MIKE SMITH: We’ve worked through the factors that affect the stratigraphic sequence: the nature of the sediments, the layers, the interfaces, the features that cut across all of these, the post-depositional factors that overprint the stratigraphic sequence. Let’s go back to our archaeologist sitting in a trench, staring at a wall of red-brown, undifferentiated sediment. What do you do? Of course, yes, it is useful to record colour, texture, and pH, but do this towards an end. Have a look at how the texture of the sediments varies either across the sequence or up the profile. Try to work out whether there really are any changes here. You may have a completely uniform sediment, but texture is the best guide not colour.
Look at the rocks. As I’ve stressed before, rocks are an opportunity. How are the rocks sitting in a deposit? If they are rocks that have a flat surface, how is that oriented? That will give you a guide to the slope of the deposit and the grain of the deposit.
Have a bit of an eye to the relationship between the rock shelter sediments and the exterior sediments. Are they continuous? Could they be continuous? Are the rock shelter sediments perched on a hillside somewhat?
What are the layer boundaries like - if you can see any boundaries at all? Be careful how you draw boundaries. Use solid lines for sharp interfaces. Use dotted lines for gradual changes. Don’t make all the lines on your stratigraphic drawing the same. Don’t draw a line around mottled areas as if it was a bounded layer.
If you’re outlining organic concentrations, charcoal smudges, you have to try to work out what you’re looking at and annotate your drawing. Are you looking at a diffuse hearth or just an area of charcoal enrichment or maybe a tree root?
Have an eye to the preservation of features. Features are preserved more commonly than we often allow. Many small rock shelter sites have a whole series of hearths. Well, have a look at them. How well are they preserved? The fact that they’re preserved suggests there has been minimal post-depositional alteration of the sediment. Look at how the hearths are laying. They have been laid down on the surface as well. Try and work out what layer they’ve been dug from, if they’re a pit-type feature.
Overall, I’m arguing that you should hazard an interpretation of a site. Try to look at the biography or the sedimentary history of the site. Try your hand at listing the sequence of events in temporal order. What was laid down? How was it laid down? Where did it go? What’s happened to it since? What happened to it before the next layer was laid on top? And so forth.
You can use laboratory analyses to supplement and test field interpretations, but don’t use them as a primary means of understanding the stratigraphic structure. You can do this as a field archaeologist. It’s conceptual, not technical. Yes, of course, the sort of analyses you can use to supplement your field interpretations are grain size analysis of the sediments, because that is very useful in working out what sorts of sediments you’re looking at or whether you’ve got a mixture of sediments.
Use thin sections, if you can get them, to look at the fabric of your sediment to perhaps see whether you’ve got pelletal clays or whether the voids are filled with clay, and how the minerals are aligned within the sediments.
Mineralogy is good, but in many cases it is not going to tell you very much in many Australian sites.
Global diagrams, which juxtapose particle size analysis with other stratigraphic information, are very useful for getting an overall sense of the change in a deposit. This is a global diagram from Puritjarra Rock Shelter. It shows the changing proportion of coarsed rocks and gravel and the fine sediments in terms of sand, silt, and clay, and also the proportion of grass in the phytolith column.
What it shows on a global diagram like this is that you can see a whole series of related changes between layers one and two at about 8000 years ago. The size of the rocks change. You get a lot more fine gravel. You get a lot less large rock fall. You get a change to a much more sandy deposit and a great increase in the proportion of grass. You have a whole range of changes here that relate to the changing local environment of the rock shelter, but it’s the ability to compare across these different materials that makes these diagrams useful.
At Puritjarra, for instance, originally in the Pleistocene layers we had mainly large rockfall, and the fine sediments were made of up of primarily wind-borne material that was deflaking from dune crests and interdunal corridors. But as those areas re-vegetated during the Holocene and you got a more moist, aggressive weathering regime within the rock shelter, the deposit switched over to a more gritty and less rocky deposit. The supply of clay-rich sand cut off and left us with just sand from dune crests. You can see that happening just when you look at the coarse component, the rocks and the gravel, as well as the fine sediments.
Magnetics - magnetic stratigraphy is a promising way to go. But as I have stressed, most of what we’ve covered today a good field archaeologist can do with just a trowel, a pH kit and a magnifying glass.
