Functional Categories and the Notion of a Closed Class

In Romeu (2014a) and (2015) I presented a model in which grammatical categories are considered to be only theoretical labels that gather lexical items that lexicalize similar parts of the syntactico-semantic structure of languages. In those works I give evidence based on prepositions. In this work, I show that not only this model can be applied to prepositions but also to other categories and phenomena. By these means I present different ways to explain controversial questions in the analysis of languages. I show that this model allows us to address and explain, among other questions, the differences between prepositions and particles, the real nature of stranding, why only certain languages have resultatives, the difference between restrictive and non-restrictive adjectives, what hides behind the apparently redundant nature of agreement, etc.

1999

2014

Many grammars posit a distinction between open and closed ‘word classes’ (lexical categories) in English. We demonstrate that the supposedly closed categories continue to admit new members. We document recent additions to four putatively closed categories. New determinatives are being added via branching from adjectives (e.g., various, numerous, myriad, several, other, multiple, said). Numerous new prepositions have emerged via branching from participles (e.g., allowing, counting, excluding, including, pertaining, wanting, given, gone, granted, provided, approaching, starting, ending, continuing, omitting, based, compared), branching from adjectives (e.g., clear, level, additional), branching from a noun (e.g., bush), branching from prefixes (e.g., pre, post, anti), and compounding (e.g., online, offline, take away, in spite of). The category of coordinators has been expanded by the addition of as well as, plus, versus, cum, and slash. And even the category of subordinators shows si...

2014

Categories may be functional or lexical (Jespersen 1924; Abney 1987 a.o.). Yet, some categories, such as prepositions, are known to be hybrid. To capture the in-between status of prepositions Zwarts (1997) proposed that the functional-lexical divide is structured by means of three independent privative parameters, viz. [+/-lexical], [+/-functional] and [+/-categorial]. In this article I adapt and develop this proposal to demonstrate that derivational affixes are a hybrid category as well. More specifically, I show that derivational affixes have lexical meaning and that they are an open class, which defines them as [+lexical]. Yet, they obligatorily select a complement, which makes them [+functional]. They may or may not determine the category of their functional superstructure, which renders them [+/-categorial]. It is thus argued that derivational affixes are not purely functional or lexical, just like prepositions. Yet, within the lexical-functional divide these two hybrid categor...

Linguistics, 1992

Our conventional classification of words into parts of speech is only a vague, wavering approximation to a consistently worked out inventory of experience. [.'..] And so we might go on examining the various parts of speech and showing how they not merely grade into each other but are to an astonishing degree actually convertible into each other. Edward Sapir (1982 [1921]: 117f.

1998

1998

Linguistic Typology 13-1, 95 104., 2009

Since Greenberg’s groundbreaking publication on universals of grammar of 1963/1966, typologists have used semantic categories to investigate (constraints on) morphological and syntactic variation in the world’s languages and this tradition has been continued in the WALS project. It is argued here that the employment of semantic categories has some serious drawbacks, however, suggesting that semantic categories, just like formal categories, cannot be equated across languages in morphosyntactic typology. Whereas formal categories are too narrow in that they do not cover all structural variants attested across languages, semantic categories can be too wide, including too many structural variants. Furthermore, it appears that in some major typological studies semantic categories have been confused with formal categories. A possible solution is pointed out: typologists first need to make sure that the forms or constructions under investigation do the same job in the various languages (functional sameness); subsequently this functional selection can be narrowed down on the basis of formal or semantic criteria to construct a set of elements that is similar enough to allow for crosslinguistic comparison (formal and semantic similarity).