Masturbation journal paper exposes deeper problems in research

A peer-reviewed article based on masturbation to paedophilic comics as an autoethnographic method has rightly sparked condemnation within and beyond academia.

The article by Karl Andersson, a University of Manchester PhD student, has now been removed from Qualitative Research’s website. Its ethical failings have already been scrutinised by journalists and academics. But a key question remains outstanding: how has qualitative research, ostensibly under the banner of anthropology, got to the point where something like this could even be considered scholarship in the first place?

The answer lies in a combination of cultural failings regarding methodological rigour and the politicisation of methods. To be clear, this culture cannot be blamed for the ethical failures of this case, responsibility for which lies with Andersson and those associated with the work’s publication. But I cannot help feeling that something so astonishingly stupid as a masturbation autoethnography was sooner or later going to get published as though it were real research. Quite apart from its depravity, Andersson’s article is appallingly bad scholarship, from methods to evidence to inference to conclusions. But not exceptionally so when compared to some other articles that pass peer review.

One response has been to list examples of good research in a defence of qualitative methods. Laudable, but the issue is not that no good qualitative research exists – plenty does. Rather, the disciplines involved, including my own field of social anthropology, in which Andersson regrettably lists his “expertise”, have fostered a culture in which bad qualitative research can emerge.

How has this come about? Ethnography, after all, can produce excellent, well-informed scholarship. In anthropology, it typically involves many months, if not years, of “participant observation”, living in the social context being studied while documenting people’s behaviour, conducting interviews and so on.

This presents the challenge, long acknowledged by anthropologists, of separating objective empirical observations from subjective interpretation. The best ethnographic research meets this by focusing on what can be objectively documented, assessing this comparatively and being explicit about crucial details and where subjective interpretation begins. This is work that can either stand as empirical documentation in its own right, or go on to inform wider social scientific analysis.

However, since the introduction of “postmodern” approaches to anthropology and the turn to a focus on reflexivity, in attempting to address genuine issues, anthropology has tended towards abandoning any pretence to objectivity in the first place. For some, this means that because objectivity cannot be achieved, it shouldn’t be attempted at all.

It is then a short step to subjectivity becoming a “method” and source of knowledge in its own right. A great deal of effort is to be expended by the ethnographer to explain who they are and how this might affect what they write. At the same time, subjective impressions are embraced such that any vaguely logical interpretation of the data will do, rendering the aforementioned reflexivity methodologically pointless.

If ethnography results from participant observation, but this inherently involves subjective inferences, why bother with the observation part in the first place when you can simply resort to introspection? Hence, “autoethnography”, which in essence takes subjectivity-as-method to its logical conclusion. This might be all well and good as an exercise in personal development, assuming one does not share Andersson’s predilections, but what it produces is equivalent to a diary. It might have relevance as a source of evidence for scholars interested in the internal life and ideas of the author and others like them, or the context in which they live, but in itself it is not research in any meaningful sense.

It is not just the encouragement of non-research methods that is a problem; it is also the culture surrounding methodological standards, or rather, their absence. While ethnographic research is often of high quality, it is not consistently so. Anthropology lacks established disciplinary standards for evidence and data collection, leaning heavily on a philosophy that because each field site and project is different, such things are irrelevant. So, there are no examples of broadly acknowledged best practice for note-taking, interviewing or (God forbid) incorporating relevant quantitative methods, much less for how best to gather data on specific social phenomena.

As a consequence, the raw materials for ethnographic writing take all sorts of forms, and are typically not exposed to the wider scholarly community until they appear in a curated form in an article, monograph or thesis. Unlike archives or experimental records, field notes themselves, the closest thing to ethnographic primary data, are not accessible to other scholars for assessment and critique. This means all manner of subjective interpretations can creep in, becoming indiscernible from other observations.

Anthropologists might protest that field notes cannot be meaningfully understood by others given how personal they are, or that sharing them could jeopardise participants. These are not sound objections; safeguards are easy to implement, and it is perfectly possible to document notes in such a way that they are readily understandable to others. There is no reason why general standards for field note-taking, with the aim of scrutiny by peers, could not be established.

This is doubly necessary as the nature of participant observation makes it very unlikely that another anthropologist would be able to go to a field site to judge findings first-hand. Psychology is currently undergoing a crisis as large numbers of findings have not replicated across subsequent studies – a problem, but one that is at least identifiable and addressable from within the discipline. Anthropology couldn’t have a replication crisis in the first place.

We have a situation in which subjectivity is lauded methodologically at the expense of objectivity, and where adequate standards for peer scrutiny of evidence are absent. But that isn’t the whole story. The legacy of anthropology’s complicity with colonialism has long soured practitioners’ relationship with the supposed objectivity of early approaches, but rather than seeking better ones, a tendency has been to eschew objectivity not just epistemologically but ethically.

Today, this is increasingly combined with the language of “critical” studies, expressing the conviction that objectivity is a morally suspect tenet of colonial epistemology. Ergo, subjectivity should be embraced not just as an epistemological but as a moral and political project.

This has two effects. First, it exacerbates existing methodological problems, subordinating rigour and empirical enquiry to a priori assertions, typically of abstract forces such as power relations that are underdetermined by the empirical material presented. Documenting and analysing evidence to draw conclusions or arrive at explanations is less the order of the day than a kind of theological exegesis, in which these assertions are read into ethnographic case studies.

Where we once at least attempted to separate our subjective politics from our objective scholarship, understanding that the latter might inform the former but that the two are not the same, we now conflate them. Andersson does this himself, presenting his work (ludicrously) as an edgy political project of resistance to the “yoke” of Christianity.

The second effect is to militate against adequate methodological criticism. Qualitative researchers, even while pointing out that what Andersson did does not constitute research, have focused on his failure to acknowledge his race and thus his positionality as a researcher. Note that we are talking about a man masturbating to paedophilic cartoons; it is less than immediately clear that that is primarily an issue of race and identity, even if these play a role.

Had Andersson considered his positionality, we would still have the same article, just with an additional inconsequential paragraph paying lip service to reflexivity. However, a general retreat to subjectivity makes such forms of critique the only conceivable option. There can’t be a methodological problem here, because acknowledging it would undermine the epistemological politics of subjectivity-as-method. Meanwhile, it is left to researchers in quantitative disciplines to address Andersson’s lack of rigour in any depth.

This methodological politics is potentially lethal for serious research. In a wider context of political polarisation, it results in knee-jerk defence of work such as Andersson’s and a fear of voicing dissent. At a time when cuts to humanities and social sciences pose a genuine threat to serious disciplines, it is up to academics to ensure that sub-par scholarship, and the practices that facilitate its production, are dealt with.

Failures cannot be blamed on a lack of due attention to positionality or colonial epistemologies. Doing so is symptomatic of the culture of fetishising subjectivity, eschewing empirical verification, and politicisation of methods that has allowed something like autoethnographies of paedophilic masturbation to even be entertained as credible scholarship.

William Matthews is an LSE fellow in the anthropology of China at the London School of Economics.