Productivity: Definitions and measurements
The central aim of general morphological theory is to define the notion of 'possible complex word in natural language', or, in the case of theories about a specific language, the notion of 'possible complex word in language A'. Such a theory would ideally not only describe existing complex words but also non-existing, but equally possible derivatives, as they could be formed by the speakers according to the regularities and conditions inherent in their morphological competence. In other words, any morpho logical theory should make predictions which words are possible words of a language and which words are not. Assuming the existence of morpho logical rules according to which complex words are structured or formed, one can easily observe that some rules are often used to create new words, whereas others are less often used, or not used at all for this purpose. In this sense, some rules can be called productive and other rules unproductive or less productive. The obvious question now is which mechanisms are responsible for the productivity of morphological processes. Is the productivity of a rule the result of structural factors, conditioned by pragmatic factors, or an inherent non-derivable property of any morphological rule? And how can we measure the productivity of a rule?

We attempt to shed some new light on this problem by investigating the nature of some of the mechanisms that condition the productivity of morphological processes in English. However, before we turn to the empirical side of the matter we will first take a closer look at the notion of productivity and its restrictions.

According to a standard dictionary of linguistic terminology, the term productivity is "used in linguistics to refer to the creative capacity of language users to produce and understand an indefinitely large number of sentences." (Crystal 1991:279, see also Chomsky 1957:15). Whereas the role of productivity in syntax is generally not regarded as problematic or is simply ignored¹, there is hardly any work on word-formation where productivity is not prominently discussed.

In most publications, productivity is loosely referred to as the possibility to coin new complex words according to the word formation rules of a given language. Although this definition of productivity seems rather straightforward, we will see that numerous problems remain unresolved, and that, in spite of its centrality to the study of word formation, there is still no consensus about the nature of productivity. One is therefore still inclined to agree with Aronoff when he states, "The term productivity is widely used in the studies of derivational morphology, and there is obviously some intuition behind the usage, but most of the discussion is rather vague" (1976:35). Before we try to remove some of this vagueness, some remarks are in order concerning the crucial concepts to which notions of productivity refer.

¹ Cf. Aronoff (1980b:71), but see also the critical remarks by Aronoff (1976:35), or Bauer (1983:65-74). In his discussion of syntactic productivity, the latter comes to the conclusion that "syntactic and morphological productivity have more in common than they have to distinguish them" (1983:74). Since we are concerned with morphological productivity here, this point will not be further discussed.