Adjective ordering as the reflection of a hierarchy in the noun system: A study from the perspective of numeral classifiers

Adjective ordering in English, as in other languages, is nonrandom. In English, the restrictions involve left-to-right sequence, this being a specific case of the general principle: proximity of adjectives to the noun. This article provides a syntactic analysis of such restrictions, focusing not on the adjectives themselves but rather on properties of the nouns modified by them, namely their count/mass properties. Based on the claim that count and mass are hierarchically organized — rather than dichotomous, as previously thought — adjective ordering is shown to be a reflection of the count/mass distinction. This system accounts for the universality of the ordering restriction on adjectives, the universal principle being proximity to the noun. The difference in linear ordering in English and Spanish is ascribed to the presence/absence of a functional category, this being considered as a parameter. Non-canonically ordered adjectives in English are given a syntactic account as well, thus obviating the need for a pragmatic account.