Puts forward a methodological framework called the 'p-model' that is expected to account for the data/evidence problem in linguistics. This book shows how this framework can be applied to the everyday research practice of the working linguist, and how it can increase the effectiveness of linguistic theorising.