Layers
In this section we present a skeleton of grammatical evolution. Our approach is reductionist in a number of ways, in that discussion is narrowed down to a range of notional categories and the major pathways of development linking these categories. Furthermore, space does not allow us to describe the cognitive and pragmatic foundations underlying these pathways, which have been the subject of many individual studies. Wherever possible, however, we will provide relevant information on the contextual frames that have contributed to these pathways.
We will describe linguistic evolution in terms of a set of grammatical categories that tend to be distinguished in the modern languages of the world. It is very likely, in the earlier development of human language, that these categories were not of the same kind as we find them today. For example, nouns in the modern languages usually have syntactic properties such as taking adjectives and demonstratives, or markers for number, gender, and/or case. As the following reconstructions suggest, such properties were presumably absent in the earliest stages of language evolution. The reader therefore has to be aware that the reconstructions proposed are based on the application of grammaticalization theory and can be accounted for with reference to this theory, but they are not necessarily of the same kind as those that may have characterized the structure of early language. In other words, if applying grammaticalization theory will allow us to reconstruct grammar back to an initial stage of nouns only, this does not mean that the first language(s) had nouns the way we know them from most languages nowadays, that is, as characterized by certain grammatical and distributional properties.1
We will describe grammatical evolution in terms of a set of ‘‘layers’’, that is, clusters of categories that show the same relative degree of grammaticalization vis-à-vis both the categories from which they are derived and which they develop into. For example, we observed in “The present approach” that there is a regular development from lexical verbs to functional categories for tense and aspect. Accordingly, we will say that verbs belong to a different layer than categories of tense and aspect. Furthermore, as we will see in “The fourth layer: demonstratives, adpositions, aspects, and negation”, aspect categories can further develop into tense categories, which allows us to argue that aspect and tense each represent a different layer; consequently, this example allows us to reconstruct three distinct layers.
1 The reason for that is very simple: as Maggie Tallerman and Jim Hurford (p.c.) point out to us, one cannot talk of any distinct category until there is another category to contrast it with. Rather, it is reasonable to assume that the stage of nouns only recon structed by means of grammaticalization theory corresponds to entities which served primarily the task of reference—rather than entities that served the task of predication— in the first language(s); we will return to this issue in “Layers”.
