Scott Arcenas, University of Montana
Uncertainty is one of the greatest methodological challenges facing scholars of the ancient world. Despite this undeniable fact, however, efforts to address the challenge systematically have been rare (Jew and Lavan). To the extent that ancient historians have focused on the problem of uncertainty, they have tended to operate on a macro scale and have overwhelmingly considered quantitative questions (Garnsey; Hopkins; Scheidel and Friesen; Lavan; Lavan, Jew, and Danon). Here, I address a complementary issue that has received less attention: how to handle uncertainty when writing narrative histories on a local level—in particular, the political histories of archaic and classical poleis.
In this paper, I use the term “narrative” expansively, to include even rudimentary stories of change or stability over time: for example, the bare assertion that “from 379 onward, Thebes was a democracy” (Hansen: 455). I use “uncertainty” in a limited sense, focusing exclusively on significant uncertainties that might lead us to interpret the events under consideration in substantially different ways.
The paper has three parts. In Part 1, I examine the political history of mid-fifth-century Thebes. First, I present the (potentially) relevant evidence, with particular attention to the high levels of uncertainty involved in its interpretation. Next, I consider how existing scholarship has interacted with the aforementioned uncertainties. As a rule, historians have sought to obscure, minimize, and/or resolve uncertainties as a means of developing coherent narratives. In so doing, they have created a smorgasbord of contradictory interpretations, none of which is likely to be correct (Gomme, Gehrke 1985, Buck, Robinson, Schachter; cf. Mackil, who wisely eschews narrative). Finally, I show that the inaccuracies of existing narratives should not be understood as products of bad scholarship. Rather, they are byproducts of the fact that evidence for Theban political history c. 460-440 is so uncertain that even the best possible narratives are almost certain to include fundamental inaccuracies.
In Part 2, I extrapolate my findings from Part 1. I show that the evidence for Theban political history c. 460-440 is both better and more complete than the analogous evidence for all but a few dozen polis-periods of equivalent length; and that, consequently, the levels of uncertainty involved in reconstructing the political histories of most polis-periods—including frequently narrated examples, such as late-fifth-century Argos (David, Gehrke 1985, Robinson) and mid-fifth-century Miletos (Gorman, Gehrke 1980, Rubinstein)—are even higher. On the basis of these observations, I argue that the methodological problems diagnosed above are not unique to mid-fifth-century Thebes. Nor are the resulting inaccuracies. Rather, they compromise most narrative political histories of individual poleis.
In Part 3, I use the results obtained in Parts 1-2 to propose a series of unsettling conclusions: that most of the narratives on which Greek political historians rely are fundamentally unreliable; that it may not be possible to narrate political histories of more than a few dozen polis-periods responsibly; and that, consequently, we need to develop new approaches to writing Greek political histories that foreground the problem of uncertainty, rather than seeking to obscure it.