To start: I thank the responding authors for their generosity and thoughtfulness in engaging in this debate about ‘Attending to unproof: an archaeology of possibilities’ (Frieman Reference Frieman2024) and also the journal's editors for facilitating this discussion.

To continue: here are two grounding principles for this response. First, in the spirit of many pasts and encouraging a flourishing of approaches to our study of them, I will not quibble where I see small disagreements or misinterpretations; each of the responses should be read as an important contribution to our evolving methodologies. Second, I will emphasise, as I did throughout the opening article (Frieman Reference Frieman2024), that I am endeavouring to conduct a political intervention, since our writings of the past powerfully impact the present and shape the future into which we are hurtling.

It is for this latter reason that I have named and, in so doing, reified a category of unproof. I use the nomenclature unproof deliberately because it confronts our ideas of proof, empirical knowledge, data and science. Just as in Le Guin's Reference Le Guin1969 novel, cited in the opening article, the fact of unproof entangling reality upsets the systematic observer who believes themself objective, or at least emotionally distanced from the objects of their study. Thus, I find the concept of traces, bits of incomplete pasts collected and assembled by archaeologists, proposed by Sørensen (Reference Sørensen2024, and in other publications) both unsatisfying in its description of knowledge-making and rhetorically ineffective. Archaeological materials are created through archaeological practice. Stratigraphic layering, bags of earth and pieces of stone have no innate meaning—they are neither traces nor data until we enmesh them in methods and stories. We can see this in the ways that various materials have been envalued as new methods have been developed: macrobotanical remains thanks to floatation, disarticulated petrous bones thanks to biomolecular analysis, etc. Moreover, if we are serious about recognising archaeology and archaeological practice as a form of political engagement and world-making (which I firmly am), then the words we use must do work. Unproof is unsettling enough to work hard for me. It challenges norms and habits with some aggression and forces us (myself included) to question how we make knowledge and if those methods meet the needs of our now and next decades.

Feminist scholarship emphasises the ways that dominant discourse, culture and social practices encapsulate (and efface) alternate experiences, power flows and inequalities. As Crellin (Reference Crellin2024) explores, proof and unproof are inextricably imbricated; but in pulling the unproof to the fore, I am bracketing it off and making it legible, giving it a tangible form so that others may also engage, in a substantive and robust way, with the many apertures in our apprehension of the past and the critique I have attempted to delineate of work that skates over these. In our sibling humanities disciplines, scholars have productively engaged these spaces through fictional narratives and critical fabulation—a body of work that I welcome Crellin bringing into the conversation. Feminist and post-colonial archaeologists too have created speculative vignettes to animate and illuminate the otherwise unknown (e.g. Tringham Reference Tringham, Gero and Conkey1991; Spector Reference Spector1993; Harrison Reference Harrison2002). Yet our archaeological fictions, coming as they often do from cultural outsiders perhaps many generations or even millennia distant from their protagonists, are not without ethical peril, risking an uncritical projection of present into past at best and a vicious recolonisation or appropriation at worst (Bernbeck Reference Bernbeck, Van Dyke and Bernbeck2015; Marín-Aguilera Reference Marín-Aguilera2024). Moreover, they are easily disregarded as mere speculation by colleagues grounded more deeply in scientific methods, as we see in Gibb's (Reference Gibb2024) response.

I offer the framework of unproof to create a bridge between the complex unknowns and possibilities of an archaeology grounded in the activist humanities and the scientific methods practised by colleagues at the other end of the disciplinary spectrum. Here is a tool they can use to avoid Haraway's (Reference Haraway1988: 589) scientific “view from above, from nowhere, from simplicity”, the god trick that centres the scientist and their unquestionable expertise rather than the complexity and incompleteness of the questions, data and methods available to them.

Indeed, far from being opposed to plausible scientific data analysis as Gibb (Reference Gibb2024) implies, I argue that the most robust science requires its practitioners to acknowledge and account for unknowns and to accept that multiple hypotheses may be equally statistically likely, depending on the questions asked, the data collected and the analytical methods employed. Maier and colleagues (Reference Maier, Flegontov, Flegontova, Isildak, Changmai and Reich2023), for example, demonstrate the limits of standard methodological assumptions for the computational modelling of ancient DNA data by developing an algorithmic method to test admixture hypotheses. This allows them to enumerate a range of plausible alternative models to those previously published, each equally supported by the scientific data. Even in the most quantitative realm of archaeological science, then, we find myriad paths through the data and many possible foundations for interpretative stories.

My thinking here emerges from conversations I have been having for over a decade with my Australian archaeologist colleagues on the one hand and my students on the other. In the first week of the advanced archaeological theory class I run, my students and I read a short essay by Sharon Hodgetts and Jesse Hodgetts, two Wiradjuiri-Ngiyampaa scholars who argue (as part of a larger discussion forum in Australian Archaeology) that archaeology cannot just be a science, it requires “empathy for the land and its people” (Hodgetts & Hodgetts Reference Hodgetts and Hodgetts2020: 304). It requires us to listen, to understand the connections that link then and now and to tell our stories of the past with these considerations front of mind. In Australia, and many colonised places, that means working closely with Indigenous, marginalised and descendent communities to co-create methods of analysis and interpretative practices (Atalay Reference Atalay2012; Gonzalez Reference Gonzalez2016; Franklin et al. Reference Franklin, Dunnavant, Flewellen and Odewale2020). Indigenous colleagues and collaborators are increasingly pushing the discipline into activist spaces, working to create an archaeological past animated by Indigenous values to help surface past atrocities, resist the injustice of the present and build towards alternate, better futures (Smith et al. Reference Smith2019; Supernant et al. Reference Supernant, Baxter, Lyons and Atalay2020; Laluk et al. Reference Laluk2022).

Marín-Aguilera's (Reference Marín-Aguilera2024) powerful comments concerning the ways archaeological theorists can reinforce colonial power dynamics emerge from this necessary tectonic shift in the field. I firmly agree with her critique (here and in other publications) of the ontological turn in archaeology. That said, ‘ontology’ has meaning beyond contemporary theoretical fads, and an emphasis on flat continuity is also a political position. An imagined continuity between Europe's prehistory and various contemporary Europeans, for example, is regularly mobilised politically (Hølleland Reference Hølleland2010), often by far-right ethnonationalists seeking to anchor their ideas of biological superiority and national distinction in a presumed-to-be-white European past (Hofmann et al. Reference Hofmann2021). In situations where there is no descendant community, insisting on distance between past and present, on rupture, discontinuity and dynamism, on shifting value systems and fundamentally different experiences of the world does not exoticise or dehumanise people who once lived or naturalise their experiences of the world. Even in Australia where cultural continuity extends tens of thousands of years into the past, deep-time persistence does not imply stasis, but reflects the highly dynamic ways Indigenous people found and find to build community, live on Country (an Indigenous term emphasising connection with place and with the non-human entities who occupy it), and respond to a changing world (Kowal Reference Kowal2015).

To finish: totalising grand narratives that exclude outliers and alternatives may be compelling in their simplicity, but they are fundamentally incomplete. More data does not mean greater robustness, and the data themselves are profoundly biased by the questions we ask and the methods of collection, recording and retention we use. Consequently, not only should we expect multiple, sometimes conflicting interpretations, we should welcome them as the best representation of an innately fragmentary body of material under study. The past gives us fodder to think ourselves into different futures, we should be mobilising the unknowns and incompletes to explore alternatives to an increasingly unsupportable status quo. Reconstructing human pasts, and telling these stories, is a profoundly political act and one we should conduct—with care—to address the inequities of our present reality and prefigure a world without domination.