The language of numbers can cause breakdowns in communication across teams because—perhaps counterintuitively—multiple markers from a single person can reveal information beyond the genetic profile of that specific person. This is particularly important when we consider how few ancient human remains we have access to. To understand how multiple genetic markers can be used to maximize information, we must take a step back and consider that each genetic mutation has its own history, its own distribution, lineage, and ancestry profile. As an example, we can look to the oft-quoted observation that everyone with blue eyes shares a common ancestor. There is a mutation in the genetic code affecting melanin production in the iris, one which “dilutes” brown eyes to blue. The scientists identifying this trait commented on the lack of variation across melanin expression in modern blue-eyed people—despite their high number and wide distribution—concluding that they share a single, common ancestor.²⁰ However, if we were to imagine all the blue-eyed people in the world today and to draw a family tree tracing back through time and leading to one common ancestor, such an image would be misleading. All of us have many different genetic mutations and all of us also have many, many common ancestors. We can draw many different family trees depending on how many genetic markers we look at. We are more than our eye color.

Each individual—regardless of where or when they lived—is part of the human family. From this perspective, we can assume that any two people, regardless of how distanced they are through time and space, are related in some way by virtue of their humanity. Our individuals from the Neolithic were related to their own ancestors, to each other, to other individuals from the Neolithic described in different studies, and—obliquely—to their modern-day descendants, for whom we have far more genetic resources. The question is, to what degree are all these different people, past and present, related? Direct relatives or very distant cousins? While we can learn about relative relatedness of different people using multiple genetic markers from across the genome, it is important to be aware that a genetic sequence—even thousands of genetic sequences—does not act as a perfect proxy for an individual’s ancestry. In this respect, as geneticists we are looking at an individual’s genetics, modeling multiple possible scenarios of “relatedness,” and making our “best guess” as to which one fits.

Our best guess relies on an additional discipline—biostatistics and, in particular, modeling—that allows archaeologists and geneticists to leverage known data (genetic information from existing samples) to navigate the uncertainty of missing data (sequences of DNA that we cannot decode due to degradation). Even more important than missing sequence data is the gap of who is missing—all those people contemporary with those excavated from burial sites; people whose genetic profiles we know nothing about. Knowledge of statistics, and what can and cannot be justified through their use, is a mainstay of genetic culture, but one that is often lost in translation. There is a common temptation to view statistical modeling as a “black box” into which questions are put and from which answers appear. The language of numbers—literal formulae in methodology sections of papers—can be less accessible than words. Our partners working in archaeology or the humanities were often disenfranchised from statistical knowledge-making processes by their unfamiliarity with the field and, it could be argued, by our inability to offer adequate translations. This has limited the opportunities for those without biostatistical knowledge to have meaningful input to modeling strategies, with the effect that methodologies have sometimes been designed with limited acknowledgement of key archaeological data. Also, outputs are potentially misinterpreted when reviewed by researchers who may lack the knowledge base to critically determine what makes a good model.

Of all the disjuncts we noted, this was one of the most critical, since the choice of genetic markers, samples, or calculation models can change the “answer” to a research question—sometimes dramatically so. As demonstrated by Marianne Sommer and Ruth Amstutz elsewhere in this volume (chapter 2), even the same finding can be given a radically different interpretation depending on the tools used to visualize the data. Equity of knowledge with regard to statistics, modeling, and computations is an important prerequisite for collaboration in a context where so much of the research depends on understanding the numbers. For geneticists, statisticians, and those modeling population dynamics, a challenge will be to bridge this knowledge gap. While it may not be feasible to impart a career’s worth of training, those with numerical proficiency could take responsibility for communicating the key issues, choices, and elements of statistical analysis such that colleagues across the team are able to have an active voice and input. This effort may enrich future methodologies and support a culture in which all team members are able to contribute their expertise.

Conversely, those with a scientific or statistical background have much to learn from their humanities colleagues. One example may be a humility in terms of what can be modeled statistically given the profound ambiguity of the past. Archaeogenetics as a discipline may need to accept that certain questions may never be answered in exhaustive ways, because the remains of ancient societies are always fragmentary. There is a danger that uncertainty can be amplified by gaps in both molecular and material data, creating a problem if statistics are used to “smooth over” the unknowns. The need for interdisciplinary teams to ensure a strong understanding of biostatistics and models across the team is crucial if researchers are to signpost caveats and limits to their interpretation of the data.

