Using a Word-Learning Paradigm to investigate three forms of generalisation in the acquisition of lexical knowledge

word-learning language acquisition

The ability to generalise knowledge from a limited set of exemplars is at the heart of our language abilities. One example of this productivity arises in the domain of morphology. Through experience with a small number of morphologically-complex forms (eg, unhappy, unruly), we abstract knowledge about components within those forms (eg, -un) for use in interpreting and creating new words.

Our research will gain insight into the mechanisms that underlie the acquisition of this knowledge. Specifically, we will :

a) investigate the nature of the knowledge acquired through exposure to full morphologically-complex forms;
b) delineate the time course of this learning process;
c) investigate how various properties of full forms influence the knowledge acquired.

Our studies will use a new method in which adults are trained on novel affixes (eg, -nule) presented in novel word contexts (eg, sleepnule, buildnule), and then tested on untrained novel stimuli that contain the learned affixes. Such tests permit us to assess the extent to which learned morphemic knowledge has generalised to new instances.

This research will provide data central to building a robust theory of the acquisition of morphemic knowledge, and will permit new insights into the mechanisms underlying first- and second-language learning.