Desemanticization
Desemanticization is an immediate consequence of extension: Use of a linguistic expression E in a new context C entails that E loses part of its meaning that is incompatible with C—in other words, the two are (as we observed in “Extension”) Janusian sides of one and the same process.1
Desemanticization is frequently triggered by metaphoric processes (LakoV and Johnson 1980; LakoV 1987). For example, a paradigm case of grammaticalization involves a process whereby body part terms (‘back’, ‘head’, etc.) are reinterpreted as locative adpositions (‘behind’, ‘on top of’) in specific contexts, cf. English in front of. Via metaphorical transfer, concepts from the domain of physical objects (body parts) are used as vehicles to express concepts of the domain of spatial orientation, while desemanticization has the effect that the concrete meaning of the body parts is bleached out, giving way to some spatial schema. In a similar fashion, when an action verb (e.g. English keep, use, go to) is reinterpreted as a tense or aspect auxiliary (see “The present approach”), this can be understood to involve a metaphorical process whereby a concept of the domain of physical actions is transferred to the more abstract domain of temporal and aspectual relations.2 Once again, this leads to the desemanticization of lexical meaning, namely that of the action verbs (Heine, Claudi, and Hünnemeyer 1991; Heine 1997b).
Rather than desemanticization, Hopper and Traugott (2003) prefer to describe the semantic development in terms of notions such as invited inferences, subjectification, or pragmatic strengthening. It would seem, however, that desemanticization is a fairly predictable component of the process while neither invited inferences, subjectification, nor pragmatic strengthening are: There are many grammaticalization processes that involve neither inferencing nor subjectification, nor strengthening and, conversely, there may be inferencing without there necessarily being desemanticization (Fritz Newmeyer, p.c.).
1 This view is at variance with that of Hopper and Traugott who argue that desemanticization (‘‘weakening of meaning’’ in their terminology) is not involved in the beginnings of grammaticalization—that is, that it follows extension (or ‘‘pragmatic strengthening’’): ‘‘There is no doubt that, over time, meanings tend to become weakened during the process of grammaticalization. Nevertheless, all the evidence for early stages is that initially there is a redistribution or shift, not a loss, of meaning’’ (Hopper and Traugott 2003: 94).
2 This is a simplified rendering of the process concerned; see our example of German drohen in “Assumptions”, which gives a more detailed description of such a process from lexical verb to auxiliary.
