The verification of ethnographic data: A response

Reading Robert Pool’s opening statement to this debate, I ran into this line: ‘The growing culture of accountability and audit in academic research is beginning to impinge on ethnographic research.’

Not too long ago, this sentence would have provoked me to make an argument against what I consider to be a kind of ‘institutionalized distrust’. I would have written about bureaucratic fantasies, which suggest that ever tighter control mechanisms might heighten the reliability of ethnographical accounts. I would have stated that, yes of course, mistakes are made, fraud is committed, data are ‘massaged’ to fit hypotheses, fieldwork experiences are embellished; then to add that sooner or later such issues would come to light, as the anthropological community – ever more sizeable, ever more interconnected, ever more critical – has already proven to be capable to unmask such strayings for what they are. ¹ I would then have asked a number of questions. Instead of seeking to harden the veracity of our narratives by delivering ‘the facts’ on which our arguments are built, should we not continue to thoroughly question the facticity of the fact (Taussig, 1993: xvii)? Should we not research how scientific facts are produced (Latour, 1993, 2010), and analyse the role of rhetoric, aesthetics and style in producing auras of factuality (Clifford, 1988; Van de Port, 2016)?

I still sign up for these thoughts. I do, however, realize that times have changed.

In 2016, ‘post-truth’ was selected as the Oxford Dictionary word-of-the-year. The online version of the dictionary offers a sentence so as to convey the meaning of the term, saying: ‘in this era of post-truth politics, it’s easy to cherry-pick data and come to whatever conclusion you desire’. ² I would say that pretty much sums up what we are seeing all around us. Everywhere, the talking is about ‘the demise of a common world of facts’, falsehoods, ‘alternative facts’, fact-checking, fake news, ‘fact-free worlds’, lies, ‘totally dishonest media’. Alarmingly, populist leaders seek to communicate their (often evidence-free) ‘truths’ directly with their electorate via twitter and social media. They say: Who needs the checks-and-balances of journalism? Who needs experts? Who needs anthropologists? In Slate magazine, commentator Jacob Weis states:

We never had a president who didn’t accept the existence of objective truth. We don’t know how to make him recognise that reality has an existence independent of his preferences. It’s like we are suddenly covering a post-structuralist literary theorist, albeit one with orange hair. ³

Understandably, pressures in journalism to invest more in accountable knowledge are mounting. Standing practices such as clearly distinguishing news reports from ‘opinions’ are being reinvigorated. And there are loud calls to hold on to ‘the facts, the facts, and nothing but the facts’.

Perhaps the pressure should be up in anthropology as well. Perhaps a renewed interest in making explicit what actually goes into the claim that our statements are ‘empirically grounded’ would be appropriate. I would think, however, that this can never be a return to an uncritical embrace of ‘the facts, pure and simple’, or an all-out rejection of the insights we’ve obtained from post-structuralist literary theorists – however much discomfort these may cause in the current political situation.

It is with this sense of urgency that I enter this debate. Robert Pool convincingly sketches the many difficulties a striving for accountability produces due to the particular nature of our modes of doing research, as well as the intellectual legacy of the Writing Culture debates. He ends with three suggestions. Ethnographers ought to store and archive whatever it is that makes up their basic fieldwork material. In their writings, they ought to be more transparent about the basics of their fieldwork, providing information about such elemental givens as the number of people with whom they have been working, the length of their stay in the field and their proficiency in the local vernacular. They also ought to give more voice to their interlocutors in their final reports: not merely the obligatory ‘quotes’, which are often groomed to serve the ethnographer’s arguments in the making, but longer narratives that might actually reveal up to what point their arguments coincide with what their interlocutors were saying. (I would add that this latter point would require more flexibility in the rigid formats of academic publishing, with their strict conventions of word-limits, textual lay-out, etc.)

In and of itself, these are useful directives. I do feel, however, that they do not take sufficiently into account the many intellectual divides in our discipline. Take, for instance, differences between ethnography and anthropology. In Tim Ingold’s understanding the former has the objective ‘to describe the lives of people other than ourselves, with an accuracy and sensitivity honed by detailed observation and prolonged first-hand experience’ (2008: 69), whereas the latter seeks ‘a generous, comparative but nevertheless critical understanding of human being and knowing in the one world we all inhabit’ (2008: 69). Pool’s suggestions might work for ethnography, which, however critically, holds on to a separation between the ‘material’ brought from the field and its ‘interpretation’, and subscribes to the ideal of accuracy. Yet what about accountability in the more anthropological strands of our discipline?

