Appetite for (less) Destruction

This week we had the annual research away day for the Department of Archaeology at Durham. It was the first such event I have been to since joining the department just over a year ago, and it was great to see the range of brilliant work going on across the breadth of the discipline. I presented a short provocation on what the future of archaeology might look like in an era increasingly shaped by biomolecular data. 

We are unquestionably living through what has been termed a 'biomolecular turn' in archaeology. Advances in ancient DNA, proteomics, lipids and isotopic analysis have opened up extraordinary possibilities for reconstructing past lives, movements, diets and relationships. These methods have transformed the kinds of questions we can ask, and in many cases, the kinds of answers we can plausibly give. But they also raise important challenges. As with DNA evidence in forensic science, biomolecular data is powerful, but it does not speak for itself. It must be interpreted carefully, situated within broader archaeological contexts, and integrated with other forms of evidence.

In many ways, this is not a new problem. Archaeology has long been characterised by the development of specialist subfields, each generating its own datasets, methods and interpretative frameworks. The risk has always been that these lines of evidence become siloed, rather than genuinely integrated. What is different today is the scale, visibility and perceived authority of biomolecular data, which can sometimes lead to it being prioritised over other forms of evidence.

For me, the key principle is that we should always start with the research questions. Different methods provide different kinds of information, at different scales and levels of resolution. The challenge is to bring together the combination of approaches that best addresses the question at hand, rather than allowing the availability or novelty of a particular technique to dictate the research design. This requires genuine integration across datasets, and a willingness to engage with complexity and uncertainty rather than seeking overly simplified narratives.

This brings me to the issue of destructive analysis. Biomolecular approaches typically require the physical sampling, and often partial or complete destruction, of archaeological material. In the context of a finite and non-renewable record, this raises fundamental questions about sustainability (and even ethical issues). Every sample taken is, in effect, a decision about the future. What material we preserve for future generations, and what we prioritise in the present.

I am certainly not arguing that we should avoid destructive analysis altogether. In many cases, it is the only way to access certain types of information, and it has already produced transformative insights. However, I do think we need to be much more deliberate and reflective about when and why we undertake it. Are we asking questions that genuinely require destructive methods? Are we maximising the information gained from each sample? Are we considering how future methodological developments might extract more information from the same material with less damage?

Related to this is the growing potential of non-destructive and minimally destructive approaches. Imaging techniques such as micro-CT scanning offer exciting opportunities to investigate internal structures without physically sampling at all. These methods can help bridge the gap between field observations and laboratory data, and may allow us to explore relationships within materials in ways that were not previously possible. At the same time, they come with their own challenges, not least the volume and complexity of the data they produce, and the need for new analytical frameworks to interpret them effectively.

Looking ahead, I am particularly interested in whether emerging tools, including machine learning and AI, might help us better integrate different types of archaeological data. One of the persistent challenges in our field is equifinality - the fact that similar patterns in the archaeological record can be produced by very different processes. By combining datasets and improving our ability to model complex relationships, there is potential to develop more nuanced and robust interpretations. But this will only work if we maintain a strong commitment to question-driven research, rather than method-driven analysis.

Finally, these methodological questions intersect with broader issues about the role of archaeology in society. If we are serious about learning from the past to address contemporary challenges, whether environmental change, sustainability, or social organisation, we need to think carefully about how we connect our data to more recent records, and how we communicate our findings beyond academia. This includes engaging policymakers, collaborating across disciplines, and recognising and addressing the colonial legacies that have shaped our field.

The conversation around destructive analysis at the away day highlighted just how important and timely these issues are. As a discipline, we are navigating a period of rapid methodological change, but also one that demands ethical reflection and critical self-awareness. The choices we make now about how we generate and use data will shape not only our interpretations of the past, but also the material that remains available for future research.

I have explored these questions in more depth in a forthcoming paper in World Archaeology, specifically as they relate to 'non-skeletal human remains', but I think they are relevant across all categories of archaeological materials.