contrastive distribution for other environments. It is obvious that we can not class both A and A′ into one morpheme and both B and B′ into another; this would be equivalent to saying that the word AB consists of the same sequence of morphemes as the word A′B′, which in turn would mean that AB and A′B′ are the same word even though in contrastive distribution. For example, kee.tee and kyoo.dai are different words meaning ‘brothers’. It would be possible to take one pair (say A and A′, kee and kyoo 'older brother') as the same morpheme, and the other pair (B and B′, tee and dai 'younger brother') as different morphemes. But the choice would be entirely arbitrary; the burden of distributional contrast falls equally on each pair. The obvious contrast in words of the type AB : : AB' (iči.jicú : : iči.niči 'one day, some day') or AB : : A′B (čo.ŋen̄ : : šo.ŋen̄ 'preface') is considered subsumed under this criterion. There are two further criteria for rejecting morpheme identification of some such S morphs, even when they are in non-contrastive distribution: (3) The meaning differs, slightly but consistently, with the alternation in shape. (4) Although one of the alternants is traditionally associated with a certain graph (and the morph group which it represents), it is possible to associate it with another graph (and morph group) or with some Y morph of similar shape and meaning. In other words, it is sometimes possible to reinterpret the identity of morphs in a different way from the traditional or etymological identification.
No takes yet. Share an insight, caveat, or question.
A 1952 study studied this question.
Synapse has enriched 5 closely related papers on similar clinical questions. Consider them for comparative context: