Now that we have reached the end of our – certainly partial, but, we hope, indicative – survey of the data and the theories relevant to the study of imperatives, you may wonder what should be the next step. Although we sought to be objective and even-handed in writing Part II, we did not make any particular effort to hide our sympathies or lack thereof. Of course, you probably disagree with some of our judgements, and perhaps some of our criticisms of one theoretical stance or another merely managed to persuade you how right that position is. Thus the lessons learned will probably be different from reader to reader, which is how it should be. That said, let us sum up here what we have learned in the process of writing this book.

The idea that we have probably insisted on most throughout the book is that the imperative mood should not be thought of as encoding directive force. The first reason for rejecting such an identification of form and function is methodological. Cutting the tie between mood and force enables a much more flexible relationship between sentence types and speech acts. It helps understand why the imperative may have non-directive uses: in particular, it makes non-directive ILCs less puzzling. It also dissolves an otherwise threatening cross-linguistic observation: there exist languages with no imperative sentence type. In Chapter 4, we saw again and again that various theories that treat imperatives as linguistic carriers of directive force do not manage to avoid such problems.