Explanations for language universals

Most typologists seek functional explanations for language universals. This is chiefly

because the basis of crosslinguistic comparison is how function is encoded in

grammatical form. Nevertheless, there are many aspects of language function, and

different aspects have been proposed to explain different kinds of language universals.

Proposed explanations for universals of word order (and also the order of affixes) fall

into two general categories: language processing in production and comprehension, and diachronic explanations. Most models of word order universals include harmonic

patterns, in which two or more word orders are correlated. The processing explanation is based on theories of how easy or difficult it is for speakers construct or parse utterances: put crudely, the word order combinations that facilitate parsing are more common crosslinguistically, and those that make parsing more difficult are less common (a detailed theory is presented in Hawkins (1994, 2004). A diachronic explanation of word order correlations is based on the fact that the actual grammatical constructions used for different word orders are historically related. For example, there is a very tight harmonic correlation between genitive-noun order and the order of adposition and noun: genitivenoun correlates with postpositions, and noun-genitive with prepositions (Greenberg 1966). The most likely explanation for this correlation is that adposition constructions are very frequently historically derived from genitive constructions via the process of grammaticalization (Greenberg 1969; Lehmann 1982; Bybee et al. 1994; Hopper and Traugott 2003). This can be observed in the history of English, where the prepositional phrase inside the house derives historically from the noun-genitive construction in the side of the house. Grammaticalization theory has uncovered a large number of universals of language change, specifically processes by which words in particular syntactic constructions evolve into grammatical elements.

The semantic map model also lends itself to a functional explanation for patterns of

the expression of grammatical relations (see Figure 2 in §5), spatial relations and many other grammatical categories and constructions. The explanation is based on a universal conceptual space. The conceptual space represents the degree of conceptual relatedness of the situations or concepts represented by the points in the conceptual space. The structure of the conceptual space is hypothesized to represent general properties of human cognition and conceptualization. The language-specific categories represented by the semantic maps are partly arbitrary (hence the crosslinguistic variation), but they must conform to the constraints imposed by the structure of the conceptual space (Croft and Poole 2008).

Another widely used functional explanation appeals to competing motivations of

economy and iconicity (Haiman 1983, 1985). A simple example of a language universals where economy and iconicity are involved is the expression of inflectional categories. In the category of nominal number, many languages express the singular form without any inflection (zero coding), while the plural is expressed with an overt inflection, as in English branch ~ branch-es. Other languages express both singular and plural with overt inflection such as the Zulu (Bantu, South Africa) prefixes in (28):

(28) a. umu-ntu b. aba-ntu

SG-person PL-person

Other languages, such as Lahu (Sino-Tibetan, Burma) in (29), make no distinction, or to put it another way, express both the concepts of singular and plural without any overt inflection:

(29) qha$/ ‘village/villages’

However, very few languages express the plural without an overt inflection and the

singular with an overt inflection. (In the case of languages that do, the plural is

designated a collective and the singular is a special singulative form, and indeed this

pattern is typically associated with nouns for objects occurring in groups.)

The typological pattern can again be described in terms of a table (Table 2) and an

implicational universal:

Overt Plural Inflection No Plural Inflection

No Singular Inflection attested (English) attested (Lahu)

Overt Singular Inflection attested (Zulu) extremely rare

Table 2. Attested and unattested singular and plural inflectional types.

(30) If a language uses an overt inflection for the singular, then it also uses an overt

inflection for the plural.

The variation allowed by the implicational universal in (30) can be accounted for by

economy and iconicity. Economic motivation is characterized in processing terms: the

more frequently occurring form is expressed by fewer morphemes. The singular is more frequent than the plural in discourse for the vast majority of noun referents, so it may be expressed without an inflection, as in English. Iconic motivation is characterized by the relationship between form and function: it favors a one-to-one mapping between form and function. So if the category of number is expressed in a language, iconicity motivates expression of all values of the category, as in Zulu. (Alternatively, the category may not be expressed at all, as in Lahu.)

Economic motivation is extremely common crosslinguistically: many universals of

grammatical inflection are explainable in terms of frequency differences, and economy is increasingly being invoked to explain universals of syntactic constructions as well (Bybee 2006). Iconic motivation is also extremely common. In fact, iconicity is often taken for granted. For example, the fact that referents and their modifying properties are syntactically constituents (e.g., in Two boys ate five pizzas the number of boys is two, not five) is basically assumed, but it represents the iconic motivation that conceptual relations support syntactic constituency. Also, most syntactic theories contain principles specifying that the syntactic arguments of a verb must match the semantic participants in the event denoted by the verb in number and type; this is also an iconic principle. Economy and iconicity frequently are in competition, leading to the crosslinguistic variation of the type observed in Table 2.

The competition between processing mechanisms of different kinds and the different

types of conceptual relations holding among components of the situation expressed by a grammatical construction leads naturally to a dynamic model of language and language universals. Many typologists have proposed evolutionary models of how language structures adapt to the functions they perform in communication and the constraints on comprehension and production of utterances in language use (e.g. Greenberg 1979; Croft 2000, 2003; Givó n 2002). In an evolutionary approach, a linguistic system is conceptualized as a stage in the process of language change, which can shift by the choices of speakers. Language universals are the result of universal cognitive and social interactional forces that shape speakers’ choices at all timescales, from the immediate discourse situation to the lifetime of a speaker to the transmission of language across generations. The evolutionary approach to grammar emerging from typology its integration with related approaches, including sociolinguistics (Coulmas, this volume) and the usage-based model (Bybee 2006), promises a major step forward in our understanding of language.