The Lost Art of Stratigraphy: Part 1 - Introduction, Mike Smith

Today I’m talking about the lost art of stratigraphy. It is an area of archaeological practice that’s quite dear to my heart. I grew up reading Mortimer Wheeler’s Archaeology from the Earth and of course Wheeler stressed the importance of stratigraphy in terms of understanding the relationships between deposits across a site. I also had the opportunity to work for a while on the British excavations at Danebury hill fort, an Iron Age hill fort in southern Britain, where Barry Cunliffe was renowned for his careful stratigraphic excavation of pits, earthworks and other features. As an undergraduate student, I also worked at Early Man rock shelter in northern Australia in northern Queensland at the same time as Ed Harris, the EC Harris of the Harris matrix, who was working at the site. So I have a bit of involvement with stratigraphy. I have training in sedimentology and geomorphology, so I bring a certain sort of eye to a site. I’m sure we are all familiar that, in an ideal world, an archaeological site is a layer cake of layers. This is one of the classic French upper Paleolithic sites in the Dordogne. I like it because it’s a really good example of classic layer-cake stratigraphy. We’ve got the dark, charcoal-rich occupation layers, each nicely separated by layers of limestone éboulis or frost-fractured rock. This serves as a really good example of the sort of layering we would almost ideally expect to find, each well distinguished from the layer above and below, with older layers at the bottom and younger layers at the top. An archaeological excavation, like a piece of surgery, involves careful dissection of the structure of a site. This certainly is the case overseas, but I think in Australia we’ve really lost this art. We tend to dig in quite small pits, one-metre square in 50 centimetre squares. In rock shelters we’re looking at fairly amorphous deposits and generally we’re digging in a mechanical way with spits - two-centimetre spits, three-centimetre spits, five-centimetre spits - and we’ve really forgotten how to think about a site as a stratigraphic entity. To me, stratigraphy is perhaps the major failing of archaeological training in this country. When I read through the mass of archaeological reports and look at the site reports and look at the stratigraphic sections, it’s very clear that very few people are really understanding the structure of their site and the sorts of deposits they’re looking at. Imagine a young field archaeologist sitting at the bottom of his trench, looking at the wall of his trench. It’s all red brown deposit, fairly undifferentiated, with a few rocks here and there, odd charcoal smudge, a bit of mottling. What do they do? They draw a line around the areas of organic enrichment, a line around the areas of mottling, drawing the major rocks - and that’s it. No interpretation, no sense of how that sediment built up, no sense of where those sediments came from or how they relate to material outside the site, whether those sediments are reworked or whether they’re primary deposits or whatever. I think the major problems are conceptual, not technical. I just want to give younger archaeologists a bit of a feel for how they may go about thinking about a site, what sort of questions they might ask about the structure of their deposits as a way of improving the quality of the archaeological recording exercise. Today I want to walk through the process. My aim is to get you thinking stratigraphically. I want you to visualise the sediment cycle. Sediments are the source. They’re transported to a site; they’re deposited on a site; they’re altered; and they’re removed from a site. So let’s think of a life cycle of sediments. I want to walk through the process and starting by looking at sediments; then looking at how sediments are combined in depositional units that we call layers; and then I want to look at the interface between layers. I want to look at the features, the pits and burrows that cut through layers; and I want to look at the processes that overprint the stratigraphic sequence - classically, soil development. Why should we do this? Why does it matter? The problem is, archaeological remains are intercollated, or slipped, into a sedmentary sequence that may have a complex history of its own, and we need to understand that if we’re to understand our site. Sediments are also the recording medium of the site. Imagine a video and strobe analogy. How fast you record depends on what sort of mix of material you get, whether there are gaps in a sequence, whether you get overlays of occupation or whether you get clear separation of events. So sediments are the recording medium. Sediments are also an environmental archive. They’re a bridge between the rock shelter and the wider landscape, both in terms of the stuff that’s flowing into the shelter from outside the shelter, and events that are happening within the shelter as the climate changes - how fast the shelter itself weathers. The sedimentary history of a site is basically a site’s biography. There are all sorts of reasons for getting a good, working field interpretation of a site’s stratigraphy. It needn’t be a complex process. It doesn’t require high-end equipment. You don’t need to have a portable XRF. You don’t need to be into magnetic stratigraphy. I’m talking about good, basic field recording that can be done by a field archaeologist with a trowel, a tape measure, and a hand lense. We’re talking about asking the right sorts of questions, and we’re talking about a recording process so that any good field archaeologist can make a good, competent assessment of site stratigraphy. Do you think, Mike, that the art of stratigraphy is somewhat lost because people don’t recognise the land units that they’re working on, this basic idea of what a landscape is and how a landscape is formed? It’s been partly my own experience that archaeologists will go to a place to excavate it but they won’t even recognise the geomorphic piece of landscape that they’re working on, which seems to be almost like the cover of a book before you look at the detail of the stratigraphy within it. Yes, I agree, absolutely. I think it’s a failure to read the landscape. Archaeology is a science-based humanity. It’s a very broad-based subject, and sometimes I think the science end of our training is quite poor. I think if you’re going to be a field archaeologist, you need a working knowledge of geomorphology and sedimentology. You need at least to be able to look at a landscape and say, ‘This is an area where sediments are going to accumulate. 'This is a point bar deposit. 'This is the tail end of a dune. 'This is a little rock shelter where the sediments are held in by the block fall 'at the mouth of the shelter.’ You need a bit of an eye for what’s going on in the landscape. In a sense, the conceptual framework is more important than anything else. We do enough fieldwork in such variable conditions that we should be able to see what happens when water flows across a landscape - where it goes, where the sediment ponds and accumulates. If we look at a landscape, we can see which bits of it are eroding, deflating and where the sediments are going. We need to have an eye to reading a landscape, and reading the structure of a site is part of that whole process. Do you see archaeologists as needing to be, I guess, environmental historians as much as archaeologists to understand the context of a site? Particularly, I’m thinking here of how we go about excavating these landscapes without some sort of concept of what landscape we’re looking at. I’m thinking here of people - and I’ve seen it many times myself - going to a site, not fully understanding the landscape, and then almost randomly deciding to put a one-by-one metre pit or a 50-by-50 centimetre pit down, and simply taking it off in these five-centimetre spits with no real understanding of the environmental history of an area and the need to look into that before beginning the excavation. Again, I agree quite strongly with that view. Environmental history is one of the foundation disciplines, I think, of field archaeology. It’s not an optional extra, simply because the structure of a site, where you choose to dig and how you dig, makes such a huge difference to the quality of information you recover. What you date, how you analyse your site, is absolutely fundamental to the structure of the information we gather. In practice, you should be able to stand on a site before you start excavating and have a working hypothesis of how that deposit is built up: where the sediments are coming from and what part of the landscape you’re looking at. Those sorts of readings will help you decide which part of the site you’re going to dig - where you’re likely to get disturbance, where you’re likely to find areas which have been heavily disturbed by animals, where you’ve got water runoff, where you’re likely to have good site integrity, or where you’re likely to have a complex mixture of sediments. I’ve seen cases where people have dug quite a small trench inside a rock shelter and they’ve hit what is absolutely the most complex part of the site where a whole series of layers feather out together, so that within a five-centimetre range you have cut through three or four layers which actually expand out as you move down the sequence, and so you might get a 5000-year spread just within a few centimetres where, a few metres out, that may cover a two-metre depth. Now you need to know what you’re likely to encounter when you’re digging a site.