Meaning change
A key characteristic of grammaticalisation, which accompanies and indeed can be said to give rise to form changes, is change in the meaning or function associated with a linguistic form. While some grammaticalisation scholars argue that this semantic change is the result of meaning loss, called semantic bleaching or attenuation (weakening), others argue that it is more accurate to describe the semantic change that characterises grammaticalisation, particularly in the early stages of grammaticalisation, as an instance of polysemy. Croft (2003: 262) describes this polysemy as ‘a chain of related meanings or uses’, and illustrates this point by means of the English word that, which has four functions. This coexistence of related meanings which emerged at historically different periods is sometimes called layering in grammaticalisation theory. The four functions of that are illustrated by the following examples.
As Croft observes, there may not be a single meaning that underlies all four functions of that. Nevertheless, we might plausibly argue that the demonstrative function has been extended from the domain of physical space (6a) to the domain of linguistic organisation (6b). In other words, the anaphoric demonstrative ‘points to’ another element in the discourse (the fact that George snores) rather than an entity in the physical context. Similarly, the function of that as a com plementiser is to ‘point to’ or introduce what is coming next (6c), while the relativiser that in (6d) ‘points to’ some characteristic of the noun man which the relativiser that introduces. From this perspective, the four uses of that are plausibly related to and motivated by deixis. Cognitive linguists therefore argue that the semantic change that characterises grammaticalisation involves not necessarily ‘semantic bleaching’, but the shift from lexical or content meaning to grammatical or schematic meaning, which at certain stages in the grammaticalisation process gives rise to a set of overlapping form-meaning pairings along the continuum between content units and grammatical units.
Finally, just as grammaticalisation involves phonological and morphological loss, it can also involve semantic or functional loss. To illustrate this point, we can return to the French negation construction that we saw in example (5). Here, the emphatic meaning of pas is lost as it becomes a fully grammaticalised negation particle and the negation function of ne is lost as it is superseded by pas. This explains why it eventually slips out of certain varieties of the language altogether.
The study of grammaticalisation has a rich history, dating back at least as far as the eighteenth century (Heine et al. 1991: 5). This area of language change has received most attention from philologists (historical linguists with a particular interest in establishing language families) and from typologists, and has therefore been approached more from a functional than a formal perspective. More recently, a number of cognitively oriented theories of grammaticalisation have emerged. We limit ourselves in this section to introducing and discussing three cognitively oriented theories of grammaticalisation. As we will see, these theories differ in a number of ways, but what they share is the view that grammaticalisation is essentially grounded in meaning and the view that grammaticalisation is a usage-based phenomenon. These three approaches are rep resented in Figure 21.1.