Stories

Our genetics both reflect our unique individuality and anchor us to a shared humanity, and so do our stories. How we perceive these stories forms a key aspect of how we collaborate. At the Bergheim site, the Ancestra project team assessed a structure—silo 157—where ten portions of left upper limbs and, directly above, eight bodies were stacked in various positions without any particular organization. There were two men, two women, and four children. In addition to the traces of cutting blows related to amputations, the upper limbs also bore traces of cut-marks. The remains of an individual with an amputated arm had several traces of blows, especially on the head, thought to correspond to a violent death.

Scientific neutrality is difficult to maintain in the face of human suffering and the stories that may explain it. The dispassionate evaluation of genetic material, which conjures images of colorless liquids, laboratory machinery, and databases, stands in stark contrast with the excavation environment archaeologists are familiar with. Here, physical remains are anchored to landscape and material culture, with all the emotional charge these may bring.

It is, therefore, unsurprising if research based within the purview of genetics becomes enmeshed with preexisting, emotive stories. For example, the characterization of an individual genome from well-preserved human remains dated to the Mesolithic—Cheddar Man—maps on to contemporary discussions around race.²¹ In particular, the highly evocative reconstruction of Cheddar Man, for whom sequencing analysis suggested a dark skin color and blue eyes, captured imaginations across a substantial press movement and stimulated much highly-politicized debate.²² The original paper by Brace and colleagues, which notably did not contain the iconic image, balanced the excitement of the findings with context (in the form of other sequenced genomes), raw data, and thoughtful scholarship.²³ Much of this was lost when the story became caught up in far less restrained dialogues around national identity that emerged in the aftermath of the UK’s Brexit.

Elsewhere in this volume (chapter 6), media historian Andreas Nyblom explores how another “celebrity” of ancient DNA—the Birka warrior—was determined to have two X-chromosomes (indicative of “female,” where sex is determined through chromosomal status). The Birka warrior became an iconic figure who stimulated much public discussion of gender roles, gender-identity, and feminism.²⁴ Archaeogeneticists run the risk of genetic science being reinterpreted and leveraged to offer a seal of closure on issues that go beyond biology or the remit of genetics and that deserve a balanced consideration of many different types of evidence.

During the writing of this chapter, a film—The Lost King—was released depicting the story of the much-celebrated finding of Richard III’s burial site. This provides an interesting example of how stories interplay with evidence. The final resting place of this English king had been lost since the friary he was buried in—Grey Friars in Leicester, England—was destroyed in the sixteenth century. When human remains were excavated near the place where the friary was thought to have stood, the juxtaposition of the site (a prosaic parking lot) with the cult status of the lost king made for a compelling story. Substantial evidence to support the claim that the body was that of Richard came from different archaeological sciences: the osteological analysis was consistent with historical records regarding the king’s age at death and physical deformities (scoliosis rendering one shoulder higher than the other).²⁵ As the last king of England to die in battle, perimortem wounds were assessed and found indicative of a violent demise, and radiocarbon dating suggested a death during 1485.²⁶

To provide the genetic perspective, Turi King and colleagues were able to successfully compare mitochondrial DNA from two purported living descendants from Richard III’s female line with the ancient sample.²⁷ They found a match for both. To get such a complete correlation between the ancient DNA and modern is rare enough to be considered highly indicative of ancestry. However, Y-chromosome data from five putative male relatives told the opposite story: the genetic data suggested the men were not descendants of the possible king. The genetic findings were, in other words, in conflict with each other. Given the broader framework of information, the authors concluded that the Y-chromosome data should be discounted on the grounds of likely nonpaternity. Given the number of generations between Richard III and his modern-day descendants and the potential inaccuracy of genealogy over this time scale, this seems a plausible explanation, particularly when compared against the robusticity of the other archaeological, historical, and genetic data. However, it gives pause for thought: if only male relatives had been available, would the genetic data have been used to undermine the other types of evidence, or perhaps left unpublished? The authors of this study show a balanced and extensive review of multiple information sources, but how many instances are there where authors are perhaps less assiduous in evaluating different data or do not consider the whole picture? Are genetic data underreported if they do not support the story?

