Why would a temple in swampy ground be dedicated to Artemis?
There was an extremely interesting paper in a recent edition of Antiquity, Rocks, views, soils and plants at the temples of ancient Greece by Gregory J. Retallack. He’s been looking at the topographical setting of Greek temples, along with their orientations and the local geology. What he’s concluded is that if you want to understand Greek temples you need to get to grips with the local soils. Along the way he dismisses an idea which is pretty central to my thesis, so bear in mind the commentary below may be tainted with my own bias.
As it happens if I want to dismiss Retallack’s work then I’ve got quite a task. He’s examined eighty-four classical Greek temples, eighty-three in Greece and one in Paphos, Cyprus and examined the soils associated with the temples. This is an extremely good idea, and I’m surprised that someone hasn’t done it before. Geology might me easily overlooked, but soil is essential for agriculture. In the case of the development of farming in Europe it initially spread over löss soils as part of a package we call the Linearbandkeramik Culture. Retallack shows that the ancient Greeks were keenly aware of various types of soil and their properties. So looking for correlations between cults and soils is a worthwhile project.
What must then have been a large amount of work is tabulated into about a page and a half of results, though there are extensive supplementary materials. Along with the deity and the location there’s also the geology, topography, soil and vegetation. He notes that there’s no real record of climate change for Greece so the natural vegetation should be inferable from modern observations. If you’re wondering what the difference between geology and soil is, the geology refers to the bedrock and the soil is what overlays it. Bedrocks don’t uniformly underlie the same soils. From all of this he’s able to conclude that certain soils tend to be associated with certain gods.
The soil type which should come as least surprise to anyone is that Hades isn’t associated with a soil at all, but rather rock. Hades is god of the underworld, and his sanctuary at Cape Tainaron, Laconia, is in a sea cave while his sanctuary at the brilliantly named Necromanteion, Epirus, is on a rocky hill-top. What you want from Hades site is a connection with the underworld, so rocky crevices which plunge into the bowels of the earth are just the sort of thing you’re looking for. As I recall there’s a Plutonium (another brilliant name) in Hierapolis, Turkey, built over a geological fault which emits noxious gases, so that seems perfectly sound. He also attributes Persephone to rock crevices, and this is something I’m a bit more wary of.
From the table, the attribution makes sense. It also makes mythological sense. Hades abducted Persephone and took her to the underworld to be his wife. So the queen of the underworld should be associated with rock crevices. However, Persephone isn’t just associated with Hades. She was also the daughter of Demeter and with her she was a vegetation fertility goddess. She was only said to spend a third of the year with Hades. This was acknowledged by the Greeks who didn’t just make temples to Demeter but also to Demeter and Persephone. None of the temples listed by Retallack have this shared attribute in the main paper, which seems a bit odd. On the other hand there is a mention of a sacred grove of Demeter and Kore (another name for Persephone) at Cabirion, Boeotia. This might matter because often the reporting of the dedications of temples can be quite poor. I recently read a paper which attributed Temple F in Agrigento to Juno Lacinia, even though it’s been known for a century that this attribution is wrong. Names have a tendency of sticking. I don’t think this is a serious problem, but it does suggest where more research could be done.
The other factor I find problematic is that Retallack says that the idea that Greek temples faced sunrise on the feast day of their god is falsified by the data he’s collected. I can’t recall Scully saying that this was universal, though I’ll concede Richer may have. The reason I think Retallack is wrong comes from the same data in the table. Where he can discern an orientation, he divides the temples into eight categories for the eight major compass directions. If there was a random scatter then, of the forty-two temples he has directions for, you’d expect around six to face east, as opposed to any of the other seven directions. There might be a few more or less, but the chances of getting more than ten facing east by random chance are around 100-1. In fact from that sample set over half, twenty-six, face east. That doesn’t mean that an astronomical orientation was essential for a temple, but it does indicate a strong bias. Again, rather than undermining Retallack’s case, I think it potentially makes it more interesting.
If we do have a bias to point temples towards the range of sunrise, then we can start asking what’s special about the temples where there isn’t this bias. In the case of the temples in the north-east and south-east categories they may still be facing the extremes of sunrise. The temples of Apollo at Ptoon, Boeotia and Dionysos, Eretria, Euboea, the two temples from the survey which can definitely be said to face west, clearly cannot be facing sunrise. In the case of the temple of Dionysos at Eretria, the temple is on a creek floodplain, in Xeroll. If, like me you haven’t a clue was xeroll is, it’s the sort of soil associated with pastures. Retallack explains that because it’s stable and fertile with a mix of acid and alkaline earths. It’s the sort of soil you’d want for grain, associated with Demeter or the sort of grapes which the ancient Greeks used to make wine, beloved of Dionysos. Knowing that, we can see one reason why the temple was built where it was, but can also start asking other questions? Was there no suitable land in Eretria to use that would have allowed an easterly alignment? There may have been, the temple of Apollo there faced south-east. So what was the temple of Dionysos looking towards? Why wasn’t the temple dedicated to Demeter, who is associated with xerolls too? Retallack isn’t shutting down discussion but instead opening up new opportunities.
Another example would be Artemis. Temples of Artemis are associated with Xerept and Orthent soils. Xerept is a thin rocky soil and Orthent is what you get when Xerept is eroded and there’s hardly any soil at all. Retallack limits discussion to the nature of the soils themselves. That’s a shame because it is worth contrasting with Cole’s recent discussion of Artemis in her book Landscapes, gender, and ritual space. Cole says that Artemis is goddess of boundary places. She’s found in the mountains, but in the mountain passes rather than at peak sanctuaries. In the ancient Greek world the boundaries would be in the mountains and where the soil was poor. If that’s the case then the association between Artemis and Xerepts at peripheries could also make her the natural choice for Xerepts in more central locations. I’d accept that Brauron isn’t the most central location in Attica, but at the same time it’s not hugely out of the way. It’s very convenient if you travel along the coast. Despite this, Brauron was an important sanctuary for Artemis. Retallack, who states that Brauron is built on Xerept, may have opened up the possibility of explaining why the site was connected with that particular goddess.
The other way that Retallack is opening discussion is that he published his paper in Antiquity. This might seem obvious, but if you want to change the way archaeologists think about something, you have to put your ideas in a place where archaeologists will read them. Not everyone does this. My favourite example is a paper from the Journal of the Royal Society of Medicine, ‘Stonehenge: a view from medicine‘ by Anthony M. Perks and Darlene Marie Bailey. Whether that paper is right or wrong is irrelevant, because any researcher looking for papers on Stonehenge is going to be looking in the archaeological literature. Similarly Retallack could have chosen to place his paper in a Soil Science journal where it could be seen by his colleagues. Instead it looks like he wants his work to make a difference. Kudos are due to Antiquity too, who could have simply turned down the paper as not being archaeology. Instead they’ve produced something which is clearly relevant to non-geologists and still produced the materials and diagrams of soil profiles that people who want to more closely at the work will need.
I’m not convinced it’s a universal solution for all Greek temples. In the case of somewhere like Selinunte there’s temples to (probably) Hera and a temple to Apollo and either Athena or Artemis within a few yards of each other. In the case of the many towns which had just the one temple in contrast it could well explain why they dedicated their temple to one particular deity, like Aphrodite at Akrai or Demeter at Eloro.
For another opinion on Retallack’s work, George Monbiot has beaten me by about three and a half years.
