Skip to main content

Reports of lifespans on Roman tombstones long drew attention as evidence bearing on various demographic questions. Since the 1960s, however, this body of evidence has been increasingly abandoned (and probably rightly so) in the face of internal statistical indications which suggest both that the individual texts are unreliable and that the aggregate is unrepresentative for most purposes. The present study focuses on a potentially more valuable subset of this evidence consisting of stones that spell out timespans to the day or the hour (or even fractions thereof). In fact, even this set has features that limit its demographic value, but at the same time they also open up an entirely different set of cultural historical questions and lead us to ask how to understand Roman practices of quantification in general. This paper focuses on two phenomena: (1) the use of menses, dies, and horae understood as discrete, concrete events, in addition to/instead of as abstract units of measure, and (2) an ideology of what I call “precisionism,” that is, a pressure towards precision not through greater substantive accuracy, by types of formal display.
Events vs. units. Roughly speaking, most units of time arise by metonymy. A real-world event (e.g. a rotation of the earth on its axis, orbit of the moon around the earth, orbit of earth around the sun) is equated with the duration of that event (e.g. day, month, year). Once the metonymy takes hold, those durations can be lined up (and multiplied or sub-divided as necessary) as standardized units to measure the duration of any other event. The behavior of some of these units in the tombstones under observation suggests the metonymic reading is less dominant than might have been expected. (a) In spans listing a number of days, these often fall between 32 and 61, but almost never more. Such texts represent periods that span two calendar months without including either. (b) The number of hours listed never goes beyond 11 (except in a few cases that distinguish “hours of the night”). This suggests dies are days in the sense of periods of daylight rather than units of 24 hours. [(c) Though it is not reflected in the epitaphs, the seasonal variability of the length of the hour makes that a twelfth part of the day qua event, not a fixed, transposable unit of time.]
Precisionism. The degree of precision reported on epitaphs is surprisingly ill corollated to its actual availability or correctness. (a) Even high-precision lifespans show signs of age-rounding (and, in fact, of rounding in months and days as well). That is, much of this precise data is fabricated or approximated. (b) Statistical “heaping” suggests a lot of hidden approximation (of the sort of English “a couple” or Italian “un paio,” but much more extensively used that those). Unlike the modern instances, however, the Roman approximations do not serve as simplification, nor are they readily distinguishable from more literal formulations for the benefit of readers who might want to know the difference. (c) Lifespan is rendered with more precision than marriage span, and that in turn more precisely than length of military service. This is the opposite of what one might expect on the basis of availability of accurate information; instead precise information is more culturally salient in certain contexts. In short there is a bias, at least in certain contexts in favor of representing timespans with precision, regardless of accuracy. I close with some speculations on what might give rise to this precisionism.

[I would very much prefer to present this, if accepted, as a poster.]