One of the things I try to do in my book on early Christian manuscripts is survey some of the contexts in which ancient Christian books have been discovered. A set of examples that I wasn’t able to treat in detail in the book is the small group of Christian manuscripts found at the Egyptian town of Karanis. One of these pieces is a fragmentary leaf of a codex that contained the Psalms 32-33 in Greek (P.Mich. inv. 5475b, LDAB 128586) that was found in the courtyard of a house.
It was published in 2009 in a detailed edition by Gregg Schwendner, whose deep knowledge of Karanis made for an excellent discussion of the fragment’s context.
Karanis is located in the Fayum and was excavated between 1924 and 1935 in a project run by the University of Michigan. In a decade of work, some 4,000 papyri and 6,000 ostraca (inscribed ceramic fragments) were uncovered. The excavated materials are now divided between Ann Arbor, Cairo, and regional museums in Egypt. The Psalms fragments are among the pieces now in Cairo.
By the time the Michigan team began excavating, a large portion of the kôm (or mound) of Karanis had already been destroyed in the nineteenth and early twentieth centuries. The reason was the removal of sebakh, a material consisting of the remains of abandoned ancient structures made of mud-brick (that is, Nile silt mixed with straw) as well as accumulated plant and animal refuse. The high levels of nitrates in sebakh made it ideal for use both as a fertilizer and as a component in the making of gunpowder. The central area of ancient Karanis had thus been destroyed by the harvesting of sebakh. What the Michigan team excavated were the remaining east and west sides of the city.
In terms of record keeping and documentation the Michigan excavations at Karanis were considerably more advanced than those undertaken earlier in the century by Grenfell and Hunt at Oxyrhynchus (discussed here and elsewhere on this blog). The photographic record of the Karanis excavations is quite extensive (including even photos of the daily life and characters at the dig, like Plupy). Excavated items from Karanis that were saved (including papyri) were documented–find spot and “layer”–and for many parts of the site, there are fairly detailed top plans and occasionally excellent elevation drawings. It is probably better to think of the excavators’ periodization of the site in terms of “phases” rather than “layers,” because they differentiated periods by architectural change rather than soil stratigraphy. In any event, this contextual information was not often utilized in publications of papyri from Karanis (with the exception of some of the work of Elinor M. Husselman), but in the 1990s, Traianos Gagos and Peter van Minnen pioneered the so-called “house-to-house” approach to recontextualizing the papyri within their places of discovery. Schwendner’s edition of the Psalms fragments refines this tradition.
As Schwender discusses in his edition, the Psalms fragment is one of several items with the label “1928 C87 K item A.” This tells us when and where it was excavated (1928, in House number 87 of the excavators’ C phase, in a courtyard labelled K), in the northern part of the eastern kôm of Karanis.
But this information alone might not be as helpful as it first appears. As Traianos Gagos emphasized, “the stratigraphic record of Karanis is very unreliable, because the site was very disturbed already before the arrival of the excavators in 1924.” Thus, it can’t be absolutely excluded that the Psalms fragment is a later intrusion into this context. But pans of the site indicate that our papyrus was found in something resembling a sealed context. That is to say, in the excavators’ top layer, a street and several domestic structures were built on top of House C87. C87 and its courtyard were uncovered only after the removal of these upper surfaces:
Schwendner limited his contextualization of the courtyard to the papyrus finds in this area, setting aside other classes of finds. The papyrus pieces from the C87 assemblage that have been published or publicly described are as follows:
P.Mich. inv. 5470, unpublished document, no date
P.Mich. inv. 5471a+5478c, P.Mich. 9.542 declaration, before 212 CE
P.Mich. inv. 5472+5475b, LDAB 655, papyrus roll of Demosthenes, assigned to 2nd cent. CE
P.Mich. inv. 5473, SB 12.10797 petition, 237 CE
P.Mich. inv. 5474, SB 5.7563 application, 207 CE
P.Mich. inv. 5475a, LDAB 144439 papyrus roll of Callimachus, uncertain date
P.Mich. inv. 5476a, SB 24.15886 document, assigned to 3rd cent.
P.Mich. inv. 5476b, SB 24.15963 sale of land, 131-168 CE
P.Mich. inv. 5477a, SB 24.15878 census declaration? assigned to 2nd or 3rd cent. CE
P.Mich. inv. 5477b, SB 24.15889 receipt, 265-266 CE
P.Mich. inv. 5478, SB 22.15777 letter, 271-272 CE
P.Mich. inv. 5478, SB 22.15771 account, assigned to 3rd cent.
P.Mich. inv. 5479a, SB 24.15884 list, 253-256 CE
P.Mich. inv. 5480a, document, assigned to 3rd or 4th cent.
P.Mich. inv. 5480b, literary papyrus? assigned to 3rd or 4th cent.
All but one of the items that have secure dates fall within the first three quarters of the third century. The context included one second-century document and a roll of Demosthenes (palaeographically) assigned to the second century. Two pieces, one documentary and one literary, are assigned to the third or fourth century. But, as Schwendner cautions, aside from “the published texts from Karanis one must bear in mind the very many unpublished (difficult, fragmentary) texts whose eventual publication could change the apparent certainty of the list above.” That is to say, each of the four-digit inventory numbers can refer to a whole plate of fragmentary papyri, only a small portion of which have been published. The profile of the context is subject to change. There are also the other materials found in the same context (ceramics, etc.), which, as far as I know, still await analysis. These materials could also change the picture a bit.
The mix of papyri found in the courtyard of C87 suggest a dump context, and the dated pieces suggest accumulation over the course of the third century perhaps extending to the early fourth century. Because the Psalms papyrus seems to have been found as discarded trash, we can’t know the precise setting in life of this piece. But we can know its more general setting (a relatively remote Fayum town), and thanks to the record keeping of the Michigan team, we have a decent sense of its probable date (third century).
Boak, Arthur E.R. and Enoch E. Peterson, Karanis: Topographical and Architectural Report of Excavations During the Seasons 1924-28. University of Michigan Press, 1931.
Gagos, Traianos. “The University of Michigan Papyrus Collection: Current Trends and Future Perspectives.” Pages 511-537 in I. Andorlini, et al. (eds.), Atti del XXII Congresso Internazionale di Papirologia, Firenze 1998. Istituto Papirologico «G. Vitelli.» 2001.
Husselman, Elinor M. Karanis Excavations of the University of Michigan in Egypt 1928-1935: Topography and Architecture, A Summary of the Reports of the Director, Enoch E. Peterson. University of Michigan Press, 1979.
Schwendner, Gregg. “A Fragmentary Psalter from Karanis and its Context.” Pages 117-136 and 311-312 in C.A. Evans and H.D. Zacharias (eds.), Jewish and Christian Scripture as Artifact and Canon. T&T Clark, 2009.
Schwendner, Gregg. “Literature and Literacy at Roman Karanis: Maps of Reading.” Pages 991-1006 in Frösén et al. (eds.), Proceedings of the XXIV International Congress of Papyrology, vol. 2. Societas Scientiarum Fennica, 2007.
van Minnen, Peter. “House-to-House Enquiries: An Interdisciplinary Approach to Roman Karanis.” Zeitschrift für Papyrologie und Epigraphik 100 (1994), 227-251.