The field methods for data collection in a research project on linguistic change and variation in the Philadelphia speech community, their origins and use, are described. Five working principles of the project are detailed: (1) there are no single style speakers; (2) styles can be ranged along a single dimension, according to the attention paid to speech; (3) the vernacular, in which minimum attention is paid to speech, provides the most systematic data; (4) any systematic observation of a speaker defines a formal context where more than minimal attention is paid to speech; and (5) face-to-face interviews are the only means of obtaining the needed volume and quality of recorded speech. The field methods outlined include neighborhood studies (aims, design, selection, entry, the interview and its components, and field experiments), the telephone survey (the sampling problem, the Philadelphia sample, survey design, and evaluation), group sessions, rapid and anonymous surveys, a comparison of methods for acquiring sociolinguistic data, and policy for protection of data and subjects' rights (candid recording, protection of anonymity, and access to tape recordings). (MSE)