The DH Awards include a category for “Exploration of DH Failure/Limitations”. During the DH Awards of 2025, the category did not receive enough nominations to merit its inclusion. Additionally, although it was listed as a category in last year’s call for nominations, it was absent in the voting and results, which suggests a similar pattern. Despite being a minor development, this exclusion is something I have been thinking about.
Two pieces published in The DO this year have stayed with me for a similar reason: Alice Cassalini’s “Buddhism and AI: Another Look” and Soni Wadhwa’s “Sindhi Halchal Archive”. Both are substantive pieces on their own terms, but each contains real candour around failure, with Wadhwa discussing what went wrong with a collaboration and the lessons she learned from it, and Cassalini returning to a piece she wrote for The DO in 2024 and reconsidering a statement upon which her piece rested. This caught me off guard; these reflections often seem confined to coffee breaks and corridor conversations, meaning that they remain private, privileged insights available mainly to close colleagues and those able to travel to conferences.
The lack of publication of failed or inconclusive research is a known problem in scientific fields, with almost 50% of clinical trials for health not being published (Science and Technology Committee (UK), 2018), some due to difficulties getting to publication, but often from a sense of failure and the stigma around publishing results contrary to the hypothesis (Satalkar et al., 2022). Similarly, failure and setbacks are part and parcel of digital humanities (and most other things), so why do they rarely make it to publication?
I think this is partly due to failure being a difficult topic to discuss, particularly where this doesn’t provide a tangible outcome. In research, people are usually willing to accept failure when it results from ambitious, “high-risk, high-gain” projects that have the potential to achieve something significant. They are much less willing to discuss projects that simply never finished what they intended to do, or collaborations that fell apart. We have a language for praising bold experimentation – like the well-known story of James Dyson and his 5,127 prototypes. But we have a far more limited vocabulary, and far less interest, for the quieter, more complicated, and more common kinds of failure that happen in everyday work.
I know something about this. Early in my research career, I joined a project that began as a pilot study. As can happen, it gathered momentum quickly – more funding, more collaborators, DH advancements that made more possible – but we failed to pause and ask whether our capacity and ways of working had kept pace with these changes. Decisions were made, revised, and changed without being documented. The workflow didn’t account for or adequately adapt to the scale of disruption the COVID-19 pandemic brought. Not enough time was allocated to reflect and revise as things unfolded. Eventually the project concluded without completing what it had set out to do. At first, I was uncomfortable even discussing the project, as I felt that the failure reflected something about me. I have failed since, in different ways, and have come to understand and appreciate it as part of the experience rather than an indictment of it.
What I learned from that experience was that the problem wasn’t the fault of any single individual, tool, or decision. In this particular case, it was the absence of a structure that treated the natural iteration and course-correction as expected parts of the work, rather than signs of failure. I subsequently read more around workflows and came across action research frameworks: iterative cycles of planning, action, and reflection, in which what doesn’t work is treated as data. Would an action research approach have guaranteed a different outcome? Not necessarily. But it reframes the question away from what went wrong and toward what can be learnt. It offers an understanding of failure that encompasses compassion and clarity.
This is the terrain that Anna-Maria Sichani and Michael Donnay explore in Reframing Failure in Digital Scholarship (2025) – which received an honourable mention from the DH Awards – a collection structured around four themes: innovation, technology, collaboration, and institutions. Its ambition is to develop a space for failure that is critical, constructive, and compassionate, and to move failure from the periphery of how we understand digital scholarship toward its centre (a similar aim to Shawn Graham’s Failing Gloriously and Other Essays (2019)). The collaboration section particularly resonates with me, not least because Sichani and Donnay make a point that is easy to overlook: the unwillingness to discuss when and how collaborations go wrong can leave future projects ill-prepared to face similar challenges, and keeping these conversations informal privileges those with access to the right networks while excluding those without this advantage. In my experience, DH is almost always a team endeavour, regularly drawing together wide ranges of expertise, institutions, and communities of practice. And yet we spend little time discussing how to do it well, and even less on when and why it goes wrong. The book doesn’t offer easy answers. It does something more valuable, in my opinion: it normalises such conversations.
These conversations feel overdue. As an early career researcher, I know that I have sometimes made decisions shaped by a fear of failure and of how it might reflect on me personally or on my career. Yet if the purpose of publishing in a field such as ours is to improve practice and shape future work – not simply to celebrate successful outcomes – then we need to be open to the likelihood of failure, and willing to examine and write about what did not work, why it did not work, and the context in which failure occured. What this looks like will differ for each of us, but perhaps the first step is simply to make space for reflection and discussion, both formally and informally.
And if you would like to submit a guest piece to The DO about your own learning from failure/limitations, please see the submission guide here.
References
Shawn Graham, Failing Gloriously and Other Essays (The Digital Press at the University of North Dakota, 2019)
Priya Satalkar, Stuart McLennan, Bernice S. Elger, Erik von Elm, and Briel Matthias, “Investigators’ Sense of Failure Thwarted Transparency in Clinical Trials Discontinued for Poor Recruitment,” Journal of Clinical Epidemiology 145 (2022): 136-143, https://doi.org/10.1016/j.jclinepi.2022.01.024
Science and Technology Committee (UK), “Failing to Publish Data from Clinical Trials Presents Risk to Human Health”, 30 October 2018, https://committees.parliament.uk/work/4644/research-integrity/news/100871/failing-to-publish-data-from-clinical-trials-presents-risk-to-human-health/
Anna-Maria Sichani and Michael Donnay (eds), Reframing Failure in Digital Scholarship (London: University of London Press, 2025), https://doi.org/10.63674/yrab7358
Most of the failures are “self inflicted” by academics and academic institutions that don’t agree with the results and refuse to publish.