15 ADJECTIVE CLASSES: WHAT CAN WE CONCLUDE?
John Hajek- Introduction^1
To many linguists, the claim made by Dixon in Chapter i that all languages will
have three major word classes (noun, verb, adjective) is challenging and provoc-
ative. Cross-linguistically, it is generally considered that the distinction between
noun and verb is, with rare exceptions, a robust one and relatively easy to estab-
lish, according to basic grammatical functions. Nouns prototypically head noun
phrases which function as core arguments and appear in copula complements (or
verbless clauses), whilst verbs prototypically function as heads of (in)transitive
predicates which also bear all tense, aspect, and mood (hereafter TAM) mark-
ing. Our ability to distinguish adjectives has long been known to be more difficult
across languages (e.g. Dixon 19823, Bhat 1994, Wetzer 1992, 1996), since none of
these basic grammatical functions are associated only with adjectives, as opposed
to either nouns or verbs. Grammatical differences that allow for adjectives to be
distinguished, when they occur, are often much more subtle and diffuse. As a re-
sult, and as Dixon points out, many linguists are reluctant to recognize the exist-
ence of an independent adjective class in a language they are describing.^2
Whilst it may sometimes appear difficult to find evidence that allows us to dis-
tinguish adjectives from other word types, Dixon, in Chapter i, claims that such a
task should still be possible with more subtle testing, as the cited evidence from
Yir-Yoront (Alpher 1991) and Chinese (Xu 1988) suggest. The distinction between
(^1) I would like to thank Bob Dixon, Sasha Aikhenvald, and all the authors who have provided me
with additional feedback and information necessary for the preparation of this chapter. I wish in par-
ticular to acknowledge the generosity on the part of Randy LaPolla and Chenglong Huang in provid-
ing additional Qiang data. All errors of interpretation remain mine.
(^2) There is increasing cross-linguistic evidence that the basic grammatical criteria used to distin-
guish nouns and verbs may be fuzzier than is generally thought. Dixon in Chapter i points to the
well-known case of Nootka where more subtle testing is required, but where a distinction exists
nevertheless. Anward et al. (1997) refers to more recent problematic cases, such as Tongan. Nordlin-
ger and Sadler (2004) provide substantial evidence to show that TAM marking is not as prototypic-
ally verbal as thought: many languages in fact allow such marking on nouns—in verbless predicates
and in noun phrases.