Chronostratigraphy. Of course, when you get your dates back, there may be a few surprises. Samples that are separated by only a few centimetres might be thousands of years difference in age. I think you should try to minimise the potential for such surprises by working out what you’ve got in the field. There is a real danger here of drawing stratigraphic boundaries after the event, a sort of painting-by-numbers approach, which is a bit dodgy.
This is one of the stratigraphic drawings for Purritjarra Rock Shelter in western Central Australia and it shows a number of things. First of all, only use sharp lines where you’ve got a very sharp boundary. The boundary between these two layers is a graded boundary over a few centimetres change between the silty clays and the sands of the Holocene unit.
It also shows some of the stratigraphic features. We have this pit here dug and, if you look carefully, you can see it’s dug from this level here, as indicated by the line of material. It is very important to isolate these features. If any of the TL samples had come through that feature, they would have given us a false age for the sediment.
The other feature I want to draw your attention to here is this little fire pit feature shown here and here and various cuts through it. What it shows is a feature, a pit, that has been dug very quickly. Fast-burning fuel has been lit in it, which leaves a cinder of a burnt mass and soots the margin of the pit. Here you can see it in the longitudinal section with the burnt cinder. It’s a little fire pit, probably used for heating sappy wood or something like that as you’re shaping it. It’s a very ephemeral feature but again could cause chaos if you don’t isolate these during excavation. They have also have a spatial component and occur toward the front of the rock shelter.
Overall, what we’re aiming for in the stratigraphic diagram is not so much a straight recording of what we can see but a good field interpretation. One has simply got to sit down with the section and work out what you’ve got, where it’s coming from and how it’s laid down - have a stab at doing this. Description alone is inadequate and is usually confused rather than objective. If you can’t work out what you’ve got in the field, quite likely no one else can work out what you’ve got from looking at your stratigraphic plan.
The art of stratigraphy: it is an art but it’s not a difficult art. It just requires you to ask some questions about the nature of your site and how it’s built up. This will help you both understand the structure of your site and the deposits it contains but also help guide your hand for future excavation, because excavation is effectively dissection of structure.
ROBERT PATON: Mike, when you began your discussion of stratigraphy, you said that in a sense it was like surgery. It’s interesting. I was watching a program on surgery not long ago where a surgeon was teaching younger surgeons. He opened the person up on television and when he looked in, he said, ‘No one can tell any of the organs in the body when they look in. Get your hands in and feel it. Just touch it, feel it, and feel the differences between the organs. Close your eyes and feel it.’
Do you think archaeologists need to get into a trench? And this is how I do it. I often tell the difference between sediments by just getting some on my fingers, rubbing it and feeling it. Before I decide what something is, I can often do it just by the feel of it. Is that something archaeologists should do? Should they have an apprenticeship in the field?
MIKE SMITH: There’s definitely a case for a field apprenticeship, but I think you can learn a lot of this material by working through other people’s stratigraphic diagrams and trying to sketch out the order of events and how a deposit was built up. At least that will alert you to problems and questions.
But, like you, I work my way up a section by feeling it. I think texture and fabric is much more important than colour. Colour is misleading. When you feel a section, you feel where there are real changes in the grade of the sediment. You go back to your field notes and you see where the proportion of gravel and grit is changing. You can then also go later on to particle size analysis in a lab, just the sieving of your sand fraction, and work out whether those boundaries agree with what you’ve identified in the field.
What I like to do is collect a column sample, a whole set of samples down the stratigraphic sequence. In the field I may perhaps have identified a shift to more clay-rich sediments at point X. Then I’ll go and test that in the laboratory and see if that’s where I get a change in particle size. So I use the laboratory analysis to back up my field work, my field observations.
Excavation is a very tactile process. You’ve got to have a real sense of how things are bedded as you flip them up when you’re excavating. With features too you can tap sediments and watch things fall away from the edge of a pit. You’ve got to use a range of techniques to expose that structure.
The final point here, which would lead us into a wider discussion of how we work with archaeological assemblages, is simply that we’ve got to dig enough of a site to be able to characterise both the nature of the occupation and the actual history of the site in terms of the formation of the sediment.
We often can’t do this from just a one square metre pit. We can have a go but, really, we have this wealth of Pleistocene sites in Australia, but few of them have been adequately characterised in terms of the nature of the deposit or in terms of the nature of the occupation. We really need to do more if we’re going to exploit the potential of these as environmental and cultural archives.