In Born losers: a history of failure in America (Reference Sandage2005), historian Scott A. Sandage traces how, through the course of the nineteenth century, business failures gradually morphed into personal failures. Where losing money initially meant just that by the later nineteenth century, as the narrative of the ‘self-made man’ took hold, it came to be seen by society as a personal shortcoming and framed as a moral judgement. Fast-forward to the big-tech era of the twenty-first century and failure has become a trophy rather than a scar. Silicon Valley's credo of ‘fail fast and fail forward’ entrenches failure not only as a standard element of business practice—start-ups are expected to fail, their founders slated to move forward on their path to success—but also as a commendable addition to a CV or resumé thought to reflect ambition, innovativeness and resilience (see critique in Myers Reference Myers2019). This admittedly truncated narrative of failure in America, closely intertwined with capitalist profit-seeking, serves to illustrate that failure is not a neutral concept but rather a social phenomenon, the reality and valence of which are context dependent. Moreover, like all social phenomena, failure has a history.

To point out the context-dependency of failure is not to subscribe to the cultural relativism which Price and Jaffe (Reference Price and Jaffe2023) rightly identify as unhelpful as it risks explaining away real failures in the archaeological (or ethnographic) record. I wholeheartedly agree with the authors that attention to ‘small-f’ failure is long overdue, and that traces of such failures abound in archaeological datasets if only we look for them. But we differ on how and why to look for failures. Failure is more than an error or mistake: it is a social phenomenon. I have elsewhere defined failure as instances when risk catches up with plans (Van Oyen Reference Van Oyen2023). This is an etic definition, but one that takes into account both the temporal and the social scale-dependency of failure that the authors helpfully foreground.

Seeing failure as a social phenomenon also has an impact on its archaeological legibility. It encourages us to investigate failure not of a person or a thing but of a relational assemblage—in Latourian terms, a “program of action” (Latour Reference Latour1994: 32)—by centring the question ‘what fails?’. In the ethnographic example given in the debate article (Price & Jaffe Reference Price and Jaffe2023), for instance, Ongka's first attempt at holding a moka still fails even if we extend our temporal zoom. We can say that the first moka failed (as a result of a relational cascade including gossip and someone else's death), while the second succeeded. This does not in itself tell us anything about whether Ongka failed or succeeded at holding onto power, which is a different project. Projects shift, change nature and dimension, and feed from alternating assemblages (Bennett Reference Bennett2010; on failure and relationality, Joyce Reference Joyce, Bille and Sorensen2016). Aside from being scalable, such a project-dependent framing of failure creates space for a robust notion of risk by acknowledging both the internal (mistakes) and external (misfortune) causes behind failure.

An etic definition is necessary to start identifying failures but does not in itself provide a history of failure as a social phenomenon. For instance, the failed attempt at Bronze Age cheese-making in Denmark cited by the authors is noteworthy not just because it failed but because the pot was subsequently thrown away. What was the response to the failures we identify archaeologically? Was it cheaper to throw away a pot than to clean it out and try again? Was cheese-making as a whole abandoned in the community? Or was this merely one failed attempt on a journey of trial and error? Under which conditions did failure lead to tweaking, improvement, new attempts (as per the Silicon Valley model) and when did it, instead, herald the end of a certain project? Far from trivial, small-f failures need to be slotted into their social worlds.

By referring to a relational assemblage as the scalable unit of failure, I am less interested in questions of intentionality—which become blended in a distributed matrix of cause and effect (e.g. Robb Reference Robb2013 on emergent causation)—and more concerned with questions of power and inequality. Simmering beneath Sandage's study of the history of failure in nineteenth-century America is the hidden presence of all those who could not fail, in particular women, Black people and the poor. Failing at something means being recognised as trying to do or achieve something (a ‘project’ or programme of action), which requires one to be seen socially. I have elsewhere stressed the link between failure and the privilege of making plans, which entails the ability to foster expectations against some future horizon as opposed to the drudgery of living day to day (Van Oyen Reference Van Oyen2023).

I therefore posit that archaeology's reluctance to see failures has less to do with wanting to populate the past with clever individuals and more with turning a blind eye to certain power structures. Failure is not strictly in the eye of the beholder. What counts as failure is decided by whatever is the hegemonic view of society (Le Feuvre Reference Le Feuvre and Feuvre2010: 12–13). So it is that some queer people and Black people have reclaimed the label ‘failure’, which had been disproportionately reserved for them (e.g. Wilderson Reference Wilderson, Jung and Costa Vargas2021 on Afropessimism). If, for example, the hegemonic model of success under neoliberal capitalism consists of producing profit and offspring in a heteronormative family, then queer people have no chance but to fail at this project. Some scholars, such as Jack Halberstam (Reference Halberstam2011), now ask what might it mean to stop trying to achieve a project that is bound to fail and instead to embrace the predicament of failure? This question might seem far removed from the mundane, small-f failures on which the authors aim to shed light. Yet psychological studies show poor children's inability to improve school test results: paralysed by their internalisation of failure, they are unable to see failure as a stepping stone towards growth (Claro et al. Reference Claro, Paunesku and Dweck2016). Moreover, educational literature stresses how a narrative of personal responsibility for resilience in the face of failure is interlaced with privilege (Hallmark Reference Hallmark2018). The reality of and response to small-f failures is strongly shaped by, and in turn reproduces, existing power structures.

Unearthing small-f failures in and through archaeology is, therefore, of even greater importance than the authors claim. Recognising failure offers not just a more realistic past in which things went wrong and some people were not as skilled as others, but also a past alive to the aspirations, struggles and inequalities that make up the social fabric. More than a data blind spot that needs to be filled in, failure is an epistemic that allows us to see the past differently.