Tafila-Busayra: Crunching the numbers

I’ve been going through the most recent Burton MacDonald survey trying to apply some critical thinking skills to the data for the periods after, oh, say the lower Lithic, when MacDonald and his team lose interest. The published results, in a pleasantly sturdy volume titled The Tafila-Busayra Archaeological Survey 1999-2001, West-Central Jordan are exciting but also frustrating. For one thing, much of the information is available on a website! The TBASwebpage may not be pretty, but it allows you to copy and paste, for which everyone under the age of 35 will thank their deity of preference. The formatting is a little wonky, so I cleaned up the data I’m working with in Excel and put it up as a GoogleDoc for anyone interested.

The region was surveyed using several different methods, primarily random squares and purposive site-based survey based on aerial photos (see the sweet flickr page for the Aerial Photographic Archive for Archaeology in the Middle East) and earlier surveys, namely those by Glueck and Parker. I’m focusing on the major sites identified by MacDonald (78 out of 290), for which he helpfully gives a separate list.

On of the more interesting things MacDonald does is to distinguish between three landscapes (actually, there are two more sub-landscapes), which would provide a basis for doing a comparative study on what kinds of landscapes are being utilized for what kinds of activities during different periods. If you noticed the would, however, we’ve entered the realm of the hypothetical because MacDonald only provides the zone information for sites in or near the random squares, not for the purposive sites. Most of the major sites were purposively surveyed, which is why the final column for the zone information is full of holes.

While MacDonald publishes drawings of diagnostic material for many of the sites (hooray!), he does not publish the amount of pottery from each period found at the sites. From this presentation, one can only make binary presence-absence distinctions for the different periods. Although this is a little frustrating, perhaps this is all one really should try to conclude from surface surveys.

Some early-stage analysis:

There are many more Byzantine sites than there are for any other period. This can partially be explained by the nature of the region in the Byzantine period, which served as a militarized defensive zone dotted with forts and towers. The drastic drop in sites between the Byzantine/Late Antique and Early Medieval periods is basically what I’m going to talk about in my paper for the landscape archaeology seminar, some of which will get posted here in its inception, so I won’t dwell on that. Basically it seems to be the result of a combination of plague and economic and political instability in the late sixth and early seventh centuries, along with the disappearance of the main reason people had to live there (defending the frontier).

Perhaps more interesting are the sites that MacDonald identifies as churches, which have ceramics present from a number of periods after the conquest. It’s possible that these buildings were reused for other purposes, but it’s also possible that they were reused (or even constructed) as mosques. I haven’t looked more closely at MacDonald’s write ups of these sites yet, so this idea is probably hasty and wrong. But the long lives of the buildings is interesting nonetheless.

Looking at the sites not by number, but by the proportion of sites from each period that appear to have served different functions is also pretty revealing:

The church issue is clear here as well. One fifth of the (admittedly few) sites with Early Medieval ceramics are considered to be churches. This is not implausible, after all churches continued to be constructed for centuries after the conquest. But it also makes me wonder if some of these were actually mosques, and just typologically very similar to rural Byzantine churches or chapels. I’m thinking in particular of the small mosques associated with the rural estates at Hallabat and Qastal, but why these would or wouldn’t be necessary in the Tafila-Busayra region is a question I still have to consider more carefully.

What’s kind of neat is the fact that, overall, a similar percentage of the sites was serving any given function over the course of 1900 years. One notable exception is the Roman and Byzantine emphasis on defensive structures, which are unnecessary in the Early Medieval period but become relevant again in the Later Medieval period (a connection to the crusades is likely but it’s impossible to rule out other causes).

Expect a site time line and a map, perhaps tomorrow.