Distribution (Linguistics)

from Wikipedia, the free encyclopedia

The distribution of an object is generally the distribution of its copies ( tokens ) in the environment or the environments in which it occurs. In linguistics , as in other disciplines, the term can be used in a broad sense, where one examines the distribution of any data in the subject area and thus also of the distribution of languages ​​in an area or the distribution of properties of linguistic units, such as sentence lengths , can speak in corpora .

In a narrower sense, the distribution of a linguistic unit is the totality of the contexts (or contexts) in which it occurs. Linguistic units of all levels and levels of abstraction come into question. a. Sounds and phonemes , morphemes and morphemes , words , constructions , etc. If such a unit occurs in a speech or a text , other units of the same kind are in front of and after or to the left and right of it, which together make up a unit of a higher level of complexity forms. The German morph -eme z. B. occurs after the stems beautiful , old , loud etc., generally after adjective stems , and forms the strong declension form of the dative singular masculine and neuter. It occurs z. B. not on nouns; the form of declination mentioned is from old old , but from day to days . So the distribution of -em is "occurs after adjective stems". This class of contexts is formally represented as follows: [[X] Adj __]. In the place of the X one has to think of any adjective stem; the underscore stands for the position occupied by the unit whose distribution we are talking about.

Paradigmatic relations as distribution relations

If one compares the distributions of two linguistic units A and B with one another, three distributional relationships are of particular relevance, each forming a particular paradigmatic relation :

  • If A and B have the same distribution, but replacing A with B in one context results in a difference in meaning, then their paradigmatic relation is the opposition . Example: In Standard German / d / and / t / stand in opposition, because when they are exchanged in a context like _rüben there is a difference in meaning. Likewise, although and because (as conjunctions at the beginning of subordinate clauses) are in opposition.
  • If A and B have the same distribution, but if A is replaced by B in any context, there is no difference, then their paradigmatic relation is free variation . Example: In Standard German the sounds [R] and [ʁ] are in free variation; and also the words, though and though, are in free variation. (This example also shows that synonymy can be defined as the free variation of meaningful units.)
  • If the distributions of A and B complement each other in such a way that A occurs in all the contexts in which B does not occur, then their paradigmatic relation is the complementary distribution . Example: In Standard German, the sounds [ç] and [χ] are in a complementary distribution, because [χ] only occurs after the back vowels, while [ç] occurs in all other contexts. The morphs -est and -st are also in a complementary distribution after verb stems (possibly inflected after tense), because the former only appears after alveolar plosives , the latter in all other contexts. Sounds with a complementary distribution are grouped as allophones to a phoneme , in the case of the example to a phoneme, which is usually written / x /. Morphs with complementary distribution are grouped together as allomorphs to form a morpheme ; in the case of the example it is the conjugation ending of the second person singular.

Distribution classes

Two linguistic units that have the same distribution belong to the same distribution class . Morphs that have the same distribution as -em include a. -er (as in old ), -e (as in old ) etc. In this case, the suffixes that make up the paradigm of adjective declination belong to a distribution class. (However, this assumes that this morph has been distinguished from its homonyms , which have a completely different distribution.)

Since the distribution of a unit can be objectively checked relatively well, one tries to reconstruct traditional linguistic categories as distribution classes. The part of speech 'adjective' could e.g. B. define it as a distribution class as follows: what in the context [[X] def. Article __ [Y] N ], between a definite article and a noun, so u. a. the __ day occurs in the context . The example also shows the limits of the procedure: First of all, in this way you don't actually find adjectives in German, but rather certain inflected forms of adjectives, which you would have to generalize about again in a controlled manner. Second, one assumes the categories that form the context. So if you wanted to define 'adjective' in this way, you couldn't define 'noun' as something that can come after an adjective. So there is no prospect of analyzing all categories of a certain linguistic level as distribution classes.

Nevertheless, the determination of the distribution of a linguistic unit in structural linguistics is an elementary methodological step in its description. On this basis one then analyzes the function or meaning of the unit.

literature

  • Hockett, Charles F. 1958, A course in modern linguistics . New York: Macmillan.
  • Spang-Hanssen, Henning 1958, "Typological and statistical aspects of distribution as a criterion in a linguistic analysis." Proceedings of the International Congress of Linguists 8: 182-194.

Web links