In morphology – as elsewhere – the questions we ask our theory to answer make assumptions about the form of the theory itself. Sometimes, the questions lock us in to a theory which we know to be inadequate, because we have data which contradicts a basic assumption of the theory. For example, we know that mouse does not have just one plural form, and the implication is that we should not expect a morphological component to generate just one form to fill a slot in a morphological paradigm.

In this presentation I shall take half-a-dozen such questions which, I shall try to argue, give rise to false expectations of theoretical models in morphology. The stance I take is controversial, but if it is accepted leads to a much more cognitively oriented view of what we should expect from morphological theory.

This presentation is a lightly revised version of a plenary presented at the European Linguistics Society in Poznan, Poland in 2014.