In Clotel or the President's Daughter; A Narrative of Slave Life (1853), William Wells Brown makes use of several authenticating documents that are supposed--in the context of the novel--to be real. That is, these newspaper articles and advertisements, letters, legal tracts, religious publications, and the prefatory narrative of the author's own life and escape from slavery are assumed to exist "outside" the fiction of the novel. Although these documents are sometimes as much fictions as the rest of the novel, they do tend to have "authentic" counterparts, and the existence of such counterparts seems to have the effect of authenticating Brown's story about Thomas Jefferson's slave mistress and slave daughters.
As early as 1979, Robert Stepto delineated the characteristics of the authenticating machinery in Brown's narrative, arguing that in his personal narrative Brown "authenticates himself" and the autobiographical narrative tends to authenticate the novel: "Brown's personal narrative functions ... as a successful rhetorical device, authenticating his access to the incidents, characters, scenes, and tales which collectively make up Clotel" (30). In "The Novelization of Voice in Early African American Narrative" (1990), William Andrews re-asserts the notion that Brown uses authentic discourse and argues that it calls into question or problematizes the notion of stability of the line between "natural and fictive discourse." Brown subverts "the relation of privilege that makes natural discourse the ground for fictive discourse" (27). According to Andrews, "the voice of black narrative broke most profoundly with discursive conventions and white expectations in an attempt to find new ways of authorizing itself" (24).
Other readers also elaborate the notion of authenticating in Brown's novel.(1) Richard O. Lewis, for example, argues that Brown's authentic "documentation ... supplies enough credibility in Brown's works to sustain his ironic situations and metaphorical characters" (153).(2) Despite their careful analyses of the force of historical authentication, however, none of these essays contrasts Brown's own era (the 1850s) with the first third of the century, during which the novel is ostensibly set. Nor does any one of these recent readings pay attention to the chronological and historical inconsistencies of the authenticating documents themselves.(3) Perhaps even more significantly, none takes into account what Brown chooses to authenticate and what he leaves unsubstantiated. Given Brown's need as an ex-slave to authorize himself, one must still ask questions about what the writer chooses to or decides not to authenticate.
In that he seems to apply it inconsistently, Brown's use of authentication appears problematic. The very premise of the novel's argument rests on hearsay, on the rumor that one of Jefferson's daughters by a slave woman was "sold in New Orleans for one thousand dollars" ("Sale of a Daughter" 4). As he does with other "facts" in the history of slavery, Brown could have authenticated this assertion. But he evidently chose not to. Indeed, despite all his other allusions, quotations, and references and despite the availability of newspaper accounts concerning Jefferson's having a daughter sold into slavery, Brown makes no attempt to authenticate what would seem the single most important aspect of his narrative. Rather than authenticate, he simply asserts that Jefferson's relationship existed, letting that assertion stand on its own: "The gentleman for whom she [Currer, Clotel's mother] had kept house was Thomas Jefferson, by whom she had two daughters" (64).
Brown's choice not to authenticate this aspect of the novel by reference to newspaper and other accounts raises an interesting question about his use of authenticating documents generally. If Brown uses "authentic discourse to buttress the fictive claims" (30) as Andrews suggests, the fictive claim of Jefferson's paternity would seem to be absolutely essential. …