Professor Rickford has been a faculty member in Linguistics at Stanford University since 1980 and is Martin Luther King Jr. Centennial Professor and Director of the Program in African and Afro-American Studies. He is the author of numerous scholarly articles and books, including Spoken Soul: The Story of Black English (co-authored, 2000, winner of an American Book Award), Dimensions of a Creole Continuum (1987), Sociolinguistics and Pidgin-Creole Studies (ed., 1988), African American English (co-ed., 1998), and Creole Genesis, Attitudes and Discourse (co-ed., 2000).

There will be a reception with refreshments in the Garrison Room immediately following the talk.

Abstract
One of the newest variables to be considered in the long-standing debate about the English vs. Creole origins of African American Vernacular English [AAVE] is the omission of the relative pronoun or relativizer (that or WH-forms like what, who, or which] in restrictive relative clauses, as in:  (1)  That's the man Ø (who/that/what) I saw. On the basis of a quantitative analysis of relativizer omission in "Early African American English" [EAAE], a collective designation for Samaná English, African Nova Scotian English, and Ex-Slave Narrative data from the US, Tottie and Harvie (2000) conclude (p. 225) that EAAE is derived from English stock, since relativizer omission in these varieties appears to show the same constraint patterning found in white US and British dialects. Moreover, although they have no quantitative data on relativization in creoles, the authors claim that the possibility that the EAAE relativizer system parallels or derives from creoles is slim.
    In this paper, I will attempt to fill the missing gap in this argumentation by presenting a quantitative analysis of relativizer omission in Jamaican, Guyanese and Bajan, taking into account the central constraints considered by Tottie and Harvey and others who have worked on this variable (e.g. Guy and Bayley 1995, Lehmann 2001).  These include the grammatical category, adjacency and humanness of the antecedent NP, and the category membership of the subject of the relative clause.  Among other things,  the Anglophone creole and vernacular varieties display some of the same constraint effects on relativizer omission that EAAE and other English varieties do—for instance existential constructions and definite antecedent NPs (especially of the superlative type—di oglies maan), favor relativizer omission in all these varieties.  What this suggests is not just that creole ancestry might have played a role in the development of EAAE and AAVE, but that the constraints on this variable might be so general or universal that it might be useless as a diagnostic of creole vs. English ancestry.  Indeed, in several respects, the Caribbean creole and vernacular data appear to bear out the more general language processing hypothesis adumbrated by Jaeger, Wasow and Orr (in press): the more predictable the occurrence of a relative clause is, the more likely it is to lack a relativizer.  This may be bad news for attempts to close off the long-standing debate about AAVE's creole origins, but it opens new vistas for studying and understanding variability in English, and in language more generally.

References
Guy, Gregory, and Robert Bayley.  1995.  On the choice of relative pronouns in English.  American Speech 70.2:148-62.
Jaeger, Wasow and Orr. In press. Lexical Variation in Relativizer Frequency.
Lehmann, H.M.  2001. Zero subject relative constructions in American and British English.  Language and Computers 36.1:163-77.
Tottie, Gunnel, and Dawn Harvie.  2000. It's all relative: Relativization strategies in Early African American English. In Shana Poplack, ed., The English History of African American English.  Oxford:  Blackwell.