Writing Wednesday: Contextualizing Lamps from Polis

Over the first week of my winter research leave, I’ve been pressing to prepare a draft of an article on our work at Polis. In particular, I’m looking to publish the results of our analysis of the area around a levigation pool, a kiln, and an assemblage of Roman period lamps. Much of this is based on a 2024 ASOR paper which you can read here.

One of the most interesting components of this article is an assemblage of second century lamps that appear to be stratigraphically unrelated to the levigation pool and kiln as features, but seemingly must have derived from the area. This provides a nice opportunity to explore the limits of archaeological epistemology. While the character of our data makes it impossible (or at least very unlikely) that I can conclusively link the lamp assemblage with the features associated with ceramic production, it is nevertheless highly plausible that these things have some connection.

Another little bonus is that we’re leaning into using Airtable not only to work collaboratively on our data, but also to do some basic analysis. This means we can share our work both with one another and in preliminary ways with readers here. Here is a sneak peek at our context pottery data from E.F2:S06 levels 13, 14, and 15. 

Here’s some of my preliminary analysis.

The Lamps

This consideration of the depositional processes active around the South Basilica presents context for a deposit of lamps found amid strata of predominantly Late Roman material immediately to the east of the basilica apse. Indeed, preponderance of lamps in a single area led the excavators to isolate this area and excavate it as a separate level (Level 14). In terms of stratigraphy, the lamp deposit (Level 14) was not distinct from the levels above it (Level 13) or below it (Level 15). As we have noted, excavators associated these levels with the basilica leveling fill potentially disturbed by lowest course of the apse foundations. The material in Levels 13, 14, and 15, in general dated to no earlier than the Hellenistic period and no later than the Late Roman. Within that chronological range, the material was heterogeneous with the exception of the lamp deposit.

This assemblage produced over 50 lamp fragments and a number of complete lamps. The lamps were largely two forms: Q2460 and Q2465 from Bailey’s 1975 British Museum catalogue; these are also known as Loeschcke Type V lamps. These lamps have a flat base, a small ring handle and volutes on either side of a short spout. Q2460 has a blank disk and undecorated shoulder whereas Q2465 features a peacock on a pomegranate bough and ovules on the shoulder. Bailey associates these lamp forms in the British Museum with the Kitchener/Hake excavations at either Salamis or Kourion and Oziol notes similar lamps from the French excavations from Salamis. They both date these lamps on stylistic grounds to the first half of the second century AD. This makes them loosely contemporary with the residual Roman period material found in the basilica leveling fills and, as we will demonstrate below, the material associated with the pottery manufacturing activity in the area.

It is not just the quantity of lamps concentrated in one depositional context, but the consistency of the lamps appearing in this context that set them apart from lamps identified elsewhere at Polis and in the area of E.F2 specifically. In fact, the types of lamps recovered from the fill over the kiln were exceedingly rare elsewhere in E.F2 with only one example of Q2460 (LA340) and Q2465 (LA485) type lamps appearing elsewhere. More than that, many of the lamps recovered from the fill over the kiln showed no signs of use or blackening around the wick hole. They also included several generations from the same mould as is evident in the deteriorating sharpness of the peacock on the bough. The assemblage of lamps from this assemblage is also remarkably homogeneous compared to fragments of lamps found elsewhere in E.F2. There are two contexts of particular note that produced a large number of inventoried fragments of lamps. Trench H10 to the south of the basilica produce a large number of very small lamp fragments with many being undiagnostic or dating to the Classical-Hellenistic period. There were no complete lamps. Trench M10 to the east of H10 produced numerous fragments of a handful of Roman period lamps including a well preserved fragment of a pornographic lamp (LA796). Like the lamps recovered from trench H10, however, the fragments produced few examples of complete lamps and these complete lamps appeared in relatively wide range for forms. In short, the other areas of E.F2 where excavators encountered a large number of lamp fragments (and complete lamps) produced more diverse and fragmentary assemblages of material that showed so no signs of the integrity of the S06 deposit.

Levels 13, 14, and 15 produced material that was contemporary with the lamp deposit. In particular, these levels produced a substantial number of diagnostic Cypriot Sigillata forms with a clear bias toward the so-called “Late Series” of CS dating to the later 1st through 2nd century. These included forms P11 and P29 which appear in predominantly 2nd century contexts at the Paphos Agora (Kajzer and Marzec 2020, 252) as well as other late forms in P30 and P41 dated by Hayes to the 2nd century (Lund CPSP; Hayes 1985f, Hayes 1991) and a P12 bowl which tends toward the 2nd century. Along with these 2nd century forms are earlier and more common forms P22 and P28 and a rare form P34. This material appeared alongside cooking wares in forms that Rowe has associated with Western Cypriot manufacture and dates to deposits in the 2nd century. This material complements the general impression left by the fine wares. While the samples are small, the percentage of ESA in these level is lower than elsewhere across E.F2 and while CS percentages remain similar, the prevalence of late forms suggests that the Roman period material in these levels may well reflect the presence of a 2nd century context amid the otherwise much later leveling fill. This is consistent with the presence of a lamp assemblage which Bailey has dated to between 50 and 150 AD. In fact, it seems plausible to date the lamps to the 2nd-century component of these deposits recognizing that the residual material forms a lens in an otherwise later deposit.