The paper itself cross-references and contextualizes the genetic data tightly to the archaeological information available but it also uses the discussion to go beyond key conclusions of Richard III’s identity. The authors explore the idea of nonpaternity and, in doing so, introduce a rather left-field story that exemplifies the type of claim that can set interdisciplinary teams into conflict:

One can speculate that a false-paternity event (or events) at some point(s) in this genealogy could be of key historical significance, particularly if it occurred in the five generations between John of Gaunt (1340–1399) and Richard III.… A false-paternity between Edward III (1312–1377) and John would mean that John’s son, Henry IV (1367–1413), and Henry’s direct descendants (Henry V and Henry VI) would have had no legitimate claim to the crown. This would also hold true, indirectly, for the entire Tudor dynasty.²⁸

Theoretically, this observation, if true, could undermine the historic lineage of the British monarchy. It is this type of statement that tends to polarize readers, and certainly those working in archaeogenetics. Within our own respective teams, opinions on speculative stories differed. To some, this kind of speculation—if clearly framed as such—provided interest and allowed for a stimulating discussion point. More critical voices felt that researchers, regardless of their specialty within a team, had a responsibility to resist the temptation to create or give credence to inflammable stories.

As in general life, politics invariably engenders strong views and conflict within archaeogenetics teams, and not solely between but also within disciplinary boundaries. Depending on your definition of politics, stories involving demographics can easily become political. This creates a dilemma for any archaeogenetics collaboration: How do our findings map onto preexisting stories embedded in a politicized environment?

There are some contexts where we, as researchers, may be acutely sensitive to the impact of our findings in the political world and other contexts—such as the distant past—where it may feel relatively “safe” to propose more speculative ideas. If we believe the past is a safe space in which to follow blue-skies thinking, we can propose various scenarios without fear of censure. However, this brings us to a dilemma that affects many collaborations: To what extent can we argue that it is appropriate to apply certain assumptions and techniques (for example, the use of genetic relatedness to enable categorization of ethnic groups) in some contexts but not others? If we accept that there are some stories that would incur severe ramifications on modern populations, how do we then perceive those stories that appear less hazardous but that rely on the same principles? How do we maintain our research and the valuable contribution that genetics can make while navigating the broader implications of our stories? Our teams may not have resolved these questions in the course of our research projects, but we see a value in proposing them for deeper consideration.

Answers

Ancient DNA contains information that can be reintegrated into a broader framework for understanding the past, yet, as alluded to throughout this chapter, this is no easy task. Radiocarbon dating is often cited as a comparator technology, one that offers some prospects for how DNA analysis might, in the future, become seamlessly integrated into archaeological science.²⁹ As C14 dating gives an age to remains, the hope is that archaeogenetics can provide us with a genetic library of ancient individuals. Contemplating how best to integrate genetic techniques through consideration of C14 dating seems an interesting thought experiment, yet to our archaeological colleagues it provoked significant concern and so serves as a useful exemplar for unpacking conceptual differences.

In terms of knowledge production, radiocarbon analysis is used for chronological dating to define temporal context. Genetic science is often (though not exclusively) used to define ancestral relations, migration events, and gender status. In doing so, it becomes inexorably linked to narratives that are far more complex and changeable than chronological dating. Yet, if we were to suppose that genetic analyses were to take more of a back-seat position and become a single part of a much bigger puzzle—that is, a piece of contributory evidence rather than a defining component—we are still left with a problem of inherent ambiguity. Radiocarbon dating requires a precise frame of reference to be calibrated correctly. The “reference point” for C14 calibration is a year, a commonly accepted calendrical term. No easy equivalent exists for aDNA; the options are to reference comprehensive modern genetic libraries or other ancient DNA samples, the latter often being distanced geographically and temporarily. The issue of how data are then interpreted in the wake of the temporal gap between these libraries and the populations they seek to reconstruct remains a topic for scrutiny. It is not an agreed, precise science, but more of an art form. Genetic data cannot, therefore, be as easily translated into current systems of archaeological knowledge production as C14 dating, since the information contained within the genome can be calibrated only against reference points that are themselves subjectively chosen and open to debate.

A further difference with C14 data is the route of communication. C14 “answers” are incorporated into archaeological publications, yet while these may cite archaeogenetic data, primary findings are often presented in scientific journals that impose a highly structured format and a tight word count. An unintended consequence of this may be that “answers” are not subject to a more lateral scrutiny that assesses the choices of research question, samples, or a priori assumptions. Extensive interpretation is, in scientific journals, provided in the discussion, where findings can be placed in context and uncertainty can be reintroduced. However, word counts limit the coverage. Archaeological journals offer a greater opportunity to expand on detail, favoring a more open, narrative discourse, but the methodologies and statistical frameworks needed to interpret genetic data are not usually a good match for these vehicles. An additional consideration is the relative ranking in terms of impact factor and other journal metrics that favor scientific journals—Nature will usually be a preferable target to Antiquity. The lack of a publication “home” able to straddle the breadth of information yielded in these studies means that the strengths of interdisciplinarity do not fully come into play.

This poses a serious question around interdisciplinary research for archaeologists and geneticists: if the context is a critical aspect of the story, extrapolating an “answer” from one piece of evidence without consideration of the broader framework could, in many cases, be misleading. The decoupling of the backstory may run the risk of authors (and readers) drawing meanings from their results that extend beyond what can be supported by the data. The conclusions may well be valid from the point of view of scientific procedure, and the genetic information may well support broader interpretations, but without contextual information—such as the consideration of grave goods, placement, and historic sources—the story lacks anchorage and foundation.

Gained in Translation

Noam Chomsky’s idea that the act of interpreting language is inherently creative gains purchase in the practice of interdisciplinary research.³⁰ Here, divergences in language extend into working culture, where we may find conflict. If we are prepared to sit with the potentially uncomfortable task of reinterpreting ways of thinking which we have previously taken as unequivocal, we may, in these spaces of challenge and negotiation, find the creative insights that Chomsky alludes to. However, this requires us to do something inherently alien to many of us, to be at ease with discord long enough to understand it fully through reflection.

At first glance, archaeogenetics appears to be a discipline well versed in grievance airing and reflection.³¹ In the past few years, several papers have critiqued the field of human aDNA, in some instances proposing a series of solutions designed to mitigate some of the friction inherent when researchers from diverse backgrounds come to work collaboratively. The urge to avoid friction is a very human one and it is to the credit of aDNA researchers that efforts to address the issues of the discipline are ongoing. However, while well intentioned, the desire to generate a universal common ground may lose us the opportunity for the fresh understandings and syntheses unique to each research project.

Archaeogenetics involves the meeting of two traditions—each with its distinct practices, ideals, and conceptualizations of research—and its investigations take place across vastly different vistas of temporal and geographic space. These considerations necessitate the involvement of different stakeholders. From museum curators to Indigenous communities, regional administrative bodies to local history groups, it is hard to predict the exact blend of perspectives and voices that any given research project may involve. The idea of a “research team” goes, therefore, beyond the purview of individual investigators and encompasses a richer diversity, a broader meeting of minds. This encounter—in some cases, a volatile one—poses the challenge of reconciling stark differences in perspective, differences that cannot truly be anticipated by procedural guidelines. We see a value in research teams—using the widest definition of the “team”—navigating their way through these differences together.

In our experience, what makes these differences both stimulating and exasperating is that they are often sharply personal. Academics have devoted much time and intellectual resources to the pursuit of knowledge. They may struggle to confront a somewhat humbling reality of interdisciplinary research; while they hold expertise in their discipline, they are not necessarily an expert in their own project. Key truths, sometimes painfully obtuse ones, are held by other team members—not always from the same academic background—and these truths may be critical to the nuance and accuracy of the project’s findings. In an age of financial cuts and intense competition, academia itself borders on a vocational enterprise. We are often arguing the case not merely for our research tradition—be it rooted in genetics or archaeology—but for our own life choices. It is hardly surprising therefore that our position as researchers is often entangled with our identity and deeply rooted beliefs. Our emotional investment in our project work animates our discussions, even as we recognize that it is only through collaboration that we can operate in this world of contrasts.

To conclude: archaeogenetic research is an open-ended exploration in which we can expect the world to be complicated. Our role in such a situation of exploration is to embrace and express that complexity in order to bring it to light for a wider audience. By seeking to contain critical conversation, we risk creating a silo in which problems are defined by their preformulated solutions. The resulting research culture becomes inherently reductive. Instead, there is a desperate need for profound conversations on difficult issues. Natural science, with its current command of the attention economy, has a crucial role to play in such a discussion.

Notes

1.  Sam Hamill quoted in Anne-Marie Cusac, “Sam Hamill Interview,” Progressive Magazine, April 1, 2003, https://progressive.org/latest/sam-hamill-interview.

2.  See, for example, Ewen Callaway, “Divided by DNA: The Uneasy Relationship between Archaeology and Ancient Genomics,” Nature 555, no. 7698 (2018): 573–576; Gideon Lewis-Kraus, “Is Ancient DNA Research Revealing New Truths—or Falling into Old Traps?” New York Times Magazine, January 17, 2019; Krystal S. Tsosie et al., “Ancient-DNA Researchers Write Their Own Rules,” Nature 600, no. 7887 (December 2021).

3.  See, for example, Mary E. Prendergast and Elizabeth Sawchuk, “Boots on the Ground in Africa’s Ancient DNA ‘Revolution’: Archaeological Perspectives on Ethics and Best Practices,” Antiquity 92, no. 363 (2018): 803–815; Kendra A. Sirak and Jakob W. Sedig, “Balancing Analytical Goals and Anthropological Stewardship in the Midst of the Paleogenomics Revolution,” World Archaeology 51, no. 4 (2019): 560–573; Songül Alpaslan-Roodenberg et al., “Ethics of DNA Research on Human Remains: Five Globally Applicable Guidelines,” Nature 599, no. 7883 (2021): 41–46.

4.  Anna Källén et al., “Petrous Fever: The Gap Between Ideal and Actual Practice in Ancient DNA Research,” Current Anthropology (forthcoming).

5.  Tsosie et al., “Ancient-DNA Researchers Write Their Own Rules.”

6.  Cristina Gamba et al., “Genome Flux and Stasis in a Five Millennium Transect of European Prehistory,” Nature Communications 5, no. 5257 (2014): 1–9.

7.  See Katri Huutoniemi, “Interdisciplinarity as Academic Accountability: Prospects for Quality Control across Disciplinary Boundaries,” Social Epistemology 30, no. 2 (2016): 163–185.

8.  Källén et al., “Petrous Fever.”

9.  Anna Källén et al., “Archaeogenetics in Popular Media: Contemporary Implications of Ancient DNA,” Current Swedish Archaeology 27, no. 1 (2019): 69–91; Susanne E. Hakenbeck, “Genetics, Archaeology and the Far Right: An Unholy Trinity,” World Archaeology 51, no. 4 (2019): 517–527.

10.  Päivi Haapasaari, Soile Kulmaka, and Sakari Kuikka, “Growing into Interdisciplinarity: How to Converge Biology, Economics, and Social Science in Fisheries Research?” Ecology and Society 17, no. 1 (2012): 295–320; Vincent Larivière, Stefanie Haustein, and Katy Börner, “Long-Distance Interdisciplinarity Leads to Higher Scientific Impact,” PLoS One 10, no. 3 (2015): 1–15.

11.  The search terms were “archaeogenetics,” “paleogenetics,” “ancient DNA,” “archaeogenomics,” “paleogenenomics,” “aDNA” in all languages. The survey was limited to human studies and manually reviewed (titles and abstracts) for content and relevance. Alternative spellings of key search terms were used.

12.  Källén et al., “Archaeogenetics in Popular Media,” 69–91.

13.  Haapasaari, Kulmaka and Sakari, “Growing into Interdisciplinarity,” 1–12.

14.  Martin Furholt, “Biodeterminism and Pseudo-Objectivity as Obstacles for the Emerging Field of Archaeogenetics,” Archaeological Dialogues 27, no. 1 (2020): 23–25.

15.  Janet Stephenson et al., “The Practice of Interdisciplinarity,” International Journal of Interdisciplinary Social Science 5, no. 7 (2010): 271–282.

16.  See the chapters in this volume by Magnus Fiskesjö (chapter 7), as well as by Marianne Sommer and Ruth Amstutz (chapter 2).

17.  Anna Källén, The Trouble with Ancient DNA (Chicago: University of Chicago Press, 2024).

18.  Alpaslan-Roodenberg et al., “Ethics of DNA Research”; Prendergast and Sawchuk, “Boots on the Ground”; Sirak and Sedig, “Balancing Analytical Goals.”

19.  A haplotype can be defined as a series of genetic variants that are inherited together on the same chromosome. In this example, where women share the same mitochondrial haplotype, it is indicative of matrilineal relatedness.

20.  Hans Eiberg et al., “Blue Eye Color in Humans May be Caused by a Perfectly Associated Founder Mutation in a Regulatory Element Located within the HERC2 Gene Inhibiting OCA2 Expression,” Human Genetics 123, no. 2 (2008): 177–187.

21.  Selina Brace et al., “Ancient Genomes Indicate Population Replacement in Early Neolithic Britain,” Nature Ecology & Evolution 3, no. 5 (2019): 765–771.