MIKE SMITH: We’ve looked at sediments and we’ve looked at layers. But as important as a layer is the interface, the boundary, the surface, the surface of a layer or the boundary between layers. The boundary between layers might be an abrupt boundary or it might be a gradual change in the character of sediments. You can often tell when you’re digging, when you’re coming down to a layer boundary, if the buried layer is quite distinctive in terms of colour. You might start to get a little bit of grain mottling as you approach the grain boundary because invertebrates do move fine sediment around. They can often move fine sediment around without totally disturbing or mixing the site, but you can often get a little bit of warning that you’re coming down to a layer boundary.
At Puritjarra, we found this: two or three centimetres above the layer boundary we started to get a little specking of red sand. The ants were moving up a little bit of that Pleistocene rim deposit.
Interfaces. Iain Davidson once quite nicely said that, in most sites, there’s more gap than sequence. So you need to give some consideration to the gap, to the interfaces. Interfaces between layers might be a change in a character of the sediment or the net sediment budget. Suddenly you’ve got very different material entering the site and being laid down - or a different mixture of material.
It might be due to erosion where previous sediment has just been cut away, wiped away so you get a very sharp boundary. It might be removed through mass removal, winnowing, deflation or gullying. Or an interface might be created by some other form of disturbance: pits, burrows or features. If you’re in a small trench you might not be able to distinguish between major removal of sediment and just the boundary between that and a burrow.
Interfaces need not be a significant temporal gap. They might be, but they could just be a period where the character of sediments has changed over time. Have a look at the sharpness of the contacts of the interfaces. You might find that over two or three centimetres the character of a sediment has changed from being silty sand to being a more gritty sand, or there might actually be a very distinct almost a line in the deposit. They have different implications.
Let’s look at the issue of surfaces. A surface is a type of interface that represents a plane on which sub-aerial processes may have had time to operate and on which repeated activities might have taken place. If you’ve got a surface for some time, you might get a lot of rock fall starting to accumulate on it. There may be no change in the rate of rock fall; it’s just that rock fall is falling on the one side, the one surface.
There may be evidence of bioturbation that relates to that surface, or disturbance or treadage that relates to that surface, or there may be organic enrichment. If you’ve got a surface, it’s worth having a look at it not just vertically, but cleaning back the surface, as [John] Mulvaney very nicely did in his excavations at Kintore Cave, where beneath a layer of surface dust he had a clay, terra rosa unit. I think it’s a very good example where John - in this case the excavator - has cleared back the surface of a layer, in this case an indurated clay layer with a very cracked surface.
It’s very important when you find a surface to clear back part of it so you can get a look at the surface in three dimensions because it may have root marks; it may have gullying; it may have cracking and mottling. The surface of that layer might be quite informative about the nature of that layer, and you can’t always see those things just in a section. Sometimes, you have to clean back to understand a stratigraphic structure.
Surfaces have their own dynamics. A conflation - a palimpsest of different sorts of occupations. A surface allows for the removal and recycling of cultural material. You might have treadage and scuffage associated with the surface, and that may disturb stuff five or 10 centimetres immediately below that surface. You might have a concentration of pits and hearths dug from this level. You need to be able to identify them, and we’ll talk more about features, but you need to know from what level they were dug.
One of the things I find that people often do when they’re excavating a site is you arrive at a rock shelter, you lay out your grid, and then you dig your first spit. In that first spit you’ve included everything that was on the surface of the site, and in fact you’ve mixed two different units. The stuff on that interface, on that surface of the site, represents often a very different process than material that’s included within the body of the spit.
[Michael] Schiffer talked about abandonment assemblages. When the traditional hunter gatherer economy collapsed and people walked off their sites, they may have left their grindstones that may have had usable lives on the site - little caches of material, left useful items around. It was a classic abandonment assemblage. In the normal course of events that material would have been salvaged, recycled, the grindstones would have been broken, used for grinding bush tobacco, used as abraders for wooden artefacts and then cleaned and discarded over there. You’ve got two different sorts of assemblages. You need to keep the stuff on the surface of the site quite separate from the stuff within the first spit. But it’s a really interesting example of how people neglect surfaces.
Similarly, if you’re excavating, and you find a sharp interface between two layers and you suspect it’s a surface, anything that’s sitting on that surface should be treated as an assemblage in its own right, not mixed into the material above or below.
Another type of interface is a disconformity. Unlike a surface, which might be just a still stand for some reason, sediment has stopped accumulating for some reason and everything is happening on that surface, a disconformity is where that interface is created by removal of sediment. It’s a ‘break’ in time, but it’s a break in time where there’s been removal of sediment through erosion or some other quite catastrophic event.
This is Asprochaliko, a Paleolithic site in Northern Greece. As well as showing the layer cake stratigraphy, it also shows a major disconformity here between these deposits and these more recent deposits, and the disconformity is marked by a flow stone in black. This shows there’s been a major period when deposit has built up, then been eroded away and then has built up again against this bench of older sediment.
When I read site reports and I look at people’s radiocarbon dates, they may have drawn a relatively uniform stratigraphic sequence profile where here there’ll be a date of 5000, and here there’ll be a date of 25000, separated by only a few centimetres. By definition by implication, there is a major interface there somewhere. Either sediment has stopped accumulating for 20,000 years, or sediment has accumulated up to 5000 years and then we clean back to 20,000 years. People need to resolve that, because it affects the interpretation of the site. If sediment is removed, maybe there’s a lot of material that’s younger than 25,000 years that has dropped down and been left in the top of that older unit.
Ed Harris, EC Harris, usefully positioned the interface as a major unit of analysis and from an archaeological perspective what happens at an interface is critical. Do we have lags of artefacts? Do we have a concentration of occupational material?
I want to draw on a Central Australian example of Kulpi Mara rock shelter, a beautiful Pleistocene site very well excavated by Peter Thorley in Central Australia. Kulpi Mara sits at the base of quite a large sandstone cliff. All along the base of this cliff you have rock shelters developed through a process called cavernous weathering or basal sapping. This is a classic process whereby moisture is retained at the base of an escarpment and you get a zone of active erosion which creates a very friable white sand. The sandstone is very soft, easily eroded.
Now, he has a Pleistocene layer with a veneer of Holocene material on top. Through very careful excavation he showed that in fact there was no lag of artefacts at the layer interface. So in this case it didn’t seem that we’d had erosion of fine sediments. What creates a major break here between a unit dated 5000 years and a unit dated 25,000 years? He looked very carefully at the interface and there wasn’t the concentration or lag of artefacts you would expect if they are fine sediments that had built up and then had been cut back or eroded.
But, on the other hand, in this location, there was no reason for expecting that human use of these rock shelters would directly switch on or off sediment accumulation. These rock shelters were actively eroding all the way along the escarpment. It was a natural process. So here we have an interesting example where probably a third factor is affecting both the accumulation of the layers and the use of these rock shelters, whereby perhaps moisture controls the degree of erosion and the rate of erosion, but also controls the attractiveness of these rock shelters for human occupation.
So you get periods when you’ve got active erosion and you’ve got active occupation, and then you’ve got periods which are more stable - humans aren’t using the site, nor is it eroding much. Here we need to stress the importance of an interface is that we’re not going to find lags of artefacts and we’re not going to find surfaces if we don’t look for them during excavation. The take-home message here is that the boundary between the layers, that what a layer sits on and what the surface of the layer is like are as important as the character of the layer itself.