About this Collection

The Center for Applied Linguistics Collection contains 118 hours of recordings documenting North American English dialects. The recordings include speech samples, linguistic interviews, oral histories, conversations, and excerpts from public speeches. They were drawn from various archives, and from the private collections of fifty collectors, including linguists, dialectologists, and folklorists. They were submitted to the Center for Applied Linguistics as part of a project entitled "A Survey and Collection of American English Dialect Recordings," which was funded by the Center for Applied Linguistics and the National Endowment for the Humanities.

The collection includes recordings from forty-three states, the District of Columbia, Puerto Rico, and parts of Canada. They were made from 1941 to 1984, with the bulk being recorded between 1968 and 1982. In some cases, transcripts (and partial transcripts) made by the collectors are available as part of this web presentation. The survey's documentation covers social aspects of English language usage in different regions of the United States. It reveals distinctions in speech related to gender, race, social class, education, age, literacy, ethnic background, and occupational group (including the specialized jargon or vocabulary of various occupations). The oral history interviews are a rich resource on many topics, such as storytelling and family histories; descriptions of holiday celebrations, traditional farming, schools, education, health care, and the uses of traditional medicines; and discussions of race relations, politics, and natural disasters such as floods.

In 1983, The Center for Applied Linguistics (CAL), under a grant from the National Endowment for the Humanities, initiated the "Survey and Collection of American English Dialect Recordings," which was undertaken to compile a directory of American speech samples, and ultimately to improve access to them. There were two parts to the project. The first part surveyed collectors of speech samples and resulted in a report entitled "American English Speech Recordings: A Guide to Collections" (PDF, 3MB) which describes over 200 collections. The second part of the project was designed to select from the 200 collections a representative sample of recordings with the goal of "creating a centralized source of American dialect samples and to provide for the preservation of this valuable resource that might otherwise be lost." Additional information about the project may be found in "A Survey and Collection of American English Dialect Recordings" (PDF, 6.1MB).

The recordings from the project's second phase were donated to the Library of Congress in May 1986, and comprise the Center for Applied Linguistics Collection (AFC 1986/022). The collection contains speech samples from over 50 collectors. There are 405 recordings which average about 15 minutes in length. In addition to providing a wide variety of material for the study of American speech, the recordings capture a diverse range of topics and people.

Some of the voices will be quite familiar, such as those of Franklin and Eleanor Roosevelt. Other voices, despite the speakers' fame, are seldom heard, such as those of Amelia Earhart, Jack Dempsey, and H. L. Mencken. Most of the recordings, however, are of the voices of people whose specific identities are unknown, but whose comments represent the richness of the American experience. There are Gullah speakers from coastal South Carolina, sharecroppers from Arkansas, Puerto Rican teenagers in New York City, Basque sheepherders from Colorado, Chesapeake Bay watermen, Vietnamese immigrants from Northern Virginia, and many others.

A selection of 350 of the collection's entire 405 audio recordings are available on this website; of these, 148 have accompanying transcripts. The remaining recordings, which could not be posted due to copyright issues and other restrictions, may be heard in the Folklife Reading Room.

Users may notice that descriptive data for the website varies from recording to recording. That is a result of the collection containing material from a wide variety of collectors. The data presented, for the most part, is that which was supplied by the collector.