You are here

The Impact of Evidentiary Bias on Macro-Level Approaches to Greek History

Scott Lawin Arcenas

Stanford University

The last three decades have seen vast increases in both the amount and the accessibility of information concerning the roughly 1,100 poleis inhabited by the ancient Greeks. Consequently, it has become possible for the first time to write histories of ancient Greece that are truly panhellenic in scope. To analyze this information responsibly, however, historians need to develop new methods to detect, measure, and compensate for the deep biases of the evidentiary record. In this paper, I introduce one such method.

The paper comprises four parts. In the first, I introduce four metrics for what I call prominence: i.e. the total amount of information available concerning a given polis. As my metrics, I use i) the number of columns allocated to each polis by IACP (cf. Fleck and Hanssen 2013, Ober 2015); ii) the number of named individuals assigned to each polis by the editors of LGPN; iii) the number of times each polis’ city-ethnic (IACP, pp. 62 - 63, 1253 - 1289) appears in a selection of core texts; and iv) the number of years in which citizens of a given polis play a role in at least one narrative historical source.

In part two, I evaluate my metrics. I show that all four have significant flaws, taken individually. Collectively, however, they provide an accurate measure of prominence in both absolute and relative terms.

In part three, I demonstrate the utility of my metrics by using them to reevaluate a fundamental aspect of Greek society: stasis, commonly translated as civil conflict or civil war (on the importance of stasis, see esp. Gehrke 1985, Loraux 1997, Hansen 2004, and Gray 2015). After introducing a database that contains all 413 staseis known to have occurred between 500 and 301 BCE, I identify four significant trends in the distribution of attested cases: attested staseis are rare; they are heavily concentrated in a small set of 30-40 poleis; they are heavily concentrated during periods of large-scale wars; and their numbers increase substantially with the start of the Peloponnesian War (431 BCE).

Existing scholarship has tended to assume that the aforementioned trends are historical. This assumption has, moreover, fundamentally influenced the substantial literature that seeks to understand ancient Greek political violence on a macro-level scale (see esp. Lintott 1982, Gehrke 1985, Winterling 1991, Berger 1992, and Austin 1994; but see Hansen 2004). Using the metrics developed in the first part of my paper, however, I show that all four trends—and thus, by extension, the literature that assumes their historicity—are products of evidentiary scarcity and evidentiary bias rather than historical phenomena.

In part four, I highlight the four most significant payoffs of my paper. First, I show that that the distribution of attested staseis cannot serve as a reliable indicator of the frequency with which actual staseis occurred—whether in absolute or relative terms. Second, I show that staseis were both more frequent and more evenly distributed than existing scholarship assumes. Third, I draw attention to the harms that necessarily arise when historians study premodern societies on a macro-level scale without accounting for evidentiary scarcity and bias. Fourth, I introduce both a new methodology and a new set of tools to obviate those harms.  

Session/Panel Title

Principles and Practices of Greek Historiography

Session/Paper Number


© 2020, Society for Classical Studies Privacy Policy