Roman Taxation in the Hermopolite Nome: edition of P.Stras. 902.30-65 and 903 

The taxation dossier under investigation is from the Hermopolite nome of Egypt and dated to the first years of Trajan. It comprises three rolls of papyri, reconstructed from 166 remaining fragments, which are housed in the Bibliothèque nationale et universitaire of Strasbourg. They are now known as P. Stras. 901-903 (Chang 2014) and present a total length of c. 1500 cm. P. Stras. 901 containing 38 columns has now been completely published, while only the first 29 of 65 columns of P. Stras. 902 and a sworn declaration at the end of P. Stras. 903 have been edited.

The three rolls record the collection of taxes and public rentals set for the third year of Trajan, i.e. 99/100 (Chang 2014, 92-94). In P. Stras. 901 and 902, the taxes and rentals are levied on three neighbouring fiscal districts around three main villages – Sentryphis, Titkôis and Tertonkanô – in the southern part of the nome (Chang 2014, 80-81). P. Stras. 903 seems to concern only Tertonkanô (Chang 2014, 361). P. Stras. 901 and 902 are inter-complementary, for the former deals with taxation in kind booked on the accounts of public granaries, while the latter registers taxation in money entering the public banks. Together the two rolls offer a complete picture of the system of Roman taxation, with chronological (prior to Trajan’s reform)[i] and geographical (a under-researched area in the Hermopolite nome) coherence, as well as the major fiscal components – money and crops, granaries and banks, bases of taxation.

While the land constitutes the sole basis of taxation in P. Stras. 901, non-land bases appear also in P. Stras. 902, which presents a distinctive and peculiar textual structure (Chang 2014, 50-53). As in P. Stras. 901, the columns 1-21 of P. Stras. 902 are divided into three chapters according to the aforementioned three fiscal districts, but their entries have two different types. One type is introduced by a physical person, while the other type is presented by names of land or non-land taxes; an entry of the second type is always connected to a preceding entry of the first type. The column 22 of P. Stras. 902 is blank. From column 23 down to column 62 at least, the text is organized in another cycle of three chapters with the same geographical criteria, but each chapter is constituted of two sections: one section on land taxes and the other on non-land taxes (Chang 2014, 51 & n. 70). A hypothesis has been advanced to explain the entire textual structure (Chang 2014, 41-42): the first cycle (col. 1-21) treats metropolitans owning real estate in the villages, while the second one (col. 23-62) deals with the villagers (Chang 2014, 41; P. Stras. 902, 1, 4. n). In fact, capitation taxes, which can only be levied on physical persons residing in the districts, appear merely in the second cycle of P. Stras. 902.

The completion of the edition of P. Stras. 902 (Chang 2014, 10-11, 12, 15, 21-23) would verify this hypothesis, which at the same time would explain the relation amongst the three rolls: just as the assumed metropolitan-villager textual division of P. Stras. 902 for money-taxes, P. Stras. 901 and 903 (the whole columns of the latter remain to be edited)[ii] may concern respectively and distinctively two categories of tax-payers – people from villages and residents of the city – for impositions in kind.[iii] Moreover, through the examination of capitation taxes recorded in P. Stras. 902, demographical information could be gathered, which will be enriched by newly deciphered onomastic data (person names). These data shall be combined with those obtained from the edition of P. Stras. 901, the up-coming edition of P. Stras. 903 and from another contemporary long tax roll from Hermopolis housed in Vienna and to be published by Jaakko Frösén and Tiina Purola as CPR XI (solely on taxes in money).[iv] Based on the proposed editorial work, the present investigation aims at the study of taxation and the onomastics at the same time, and adopts a prosopographical approach rather than a quantitative one, in order to understand the relation between the absentee landowners and the villagers, within an intricate network of leasing and subleasing (Chang 2014, 85-91).

[i] For this reform deduced from and confirmed by papyrological sources, see Sijpesteijn 1965.
[ii] Chang 2014, 11, 12, 16 et 23. For further information, see Chang 2016, 176 & 184.
[iii] A preliminary transcription of P. Stras. 903 does not give out any communal agriculture in the rural area as attested in P. Stras. 901 (Chang 2014, 115-17). This may support the hypothesis about the complementarity between P. Stras. 901 on villagers and P. Stras. 903 on metropolitans.
[iv] A research framework questioning the tendency of naming people in the Hermopolite nome from the beginning of Hellenistic era down to the high period of the Islamic dominion has recently been set forward in Minnen 2015, 322-320.


Ruey-Lin Chang
Tel.: +43-1-53410-325 (ÖNB)