Polysemy and monosemy as a cline
The fact that none of the proposed tests of polysemy seems to deliver reliable results has led many linguists to dismiss the polysemy/monosemy contrast as a false dichotomy. One of the earliest to do so was Geeraerts, who rejects the idea that we should think of meanings as ‘things, prepack aged chunks of information that are contained in and carried about by word bags’ (Geeraerts 1993: 259; see also Tuggy 1993, Allwood 2003). This idea is compatible with the ‘conduit metaphor’ discussed in 1.6.2, and once we abandon it, it is no longer important to know whether a word carries around one prepackaged information chunk (monosemy) or several (polysemy).
One possible alternative to the view of words having a determinate and finite number of senses would be to think of a word’s meaning as a continuum of increasingly fi ne distinctions open to access at different levels of abstraction (cf. Taylor 2003: Chapter 8). Depending on the level of abstraction at which a word’s meaning is considered, different elements of its meaning may appear as distinct or not, with the word consequently appearing variously polysemous or monosemous on the different polysemy tests. For example, consider the dialogue in (39), adapted from Tuggy (1993):
If Jane has been painting a portrait and B has been painting stripes on the road, this answer will be misleading since it suggests that they have been engaged in the same type of painting; as a result, B’s reply could only be uttered facetiously, punningly, or with the intention to mislead. On the linguistic criterion discussed above, paint would thus be polysemous between two senses which we could provisionally gloss as ‘engage in artis tic activity involving the application of paint’ and ‘engage in a non-artistic activity involving application of paint’. In other contexts, however, the linguistic test does not point to different senses of paint, suggesting that it is in fact monosemous (general) between the portrait and road stripe painting senses. Thus, imagine in (40) that Franz is painting a portrait, and that the speaker is painting stripes on the road:
How can this clash between the test results be resolved? One answer would seem to be that (39) and (40) invoke differing levels of abstraction of the concept of painting. The verb paint can be used to refer to a broad continuum of different activities (as well as road and portrait painting, there is face-painting, painting of walls, rust-proofing, nail-painting, etc.). Strictly speaking, none of these individual instances of painting is absolutely identical to any other: two acts of wall-painting, for example, will differ in the details of their physical and temporal locations. The function of the verb paint is thus to categorize all of these different referents together (Taylor 2003; see 7.1 for further discussion). The relative importance of individual instances of painting is not, however, stable. When, as in (39), an accurate description of the type of activity being undertaken is called for, then painting a portrait and painting stripes on the road will be seen as fundamentally different activities: one is an artistic pursuit often associated with the leisure activities of amateurs, while the other takes place in the context of professional employment. Given the differing values of the two types of painting in our society, their common description by the same verb would be misleading. In (40), however, painting is considered not in terms of its wider socio-cultural import, but in terms of its actual mechanics. In this context, the differences between road-stripe painting and portrait painting disappear, since even application of colour is equally relevant to both; consequently, the verb paint may be used to refer to both types of situation without any punning, awkwardness or risk of misinformation. It is as though paint comprehends a variety of related notions, such as portrait painting, painting road-stripes, painting walls, painting the face, etc., which may be ‘zoomed’ in on and out from. When what is required is a fi ne-grained description of the type of activity in question, a ‘close-up’ view of the notions covered by paint makes each one stand out as a distinct unit, in the same way that a photographic close-up will reveal the detailed structure of an object. But when the focus is wider, the differences between the internal constituents become blurred and lose their distinctness. Accordingly, paint will appear monosemous or polysemous as a result of the level of abstraction or resolution at which its meanings are accessed. To think of a lexical item like paint as either monosemous or polysemous is therefore to ignore the fact that meanings can be accessed at a variety of levels. Rather than being absolute alternatives, monosemy and polysemy name the end points of a cline of semantic separateness.
This type of answer has found a number of recent adherents in discussions of polysemy (see for instance Taylor 2003: Chapter 8). In one sense, however, it does not resolve the problem, and for a similar reason to the one for which we rejected the linguistic test of polysemy: it ignores the distinction between the sense and the reference of paint. The cases dis cussed in (39) and (40) constitute different situations to which paint refers. But how do we know when a different situation corresponds to a different sense of the verb? Might not all the occurrences of paint we have discussed be examples of a single, schematic sense along the lines of ‘apply paint to a surface’ (which will cover both the portrait and the road-painting cases), even at the most fi ne-grained level of resolution? Difference of reference does not automatically entail difference of sense; if it did, the very distinction between sense and reference would lose its point. As a result, the mere fact that paint can be used to refer to a variety of different situations tells us nothing about the number of senses involved.
By now it will be obvious that this issue involves a number of complex questions. For some investigators, the phenomena discussed in this section problematize the very objectivity of meaning as a linguistic phenomenon (Geeraerts 1993; Riemer 2005).
