Typology

Typology is the systematic study of the ways in which languages are similar to and differ from one another. More specifically, the study of language universals concerns those properties that are common to all languages (see LINGUISTIC UNIVERSALS), whereas typology is concerned with the systematic variation across languages, although in practice both studies must be carried out simultaneously and are often subsumed under either the term typology or the term language universals research. As such, the subject matter of the typological approach to language does not differ from that of other approaches, such as GENERATIVE GRAMMAR (as illustrated in Haegeman 1997). Methodologically, however, the two approaches tend to differ in the extent to which each is data driven (typology) versus theory driven (generative grammar), although in recent years the two approaches have tended alternately to approach and recede from one another. Given the rate at which languages are dying out -- an estimated 90 percent of the world's languages will be extinct or moribund by the end of the twenty-first century (Hale et al. 1992) -- the data-driven typological approach assumes increased importance as much of the very data on which linguistic theorization is based is disappearing before us. Overviews of the field of typology are provided in Comrie (1989), Croft (1990), and Whaley (1997).

The relation between language universals and language typology can be seen in two parameters that are often used to distinguish different kinds of language universals. First, language universals can be absolute, that is, exceptionless, or tendencies, that is, occurring with a frequency that cannot plausibly be attributed to chance or other external factors (such as demographics). An example of an absolute universal is that all languages have consonants. (Examples will necessarily be somewhat banal, to avoid having to go into too much detail.) An example of a universal tendency is that nearly all languages have nasal consonants, although a few -- 3.2 percent of the languages in the sample used by Maddieson (1984: 60) -- do not. In practice, it is often impossible to tell whether a universal is an absolute or rather a very strong tendency to which exceptions happen not to have been found. A second distinction is that between implicational and nonimplicational universals. The two universals previously cited are nonimplicational, inasmuch as each refers to only one linguistic property without relating it to other properties. An implicational universal, by contrast, has the structure "if p, then q," where p and q are two linguistic properties. An example of an absolute implicational universal is: If a language has distinct reflexive pronouns in the non-third person, then it also has distinct reflexive pronouns in the third person. "If p, then q" is to be interpreted strictly as material implication, that is, three possibilities are allowed: "p & q," "~p & ~q," "~p & q," while one is disallowed: *"p & ~q." Thus, in the example cited, there are languages like English with distinct reflexive pronouns in the non-third person (e.g., me versus myself) and in the third person (e.g., him versus himself), languages with no reflexive pronouns at all (e.g., Old English me "me, myself," hine "him, himself"), languages with reflexive pronouns only in the third person (e.g., French me "me, myself," but le "him," se "himself"), but no languages with distinct reflexives only in the non-third person. An example of an implicational tendency is that languages with verb-initial word order in the clause nearly always have prepositions ("if verb-initial, then prepositional"), although a handful of verb-initial languages have postpositions, such as Yagua, spoken in Peruvian Amazonia. Implicational universals provide a typology of languages by dividing languages into the three types allowed by the universal, plus, in the case of an implicational tendency, the rare fourth type.

Although typology has a history going back to the eighteenth century (Greenberg 1974), a major impetus to the modern study of typology was the volume Greenberg (1966b), in particular Greenberg's own contribution to that volume (Greenberg 1966a). In the early work inspired by Greenberg's model, the emphasis was primarily and explicitly on the empirical side of typological work, attempting to find out what nonimplicational universals could plausibly be put forward, and to find out what correlations among different linguistic features might plausibly serve as the basis of implicational universals. More recently -- for a survey of recent approaches to typology and universals, Shibatani and Bynon (1995) may be consulted -- the importance of explaining language universals has come increasingly to the fore (e.g., Butterworth, Comrie, and Dahl 1984; Hawkins 1988). A number of different kinds of explanations have been proposed.

As suggested also by generative grammar, some universals probably reflect innate properties of the human cognitive apparatus. For instance, it has been observed that syntactic properties are, with only a few exceptions that can probably be accounted for in other ways, "structure-dependent" (see SYNTAX). By this is meant that they require identification of elements of syntactic structure for their operation. In English, for instance, questions can be formed from corresponding statements by inverting the order of subject and auxiliary verb, as with can the new professor speak Polish? in relation to the new professor can speak Polish. A priori, a much simpler rule would be to invert the first two words of the sentence, or to put the words of the sentence in the inverse order, yet human languages invariably or almost invariably go for structure-dependent rules like the English rule just discussed, which involve sophisticated parsing of syntactic structure. Given that there is no aprioristic reason for this preference, it almost certainly reflects an innate constraint. (This leaves open, incidentally, whether the constraint relates specifically to the language faculty or whether it is a more general cognitive constraint. At least in this case, the linguistic constraint is probably a special case of the general pattern-seeking preference of human cognition, for which arbitrary strings are difficult to handle, while imposition of structure facilitates processing, as seen trivially in the breakdown of seven-digit telephone numbers into groups of three plus four.)

