The Historical Evolution of Earlier African American English (Kautzsch)

Ph.D. dissertation University of Regensburg, 2000

(published as : Kautzsch, Alexander. 2002. The Historical Evolution of Earlier African American English. An Empirical Comparison of Early Sources. Berlin and New York: Mouton de Gruyter.)

This project deals with the historical reconstruction of Earlier African American English (AAE) as spoken between the second quarter of the 19th century and the first quarter of the 20th century. It provides detailed quantitative analyses of negation patterns, copula usage, and relative marker choice, which are interpreted in terms of temporal change, regional diversity, and gender variation in Earlier AAE.

What differentiates the methodological approach in this project from common practice in the field is that for the analyses an electronic corpus was created from a variety of relevant early sources, focussing on two genres. The first analytical part is based on interviews with former slaves to represent speech. The second part uses non-standard letters to account for variation across speech and writing.

The texts for the first part are selected from transcribed interviews with elderly African Americans conducted in the 1930s, 1940s, and 1970s, and published in Bailey et al. (eds., 1991), Hyatt (ed., 1970-78), Perdue et al. (eds, 1976), and Rawick (ed., 1977/79). When reconstructing spoken Earlier AAE, a combination of several sources has a range of methodological benefits. First, the sources can be analyzed separately and a direct comparison of the quantitative and qualitative findings across sources facilitates the assessment of potential similarities and differences between them, which ultimately underlines their usefulness as linguistic data. Second, due to the broad database (approx. 470,000 words from 117 informants) it is possible to re-distribute the informants according to gender, state of residence, or birth decade, and interpret the findings along these extra-linguistic factors. Third, a comparison of these transcribed sources with results from spoken data used in other studies provides evidence for the usefulness of properly selected written data for the analysis of speech. The most intriguing results are obtained from the analyses by gender, state, and birth decade: It appears that variation in Earlier AAE is not due to gender. Moreover, it is shown that this variety of English was not regionally homogeneous in the 19th and early 20th century, as is sometimes assumed. Finally, an apparent time analysis based upon the birth decades of the informants provides evidence that linguistic changes in Earlier AAE can be widely located in the second half of the 19th century, which has been argued for theoretically from a socio-historical point of view before (cf. e.g. Bailey 1993, Labov 1998, Mufwene 1999, Poplack 1999), but is backed up by empirical data for the first time in this study.

Part two uses two sets of non-standard letters (a total of 345 letters written by 112 people, Berlin et al. eds. 1982, 1985, 1990, 1993, Miller 1978 and Wiley 1980) to account for the influence of literacy on non-standard writing. It is shown that non-standard structures that are frequently present in speech (transcribed interviews) become severely reduced in number, if not suppressed completely, in writing (e.g.ain't, zero copula, or what as a relative marker). Therefore, from a methodological point of view, some caution has to be called for when non-standard letters are used for the reconstruction of speech.