Memory and Description in the Ancient Novel

In the introduction to his translation of The Adventures of Leucippe and Clitophon, a Greek novel of the second century C.E., Jack Winkler gives the following advice to its readers:

The unanswerable enigma of its contradictory style should be enjoyed directly as a lascivious surface and nothing more, making us conscious that the quest for depth, for meaning, and for unity is a fraud of the ages. That cannot quite be the author's intention, since he could not have foreseen the time when tentative endeavors for unity of meaning, perspective, and point of view would one day have created a system of seeing and reading that knew no other possibility, but as children of that system it must be part of the author's meaning for us. Otherwise we will misread his stressful irresolutions as bad rather than purposefully ineffable.¹

Winkler here faces up to a difficult fact about many kinds of ancient literature: the circumstances of their production and circulation are either unknown or alien to us, and hence our ordinary modes of reading are necessarily ahistorical to some degree. In the case of the ancient novels, the application of our own standards for a well-made story necessitates either the judgment that they are incompetent, or it necessitates a more subtle reading that recuperates a "unity" of questionable historical provenance. The old battle between the "Analysts" and "Unitarians" in Homeric scholarship is analogous: both sides took for granted the same standards of reading, the former rejecting various parts of the received texts of Homer as the interpolations of an incompetent Arbeiter, the latter spinning out ever more complicated justifications for the inclusion of those suspected parts. It is now well known how the Parry-Lord theory of oral composition provided a third way of viewing Homer, one in which the process of production and circulation itself provided the basis for viewing the texts as self-consistent in a different way. Is it possible to find a third way of looking at the ancient novels based on a "theory of prosaic composition" analogous to the theory of oral composition for Homer? I would like to advance such a project by investigating the status of description in the novel of Achilles Tatius to which Winkler alludes in the quotation above, taking my cue from two studies that correspond roughly to an "analyst" and a "unitarian" approach to Leucippe and Clitophon. Dorit Sedelmeier identifies a clear structure in the novel once the various descriptions and digressions have been discounted.² Shadi Bartsch, on the other hand, focuses on these very elements of the text to identify a complex authorial strategy of hermeneutic hide and seek in the novel (Bartsch 1989). Each of them takes for granted that the meaning of the novel must be grasped as a whole, that the parts either resonate with the whole or are to be set aside as extraneous or "merely" ornamental.

In assuming these positions, Bartsch and Sedelmeier implicitly rely on the general view that description, when it occurs in narrative, is either ornamental or symbolic: in the first case, description produces enargeia, vivid visual realization that contributes to some specific rhetorical impact; in the second, description serves to foreshadow events, delineate character, or otherwise reflect or figure the key interpretive issues raised by the text.³ In the latter case, description is sometimes seen as a miniature version of the themes or poetic design of the text, a mise-en-abyme "reduplicating, reflecting, or mirroring (one or more than one aspect of) the textual whole."⁴ Such accounts, however, often take for granted an instrumental view of the switch from narrative to description: description is selected at some point in a narrative from a set of available options because it suits the author's overall purpose best. I will argue that in Leucippe and Clitophon descriptions can perform a generative role in the unfolding of the story: rather than "reflecting" or "mirroring" the narrative as a whole, descriptions sometimes present a tentative "first draft" or preliminary sketch of it. That is, rather than being elements of...