Processing considerations seem to be a major factor constraining cross-linguistic variation. An example from the early generative literature is the difficulty of processing self-embedded constructions, for instance where a relative clause is included internally within another relative clause, as in the boy [that the man [that I saw] caught] took the apple, although a slight change of construction to avoid the self-embedding produces a readily interpretable sentence: the apple was taken by the boy [that was caught by the man [that was seen by me]]. A detailed theory relating word order universals (both absolute and tendencies) to processing constraints is developed by Hawkins (1994).

This leads into the area of functional motivations for language universals, including not only the needs of processing but also other considerations from SEMANTICS and PRAGMATICS. For instance, the universal stated above that "if a language has distinct reflexive pronouns in the non-third person, then it also has distinct reflexive pronouns in the third person" does not have any obvious formal explanation; this universal is no simpler in formal terms than its empirically incorrect opposite "if a language has distinct reflexive pronouns in the third person, then it also has distinct reflexive pronouns in the non-third person." But as soon as one starts thinking about the semantic function of pronouns, a plausible explanation emerges. First- and second-person pronouns are uniquely determined by the speech situation, with the speaker referred to in the first person and the hearer in the second person. They thus do not change within an utterance, and whether a language says I hit myself or I hit me does not affect the content, that is, marking reflexivity is in a sense redundant in the first and second persons. Third person pronouns can potentially refer to any other entity in the universe of DISCOURSE, so it is useful to have different forms that enable distinctions among potential referents to be maintained, as in the case of he_i hit himself_i versus he_i hit him_j. Quite generally, as predicted, pronoun systems tend to make more referential distinctions in the third person than in the other persons, as when English distinguishes gender in the third person singular (he, she, it) but not in the first person (I) or the second person (you). Comrie (1984) shows how certain language universals can plausibly be related to pragmatics, for instance universals of imperative formation to the pragmatic function of imperatives in encoding the speech act of directive (Searle 1969): because directives require that the addressee carry out a certain action, many languages have a constraint that only imperatives with the addressee as agent are possible, that is, they allow the equivalent of eat the bread! but not of be eaten by the lion!; no language shows the inverse pattern.

Additional links

-- Bernard Comrie

References

Baker, M. (1987). Incorporation. Chicago: University of Chicago Press.

Butterworth, B., B. Comrie, and Ö. Dahl, Eds. (1984). Explanations for Language Universals. Berlin: Mouton.

Comrie, B. (1984). Form and function in explaining language universals. In B. Butterworth, B. Comrie, and Ö. Dahl, Eds., Explanations for Language Universals. Berlin: Mouton, pp. 87-103.

Comrie, B. (1989). Language Universals and Linguistic Typology. 2nd ed. Chicago: University of Chicago Press.

Croft, W. (1990). Typology and Universals. Cambridge: Cambridge University Press.

Greenberg, J. H., Ed. (1966a). Universals of Language. 2nd ed. Cambridge, MA: MIT Press.

Greenberg, J. H. (1966b). Some universals of grammar with particular reference to the order of meaningful elements. In J. H. Greenberg, Ed., Universals of Language. 2nd ed. Cambridge, MA: MIT Press.

Greenberg, J. H. (1974). Language Typology: A Historical and Analytic Overview. The Hague: Mouton.

Haegeman, L. (1997). The New Comparative Syntax. London: Longman.

Hale, K., M. Krauss, and L. J. Watahomijie. (1992). Endangered languages. Language 68:1-42.

Hawkins, J. A., Ed. (1988). Explaining Language Universals. Oxford: Blackwell.

Hawkins, J. A. (1994). A Performance Theory of Order and Constituency. Cambridge: Cambridge University Press.

Maddieson, I. (1984). Patterns of Sounds. Cambridge: Cambridge University Press.

Searle, J. (1969). Speech Acts: An Essay in the Philosophy of Language. Cambridge: Cambridge University Press.

Shibatani, M., and T. Bynon, Eds. (1995). Approaches to Language Typology. Oxford: Oxford University Press.

Whaley, L. J. (1997). Introduction to Typology: The Unity and Diversity of Language. Thousand Oaks, CA: Sage Publications.

Typology

See also

Additional links

References

Further Readings