The historical linguist has to rely on written records of a speech event for evidence regarding spoken interaction of the past. The authors consider records of trial proceedings and witness depositions, two genres that are included in A Corpus of English Dialogues 1560-1760 (currently under compilation). They present evidence both for and against the reliability of these written records as representations of past speech. The role of the scribe, as well as the printer and editor, is discussed. The authors also offer evidence to suggest that records that are contemporaneous with the speech event are preferable to later printings and editions for linguistic research. They conclude that the study of speech-related texts is of value in its own right, although these texts cannot be equated with present-day audio recordings of spoken interaction.Texts purporting to record spoken dialogue are especially important for the historical linguist as they give access to perhaps the closest thing we have to authentic interaction between speakers from a cross section of society in the days before audio recordings. Among these speech-related texts are trial proceedings presented in dialogue format and witness depositions-which are the focus of this study-as well as trial reports and minutes (which sometimes include language presented as dialogue), parliamentary debates, church council proceedings, and the like. We now have access to such texts in English, dating mainly from the sixteenth century onwards, in formats ranging from manuscripts, early imprints, later imprints, reprints, facsimiles, and text editions. The problems of working on historical linguistic data are well known to us, but additional complicating factors must be taken into AUTHORS' NOTE: We would like to thank Anne Curzan, Jonathan Culpeper, Peter Grund, and the two anonymous reviewers for their helpful comments on the draft version of this article.

No abstract

No abstract

No abstract

Cet article décrit une ressource électronique contenant des textes qui représentent le discours oral du passé de l’anglais moderne naissant. La ressource en question est A Corpus of English Dialogues 1560–1760 , un corpus de 1.2 million de mots comprenant cinq genres : les rapports de procès, dépositions de témoins, pièces de théâtre, œuvres didactiques et la prose fictionnelle. Les procès et les dépositions représentent des dialogues authentiques avec l’intervention minimale d’un scripteur et les autres genres constituent des dialogues construits par des auteurs connus ou anonymes. Nous décrivons les propriétés des genres inclus dans ce corpus et les illustrons en commentant des extraits tirés du corpus. Après avoir présenté les recherches effectuées jusqu’à présent sur la base de ce matériel, nous concluons en décrivant les différentes versions du corpus disponibles pour les utilisateurs.