Houses and Society in the Later Roman Empire by Kim Bowes (review)

This short book, one of a new Duckworth series titled “Debates in Archaeology”, aims to give the reader a clear state-of-the-art view of where the study of late Roman houses now stands. As such, it is more about the study of houses and about different methodological approaches to the topic than about the houses themselves. According to the publisher’s blurb, the book is targeted equally at both “students and serious scholars” (one hopes these are not mutually exclusive categories). In practice it is geared more towards graduate than undergraduate students; it is not a book for the casual reader.

Bowes’ methodological territory is quickly staked out. The study of late Roman houses, she argues, has moved away from traditional text-based interpretations towards a more archaeological and theory-based approach(19). Bowes acknowledges the important role played in this process by two earlier scholars, Yvon Thébert and Simon Ellis (the latter “arguably the most influential scholar on the later Roman house,” 28). Nonetheless the model of Late Roman housing proposed by these scholars, she claims, is misleading. Whereas they argued for a bold “new” style of Late Roman élite housing based on new patterns of social behaviour, Bowes asserts rather that the houses of the rich in late Antiquity were built to accommodate patterns of social behaviour which were little changed from previous centuries.

This is perhaps correct but, as with the work of Thébert and Ellis, one has to question the evidentiary basis on which Bowes builds her argument. Her source base, it appears, is almost exclusively the material blueprint of late Roman domestic life, the structures and instrumenta (‘pots and lamps,’ 85) of excavated houses and villas. But this gives at best only a partial picture. Patterns of social behaviour are not easily reconstructed from house plans and small finds alone. A large hall with an apse and an opus sectile floor can be interpreted many different ways: as a reception room, a dining room, a library, an auditorium, even a courtroom. How are we to choose? Interpretations based on spatial analysis or artifact distribution can only take us so far. They do not explain social behaviour or household ritual. Ultimately we are going to need every scrap of help we can get from contemporary written sources, however flawed these may be. Of course written sources are problematical; writers have agendas, they have imaginations. But this is not a reason to devalue them; what matters is how we use them.

This is where Bowes runs into difficulty. She wants to distance herself from the written evidence which she views as unreliable and potentially misleading (e.g. her comments on Sidonius’ description of élite dining, 57-58) but by doing so she deprives herself of a rich source of cultural information which can provide important clues to understanding why houses were built the way they were. It was precisely this problem which undermined the earlier work of Thébert and Ellis. In his analysis of the so-called “Palace of the Dux” at Apollonia in Libya (used by Bowes as the front image of her book), Ellis introduced all sorts of ideas about ceremony and hierarchy within the late Roman house but never explained what exactly these ideas involved. Bowes tends to follow along the same path; there is little discussion here of the actual social practices which influenced the design of late Roman houses; no discussion of late Roman domestic protocols or of how these protocols may have varied from one region to another. Instead, what we have is a good theoretical discussion of late Roman housing based squarely on the material evidence afforded by excavation but little comment on the lives of the individuals and households for whom the houses were created.

Nowhere is this more apparent than in the author’s frequent comments about apses. In late Roman houses apses were everywhere: “[a]t Carranque in...