The Text in Motion: Eighteenth-Century Roxanas

Closure has a bad name. Yet to critique closure is not the same thing as to do away with it. Humans, on the evidence, have a need for closure, even those who know it is philosophically or ideologically suspect. That, at least, is Frank Kermode's thesis in his widely-read book of 1966, The Sense of an Ending.¹ The theological fiction of an apocalypse, in which the end of time would be consonant with its divine origin—for every alpha an omega, for every tick a corresponding tock—is never disproved for believers when a predicted date of the End fails to conform to human calculation. Yet while Kermode very evidently takes pleasure in writing bone-dry sentences about those who are waiting expectantly for the loosing of the beast, he is equally distanced from the intellectual impulse to merely make fun. For just as we need to recognize fact and reality, he states, we also need to live according to patterns. Thus, "the desire for consonance in the apocalyptic data, and our tendency to be derisive about it, seem to me equally interesting."² Turning to literary plots, Kermode finds the same coexistence of naive acceptance and clerkly skepticism as he does in the apocalyptic. The popular story sticks close to established conventions, but novels that are called "major" merely vary them and reach their ends through more indirect ways. Expectation is frustrated only to be answered in a different fashion, so that our sense of reality and our sense of pattern are both satisfied.

Peter Brooks on the whole agrees with Kermode's point, but he explains the need for patterning in terms of what he calls "narrative desire."³ Plot, in Brooks's view, is not adequately explained by formalist models that spatialize the text, but requires an analysis more tuned to the dynamic experience of reading in time. In this model, reading has something of the doubleness of Eros/Thanatos, what Brooks calls "the contradictory desire of narrative, driving toward the end which would be both its destruction and its meaning."⁴ Plots bind the chaos of desire through repetition and difference, leading it towards an end that is, momentarily, both satisfaction and cessation. Yet Brooks also observes that the end that desire seeks is arbitrary and fictive, and that, when it arrives, it may "disappoint and baffle," forcing us to reread in order to seek coherence.⁵ Like Kermode, and speaking of twentieth-century literary production, Brooks too makes a distinction between the popular and the sophisticated, the one following formula—a distribution of rewards and punishments, the settling of the characters' lives through either marriage or death, and so on—while the other reflects self-consciously on formula and its devices, and does not so much come to a close as run out of pages.

Such influential discussions illuminate the issues for us, but they take one only so far because the models they formulate are highly generalized and unhistorical. Their opposition between the popular and the sophisticated, for example, is not borne out by the literature of earlier periods. In Defoe's Roxana, which was undoubtedly a popular story, the first edition of 1724 ends without conventional closure. Moreover, several eighteenth-century editions of Roxana actually have different endings. Both J. Paul Hunter and Stuart Sherman have addressed the ways that eighteenth-century writing refuses to conform to twentieth-century narratology's notions about closure and aesthetic form, which appear to derive from the nineteenth-century novel. Hunter agrees that eighteenth-century novels "are in fact additive, digressive, lumpy and resistant to closure defined in the generally accepted sense," but refuses to see these features as faults; rather they are "intentional, inevitable, and significant" to their power as narratives.⁶ Hunter goes on to list six ways in which eighteenth-century narratives offer resistance to closure, the most important for my present argument being the frequent occurrence of sequels and continuations. Stuart Sherman, in a surprising move, historicizes the sequence "tick-tock" by showing that the late seventeenth and eighteenth centuries knew only a continuous series "tick, tick...