This book is a selection and integration of excerpts and complete narratives from the Slave Narrative Collection of the Federal Writers' Project, whose history and method are discussed below. The present note concerns my own part in a work which has passed through many hands and minds, as well as several stages of completion.
Since the work of collection had been nearly completed when I became folklore editor of the Federal Writers' Project in 1938, my efforts have been devoted chiefly to utilization. From 1939 to 1941, as chief editor of the Writers' Unit of the Library of Congress Project, I planned and supervised the work of organizing and preparing the collection for deposit in the Library of Congress.
In February, 1944, I was asked to prepare for publication a selection of the narratives "which would give at once the flavor of the entire collection and the social patterns revealed in the series, while keeping literary excellence to the forefront." This meant not only a reduction in scope and scale but also a revision of my thinking about the narratives to the point where I could see them not as a collection of source material for the scholar but as a finished product for the general reader. Specifically, it meant (1) reducing the bulk of the collection from over ten thousand to five or six hundred manuscript pages; (2) concentrating on its broadly human and imaginative aspects and on those oral, literary, and narrative folk values for which in 1928 I coined the word "folk-say"; and (3) fitting the selections into some sort of sequence that would give pattern to the book and to slavery.
To do all this in the narrow compass of some three hundred pages required the use of excerpts in preference to complete narratives, except where the latter constituted an irreducible minimum, as in the autobiographical section, "Long Remembrance," which balances the picture of slavery with the picture of the individual life. At the same time, the excerpts required considerable pointing up, by careful cutting, arranging, and titling, to give proper emphasis where there was none or where, out of context, there might be a wrong emphasis.
In accordance with the same criteria of truth and readability, the original attempts at dialect-writing, successful and unsuccessful, were abandoned, except for a few characteristic and expressive variations. Although they had been instructed to put truth to idiom first, many interviewers were betrayed by their zeal for accurate recording into stressing truth to pronunciation, which