Between Morphemes and Morphs

One more problematic aspect of the form of morphological elements is the question of locating the boundaries between morphemes. Many words are quite straightforward in this respect, but it’s fairly easy to find forms which do not admit of such simple segmentation.

Take, for instance, singular-plural pairs such as “man - men”, “woman – women”, “sheep – sheep”.

It is not possible to isolate any phonemes and describe them as the allomorph of the plural morpheme. But from a grammatical point of view such pairs of words behave just like any other singular-plural pair. Numerous different solutions for this sort of problem have been proposed, most of them sharing the common characteristic that the notion of morpheme becomes more abstract; not only may a morpheme be realized by a number of different allomorphs, but these may include a lack of any form at all (zero allomorphs), as in “sheep”, or even a possibility that an allomorph is not a segment at all, but some more complex construct. Among the possibilities which have been suggested we note two: the plural morpheme in the case of “man” is an instruction to change the vowel.

Such suggestions remain contentious, but in general it’s clear that if the notion of morpheme is to be extended to cover such cases, it’ll need to be made more abstract in some way.

Perhaps the easiest solution to this problem is to distinguish two levels of structure which usually correspond. This correspondence does not always obtain, however. At one level, there will be morphs – the smallest sequences of phonemes which correspond to some meaning – and the other, more abstract level, morphemes, which are meaningful units which need not be straightforward but are realized as a sequence of phonemes, i.e. as a morph. In the simplest case, sequences of morphs can be put in direct correspondence with sequences of morphemes. Other morphs will correspond to two morphemes simultaneously, without the possibility of segmentation

e.g. men (a morph) will correspond to the pair of morphemes {man} and {plural}.

It’s also quite possible that there are morphs corresponding o no morpheme (empty morphs) – examples of this are usually more contentious than other types but in French we find:

aff. Inter

j’ai ai-je

tu as as-tu

il a a-t-il

The first two of these examples invest in a simple way; in the third, however, “t” is inserted. This letter may be described as a morph with no corresponding morpheme. The same is true about the cases when linking r is pronounced when there is not corresponding physical letter to represent it.

So, the precise relationship between morpheme and morph is open to a wide variety of interpretations. This scheme gives us 4 types of words which can be summarized as follows:

The languages of the world differ in the kind of words which predominate. A language such as Chinese, for example, which is sometimes called an isolating language, has predominantly single, free morphemes with no compound or complex words as such, whereas a highly inflected language such as Latin or Greek will clearly use many complex words. English uses all types but compound words are usually restricted to two constituents. German, on the other hand, has words made up of several free forms concatenated to make a single unit.

Поиск по сайту: