AEA conference 2015 - some thoughts on taphonomy, equifinality, and multi-proxy approaches

This weekend I went to my first AEA conference. As a student, then as a fixed term post-doc, it is difficult to fund conference attendance, and also to find the time when there are so many conferences to choose from. Having done a few years of the big conferences (SAA and EAA) to maximise audience and networking opportunities, I’ve decided to spend some time at the smaller ones, where I can focus on my specific interests. Overall it has been an enjoyable weekend, and it was great to be back in York and catch up with old friends and colleagues, including Matthew Collins, who I can’t thank enough for writing me many references over these past few years (I bet he’s relieved that I finally got a job so he can stop writing them!).

There are three ‘themes’ that stood out for me at the conference. The first was the study of taphonomy, and how wildly different this is between different techniques in environmental archaeology. The second was the recognition of the advantages of integrating methods, approaches, lines of evidence, whether this be through the use of multiple environmental proxies, or using archaeological alongside historical evidence. The third was the social media side of things – a relatively highly tweeted conference, there was some discussion over the etiquette of live tweeting. My thoughts on the first two feature here, the latter I will write about in a different post.

A little bit of background for the unfamiliar, taphonomy is the word we use to describe the processes that have had an impact on the archaeological record. What has caused the patterns that we see? It is not a straightforward process of counting bones or plants, we have to consider things like preservation – how much has been eroded away? Have some materials preserved better than others? Has material been physically altered in some way? All of these things ‘blur’ the human activity signal that we are trying to ‘read’ from the archaeological record.

I was happy to see the glamorous world of dung analysis featuring on several occasions, the first of which was a paper by Don O Meara, in which he conducted his very own taphonomic experiments on the impacts of the human digestive system on the preservation of various botanical remains. I will spare you the methodological details (it’s exactly what you think). The results were very interesting. He showed that the survival of seeds within the digestive system is incredibly variable, with some species not surviving at all, whilst others preserved differently depending on the extent to which they are chewed etc. The most notable result was the lack of standardisation with such processes, the lack of consistency in the patterns, making it difficult to apply these experiments to archaeology in general. This contrasts heavily with zooarchaeology, where these sorts of experiments can produce repeatable results and patterns which can then be identified in archaeological material.
Unsurprisingly this got me thinking about phytolith taphonomy. Taphonomic studies seem to be taken much more seriously in other specialist areas, perhaps because it is possible to do repeatable studies and get meaningful consistent data? O Meara made the distinction between biostratinomy and diagenesis i.e. some material never even enters the archaeological record, whilst other material is modified after it enters the record. Both aspects of taphonomy influence the archaeological assemblage in different ways. With phytoliths we have the added problem that, unlike seeds, we really don’t have a clear idea of the variability in production between different plants, and the wide range of environmental and other factors that influence this. If a seed were studied as phytloliths, one seed would equal thousands of phytoliths – how then do we say anything meaningful with a sample set that could represent a fraction of a single plant? On a more positive note, O Meara also suggested that taphonomy is not simply destructive, but we can think of it as adding a layer of information.

The second ‘dung’ themed paper was by Eline van Asperen, who studies dung fungal spores as a proxy for past herbivore abundance. The potential of dung spores as a proxy is as an indicator of openness in the landscape. This is something that is difficult to achieve with pollen studies, which can give you an idea of vegetation composition, but not the ‘pattern’ of that vegetation within a landscape. Again the question of taphonomy was raised, and I was very impressed by van Asperen’s systematic experimental work, and the willingness to admit that interpretation of fossil data can be very problematic. As one example, spore abundance is assumed to correlate with herbivore abundance, but her studies showed that spores can actually show up, in large numbers, in areas that animals have not had access to. This is despite the fact that spores only travel c. 50cm. Something else which is directly relevant to phytolith studies was the impact of processing methods on the spore asssemblage. Spores have often been extracted alongside pollen, but she showed that this can alter the morphology of the spores to the extent that thy cannot be identified, whilst some types do not survive the process at all. Of great concern when you are looking at ratios of different fossil types, and when you are using (often subtle) morphological features to identify the genus or species. I look forward to reading more about her experiments, which are looking at the depositional pathways of the spores. Something else that really needs to be done for phytoliths.