Allow me to take my own work as an example. If I were to choose between Ingold’s categories, I guess I would classify it as anthropology. The questions that shape my research are more along the lines of ‘how do human beings handle the predicament that the world does not fully comply with their narrations of it’ than ‘what is the role of secrecy in Candomblé’? I may ask the latter question, and try ‘to get it right’ (as Kirsten Hastrup (2004) once described the ethnographic striving for veracity), yet the answers I will come up with are taken to issues not restricted to Candomblé life worlds. This in no way diminishes the crucial importance of fieldwork for my writing (and more recently filmmaking). Yet the field is not an object I seek to conquer with the tools that scientific methods offer me, then to make authoritative statements about it in the vocabularies of academic discourse. It is first and foremost the site of an encounter with otherness, a place where I meet up and engage with people who – up to a point – think differently, feel differently, do things differently. This encounter provokes a critical questioning of idées reçues, taken for granted viewpoints and ‘common sensical’ understandings – including those of academia – and it is on this questioning that I report.

Unsurprisingly, then, my preferred mode of report is the essay. I take the essay to be a kind of thinking-out-loud, geared towards the production of food-for-thought rather than conclusions about the life of others. It is a form that is open to inquisitive thought experiments, or, as Adorno famously argued, ‘the essay’s innermost formal law is heresy’ (2000: 110). The essay form recognizes subjectivity in its use of the authorial voice. It often introduces what one might call a critical auto-referentiality, as when the essay writer critically observes his own attempts to put the world into words, and explicitly addresses the problems to which such attempts give rise.

In addition to written essays, I make essay films. One of the great charms of this particular genre is the possibility to think-out-loud in the presence of filmed footage, whereby the co-presence of words and images in a single frame – a juxtaposition of abstract thought and concrete digital recordings – reveals the very tension of attempts to narrate the worlds we study. One might be inclined to think that video footage offers direct ‘proof’ of having been present in the field, and solves some of the basic evidentiary information Robert Pool is asking for (the local voice; glimpses of the kind of rapport between anthropologist and interlocutor; the proficiency of the anthropologist in the local language, etc.). However, debates in visual anthropology have made it abundantly clear that many questions must be asked about the kind of ‘evidence’ footage presents. The essay film asks these questions, not after the film has been shown, but in the film.

In the kind of anthropology that I do, what would ‘accountability’ be? Or to phrase it in Pool’s words: how could I be called to account for the nature and status of the data on which I base my interpretations? I realize that the way ‘data’ appear in my kind of anthropology recalls the way Matthew Engelke (2008), in a special issue of the JRAI on evidence, described the ‘oddity’ of evidentiary protocols in our discipline:

… the penchant we have for using a vignette or anecdote about what we observed one Tuesday morning in an open-air market outside Timbuktu eighteen years ago to explain the workings of political power in Mali, or African economies, or globalization. (2008: 12)

It is true that my data are first and foremost recollections of intercultural encounters – partly written up in diaries and note books, partly simply ‘in my head’, where they get triggered by the process of thinking itself. ⁴ Seen through the prism of ethnography this may sound like a rather flimsy base to produce significant statements (and I am always aware how lucky I am to study a field with a tremendously rich ethnography that I can consult). Seen through the prism of the kind of anthropology that I perform, however, these recollections prove to be an incredibly rich source for thinking.

I’m inclined to argue that there are other forms of ‘accountability’, which do more justice to the quality (or lack thereof) of anthropological musings on human existence. What my brief exposition of the essay form has sought to underscore is that the problem of accountability is always already present in the discussion, as the inquisitive voice brings up thoughts and observations, which will then be questioned, nuanced, amended, redirected, rejected or replaced by better ones. The argument might typically be: ‘So this is what I saw, but up to what point is what I saw indeed what was happening?’. Or: ‘The priest told me that Candomblé is a religion, but how did the interview situation produce this statement?’ Accountability would here be the thoroughness of the self-interrogation, the commitment not to settle for the ‘easy’ answers, and the transparency with which the thought process is presented. I would add that the essay’s stress on subjectivity also enhances accountability, as the author will reveal from what position he or she is speaking, and question how that position might have impacted the pronouncements that are being made. Essayistic transparency, one might say, might generate an alternative kind of ‘accountability’: one which allows the reader (or viewer of an essay film) to be witness to an argument in the making, and assess where the arguments that are being put forward come from.

Which brings me to a last point. One could say that the propositional nature of the essay, it’s steering away from conclusions, immunises this form of report from fraud. If the essay form is indeed resistant to the kind of fraud audits seek to prevent, it is because of what the genre pretends. It never promised authoritative statements about some elsewhere. Doubt, provisional knowledge, suggestions, informed speculation is what it is made of. This suggests that, in our ‘post-truth era’, anthropologists could actually state what kind of veracity they claim for their text. Do they try to proof something, or offer their readers food-for-thought? Do they pursue descriptive accuracy or creative argumentation? Such explicit statements will then generate different sets of criteria as to what accountability would be. