Reconsidering Prepositional Polysemy Networks: The Case of Over

This article explores lexical polysemy through an in-depth examination of the English preposition over. Working within a cognitive linguistic framework, the present study illustrates the nonarbitrary quality of the mental lexicon and the highly creative nature of the human conceptual system. The analysis takes the following as basic: (1) human conceptualization is the product of embodied experience, that is, the kinds of bodies and neural architecture humans have, in conjunction with the nature of the spatio-physical world humans inhabit, determine human conceptual structure, and (2) semantic structure derives from and reflects conceptual structure. As humans interact with the world, they perceive recurring spatial configurations that become represented in memory as abstract, imagistic conceptualizations. We posit that each preposition is represented by a primary meaning, which we term a protoscene. The protoscene, in turn, interacts with a highly constrained set of cognitive principles to derive a set of additional distinct senses, forming a motivated semantic network. Previous accounts have failed to develop adequate criteria to distinguish between coding in formal linguistic expression and the nature of conceptualization, which integrates linguistic prompts in a way that is maximally coherent with and contingent upon sentential context and real-world knowledge. To this end, we put forward a methodology for identifying the protoscene and for distinguishing among distinct senses.*

1. Principled polysemy

We focus here on the issue of semantic polysemy, the phenomenon whereby a single linguistic form is associated with a number of related but distinct meanings or senses. In particular, we consider how the notorious polysemy of the English preposition over might be accounted for in a principled, systematic manner within a cognitive linguistic framework. At base, we argue that the many senses of over constitute a motivated semantic network organized around an abstract, primary meaning component, termed a protoscene. The many distinct senses associated with over are accounted for by interaction of the protoscene with a constrained set of cognitive principles. Accordingly, our more general claim is that the lexicon is not an arbitrary repository of unrelated lexemes. Rather, the lexicon constitutes an elaborate network of form-meaning associations (Langacker 1987, 1991a, 1991b), in which each form is paired with a semantic network or continuum (Brisard 1997). This follows from two basic assumptions, widely demonstrated within the framework of cognitive linguistics. First, semantic structure derives from and mirrors conceptual structure (see, for example, Fauconnier 1994, 1997, Heine 1997, Jackendoff 1983, Lakoff 1987). Second, the kinds of bodies and neural architecture human beings have—how we experience—and the nature of the spatio-physical world we happen to live in—what we experience—determine the conceptual structure we have (Clark 1973, Evans 2000, Grady 1997, Heine 1993, 1997, Johnson 1987, Lakoff & Johnson 1980, 1999, Svorou 1993, Sweetser 1990, Talmy 1983, 1988, 1996, 2000, Turner 1991, Varela et al. 1991).

This model of the lexicon generally, and the model of polysemy proposed here in particular, contrasts with traditional models in a number of ways. The traditional view holds that all regularity and productivity are in the syntax, with the lexicon serving as a repository of the arbitrary. Aronoff (1994) points out that Bloomfield articulated this perspective as early as 1933. More recently, Chomsky has reasserted this stance: ‘I understand the lexicon in a rather traditional sense: as a list of “exceptions”, whatever does not follow from general principles’ (1995:235). Models within this framework have tended to represent different word senses as distinct lexical items (Croft 1998). Polysemous forms are simply represented as an arbitrary list of discrete words that happen to share the same phonological form.

Over the years, this stand has been criticized for failing to account for systematic ways in which numerous forms are clearly related (Jackendoff 1997, Langacker 1991a, Levin 1993, Pustejovsky 1998). Croft (1998) notes that a number of linguists have argued for some type of derivation within the lexicon that would represent distinct senses as arising from a primary sense via a set of lexical operations. By and large, these analyses have focused on polysemy involving changes in the argument structure of verbs or alternatively in category changes...