Another paper I liked on Saturday was by Emily Johnson. It was a zooarchaeology paper looking at different types of bone fractures, but reminded me of the work we did on the Feeding Stonehenge project. There were a number of points I thought were interesting, and were lessons for the wider subject beyond animal bones. The first was to do with the way that we measure and present data, and the influence this can have on the interpretations we make. For me it also highlighted how essential it is for other studies, such as pottery lipid analysis, to integrate evidence to get a better picture of what was going on in the past. With the Stonehenge data, we know about the volume of animals and butchery through the zooarchaeology, whilst pottery lipids tell us more about how the animals were used and processed for a specific activity. Likewise with Johnson’s study, I wondered about the lipid signal for grease processing in pots, could this be distinguished from other types of fat from the same animal? Also the question of spatial distribution – often in pottery residue studies, a sample of pots is taken from a site, and considered as a single category. Both the Stonehenge study and Johnson show that specific context is hugely important - different activities are occurring, and associated waste is being deposited, in different locations. If we don’t consider the spatial dimension, we are only seeing a fraction (of a fraction) of the overall picture.
Another paper I really enjoyed was by Suzi Richer. From an archaeobotanical perspective, she questioned the way that commercial archaeology often puts forward one interpretation as objective fact when producing an environmental report, when there may be multiple possible interpretations. I don’t think this is a problem just in the commercial world, we see it all the time in academic archaeology as well. Suzi asks what is constraining us? Are we afraid to admit we don’t have an objective, definite explanation for what we see? In the commercial sector, will other stakeholders (such as engineers) not take archaeology seriously if it doesn’t say something definitive? In academia, why are we worried that admitting there may be more than one way to interpret the data? Do we choose the interpretation we like the best, or that is more exciting? I agree hold heartedly with her final point, that by integrating multiple lines of evidence we can help reduce the problem of equifinality.
This leads on nicely to the geoarchaeology papers. I think geoarchaeology is inherently multi-proxy, as it covers such a wide range of methods, compared to archaeobotany and zooarchaeology. So does this help narrow down the interpretations? The geoarchaeology papers variously combined micromorphology, phytolith analysis, geochemistry (µXRF, pXRF, ICP-AES). The first, which was commendable in its attempt to integrate geoarchaeology data with spatial patterning of lithics, was Charlotte Rowley (et al)’s study of Flixton Island. However I did have some concerns. The approach reminded me of my own PhD, where I spent the best part of my first year doing a similar ‘look and see’ geochemistry on sediment samples. It turned out that the (mountains of) data didn’t show much of interest, and/or there were so many possible explanations for the data as to make it of little use. This is a real problem with spatial geochemistry studies, discussed recently in a review by Matthew Canti and Dirk Huisman.
Barbora Wouters presented a range of examples from towns in medieval northern Europe, highlighting how integrating multiple methods can provide a better explanation of processes. As this has been my own methodological approach I am very keen to see that there are others working on this area. Overcoming (some) of the problems of phytolith taphonomy, but at the same time showing the daunting complexity. The µXRF showed the huge variation that you can get in geochemical data over a small area. If this is what we are seeing when we focus on specific points, what are we missing when we measure bulk samples? An averaged signal from hundreds of activities, a chemical palimpsest?
Hayley McParland’s study on phytoliths and macrobotanics at Songo Mnara looked for activity areas in different parts of the site. It offered some tantalising hints of the variety of plant use, and it is great to see the whole range of plant remains being analysed simultaneously across a site. As with other phytolith studies, I feel we are still unsure of the complexities of depositional pathways and whether we are seeing a direct activity indicator or something else? The more phytoliths I see from different environments, the more I realise that preservation at Çatalhöyük is highly unusual. To me this is an example of where taphonomy can be seen to be adding information – if phytoliths are degraded and worn, what is that telling us about the processes? At Çatalhöyük we only ever see phytoliths like this where they have been deposited in exposed external areas and have become highly trampled and weathered. Is there a difference in degradation index between different areas – external versus internal space?
Lots to think about! Hence the tl;dr post - thanks if you made it this far :)