LINGUISTIC TYPOLOGY Also language typology, typology of language. The classification of human languages into different types on the basis of shared properties which are not due to common origin or geographical contact. Linguistic typology therefore complements the long-established tradition of genetic classification, in which languages are assigned to a family on the basis of their presumed historical origin. The criteria used for dividing languages into types depend to some extent on the purpose of the classification, since a typology based on sound structure does not necessarily correlate with one based on word order. The most common classificatory criteria are morphological (word structure), syntactic (word order), and phonological (sound patterns).

Morphology

Investigation of the way in which different languages combine grammatical units (MORPHEMES) within WORDS is the longest-established aspect of typology. In the 19c, there was an attempt to assign languages to a number of basic morphological types, most commonly three, which divided languages according to the degree to which morphemes are fused together: (1) Analytic or isolating languages, in which each morpheme tends to form a separate word, as in Vietnamese Com nâú ngoài troì ǎn rât nhạt (rice cook out sky eat very tasteless: ‘Rice which is cooked in the open air is very tasteless’). (2) Agglutinating languages, in which several morphemes are juxtaposed within a word, as in Turkish adamlardan (adam-lar-dan, man-plural-from: ‘from the men’). (3) Fusional languages, in which morphemes are fused together within a word, as in Latin servorum (of slaves), where the ending -orum is a fusion of possession, plural, and masculine. In practice, few languages are pure types, since many use all three processes, even though one favoured method tends to predominate. English has a tendency towards isolation (as in I will now go out for a walk), but both agglutination (as in clever-ly and high-er) and fusion (as in gave, in which give and past are fused) are also found.

Alternatively, languages may be classified morphologically according to the number of morphemes within a word: analytic languages (ideally, one morpheme per word, such as Vietnamese, with an estimated 1.06 morphemes per word) are opposed to synthetic languages (two or more morphemes per word, such as SANSKRIT with an estimated 2.59 morphemes per word). The most extreme form of synthesis is found in a polysynthetic language, such as Inuit (Eskimo), which has an estimated 3.72 morphemes per word. On this scale, English comes out as mildly analytic with 1.68 morphemes per word. Inflected languages are a variety of synthetic language in which a word takes various forms, most usually by the addition of suffixes, which show its role in the sentence. Many languages have some inflection (such as English boys, the boy's mother, play/played) but in a highly inflected language, such as LATIN, this process predominates. There are therefore two morphological scales, one which measures degree of fusion (isolating—agglutinating—fusional), the other degree of synthesis (analytic—inflected—polysynthetic). Since an isolating language is inevitably also an analytic language, English is at the low end of both scales.

Syntax

In the past quartercentury, basic word order has been the main criterion for classifying languages. In the early 1960s, it was observed that of the possible combinations of subject (S), verb (V), and object (O) within a sentence, only certain ones actually occur, and that these are not all equally likely. The commonest are those in which the subject comes first (SVO as in English, SOV as in Turkish), less common are those in which the verb comes first (VSO as in Welsh, VOS as in Malagasy), and least common are those in which the object comes first (OVS as in Hixkaryana, spoken in northern Brazil, OSV of which no sure example has yet been found). Many languages have mixed word orders, and not all languages have a firm order, so this classification has its flaws. However, English with its SVO structure, such as The rabbit (S) gnawed (V) the carrot (O), is a language with one of the two commonest word orders, even though some subsidiary orders are possible, such as Up jumped the rabbit.

The relative order of verb and object is often considered to be most important from the point of view of typology, since not all languages express overt subjects. The main interest in classifying languages in this way lies in the implicational relationships, in that certain other constructions are statistically likely to occur in each type. A VO language, such as English, is likely to have prepositions rather than postpositions (such as up the tree rather than *the tree up), and auxiliaries before main verbs (such as Bill may come rather than *Bill come may). It is also likely to have relative clauses (beginning with who/which, etc.) after the noun they refer to, such as The burglar who stole the silver escaped rather than *The who stole the silver burglar escaped. The general principle behind these observations appears to be a preference for consistency in the position of the head (main word) in any construction with regard to its modifiers (items attached to it): so a VO language such as English is a ‘head first’ language and an OV language such as Turkish is a ‘head last’ language.

A controversial proposal by Noam CHOMSKY is that humans are genetically ‘hardwired’ with some universal features of language, but that these are supplemented with a number of options which have to be selected on the basis of exposure to a particular language: see LANGUAGE ACQUISITION DEVICE. The choice of one rather than another has complex ramifications throughout the language. Different language types are therefore the result of a number of fairly simple choices which are automatically available to humans. One proposal for such an option is between a prodrop language (one which can optionally drop pronouns at the beginning of sentences, as in ITALIAN Io sono Italiano/Sono Italiano I am Italian/Am Italian) and one which does not usually do so, such as English. Pro-drop languages seem to behave somewhat differently over a range of constructions from languages which do not drop their pronouns.

Phonology

Phonological typology has received somewhat less attention, though some interesting work has been done on types of vowel system. In addition, a number of studies have proposed implicational hypotheses, such as if a language has fricative consonants, it will also have stop consonants. With regard to RHYTHM, some linguists divide languages into: (1) Syllabletimed languages, such as FRENCH and Japanese, in which the rhythm appears to be fairly even, with each syllable giving the impression of having about the same weight as any other. (2) Stress-timed languages, such as English and ARABIC, in which stressed syllables recur at intervals. In recent years, a somewhat ‘weak’ version of this view has gained ground. The absolute division has been replaced by a sliding scale, in which there are few pure types, though many which can be placed towards one or the other end of the scale. There is no doubt that English is on the stress-timed end of the scale. Another distinction is sometimes made between tone or tonal languages, such as Mandarin Chinese, and intonation or non-tonal languages, such as English. In a tone language, the pitch level of any syllable is of critical importance, since words are sometimes distinguished from one another purely by the tone, such as Mandarin ma with level tone (mother), with rising tone (hemp), with a dipping tone (horse), and a falling tone (scold). In a language such as English, however, sentence intonation plays a crucial role, as in You saw him! versus You saw him?, where difference in meaning is signalled by the intonation.

Conclusion

Linguistic typology is currently in a state of considerable flux and controversy. A wide range of criteria apart from those outlined above are currently under discussion, and it will be some years before reliable methods of classifying languages are firmly established. See LANGUAGE FAMILY.