Free Death to Tyrants! by David Teegarden Page B

context, however, I also provide data on the success of other regime types during those same periods.
    The data are culled exclusively from Hansen and Nielsen’s Inventory of Archaic and Classical Poleis (2004), in particular from its appendix 11. I entered the data into a database that allows simple yet fundamental queries relating, most importantly, to the total number of cities for which we have regime-type information and the percentage of those cities that experienced the different regime-types in a particular period of time. The result, I believe, is a fairly compelling rough sketch of the relative success of the major regime types over several centuries.
    I have not examined the evidential basis for all of the data that I included. Acknowledged experts wrote the entries for each of the poleis contained in the Inventory . If they suggest that city x experienced a democracy (or another regime type) in time y , I took their word at face value.
    I have divided the data into periods of a half century. Using shorter units of time is unfortunately not practicable. There is one problem, however, in dividing the data even into periods of a half century. One of Hansen and Nielsen’s temporal categories refers to the “middle” of a century, which corresponds to the forth to sixth decades of that century. For example, C7m (“middle of seventh century”) refers to the years 660–640, C4m refers to the years 360–340. Yet Hansen and Nielsen also have temporal categories that refer to the first or second half of a century. For example, C7f (“first half of seventh century”) refers to 699–650, while C7 (“second half of seventh century) refers to 649–600. Likewise C4f refers to 399–350, while C4 refers to 349–300. The difficulty (for this appendix) arises when there is evidence for a particular regime type in a city during the “middle” of the century, since that city could have experienced that regime during the first or the secondhalf of the century, or both. I have decided to have it refer to both halves of the century. Thus, for example, if there is evidence for democracy in city x in C4m, I mark it as experiencing democracy in both the first half and the second half of the fourth century.
    I must also comment on what I mean by a city having “experienced” a particular regime type. First, that experience need not have been long in duration: if a democracy controlled city x for a month in, for example, C4f, that city experienced democracy in the first half of the fourth century. Second, a polis can experience several different regimes types in one half century. Thus the combined number of regime types experienced by all cities in a particular half century will be greater than the number of poleis for which we have regime type information for that half century; likewise, the sum of the percentages of cities experiencing all of the various regime types in a given half century will be greater than one hundred. For example, in a particular half century we might have regime-type information for ten cities, each of which experienced democracy, oligarchy, and tyranny. In that case there would thirty different regimes experienced by ten poleis, and the sum of the percentages of cities experiencing the various regime types for that half century would be 300: 100 percent experienced democracy, tyranny, and oligarchy.
    I here present two charts that efficiently capture the data. Figure A1 indicates both how many cities are known to have experienced a particular regime type in a given half century and the percentage of cities known to have experienced that particular regime type during that half century (out of all cities for which we have regime-type information for that half century). Figure A2 indicates the percentage of geographic regions known to have contained at least one city that experienced a particular regime type in a given half century. The two