This proposed research-project takes as it point of departure the following paradox: on the one hand, as a generative grammarian one routinely considers the grammatical core of a given language in place say at the age of five, if not earlier. On the other, in lexical phonology one customarily makes a theoretically significant distinction, for languages such as English and Dutch, between two word-classes of derived words, cohering and neutral. An example of a cohering affix occurs in the word-pair anonRem/an\nim-us, in which the derived words behaves as if underived for the purpose of stress-assignment, cf. jerdzalem; an example of a neutral affix can be found in the word-pair \lifant/\lifant-je, in which the diminutive suffix is allowed, by being stress-neutral, to create a stress-pattern which is completely absent from the stem vocabulary. One of these latter classes, though, specifically the `cohering' one, covers material that is to a considerable extent acquired only as late as a native speaker's (late) teens.
Given these observations, it strikes one as an interesting challenge to try and probe the acquisitional status of the wordclass with cohering affixation. More specifically, this challenge is proposed to reside in the pursual of the idea that the cohering wordclass, at its generally late stage of arrival, comes to occupy a `slot' in the lexicon that is present from the outset but simply happens to be filled at a late stage.
This research question is approached by trying to find out whether young (Dutch) children have the ability to handle cohering material of their native language, even though this is relatively unfamiliar to them. If they can handle that material, this may be viewed as broad support for the cohering/neutral distinction and the theoretical consequences this distinction entails in lexical phonology; if not (assuming adequate testprocedures), this distinction is less well supported and one is then invited to come up with an explanation of why the empirical situation is as found. Under either result it will be an interesting to try and find an explanation of the late appearance of coheringly affixed words (semantic features may well be involved here, such as the abstract/concrete distinction).
Empirically, the project will make use of three sources of data. First, existing corpuses of spontaneous child speech will be investigated in order to establish a picture with respect to `morphological creativity' in child language; second, data will be elicited in controlled interview situations; and finally, specific experiments will be designed towards testing out creativity in the area of investigation.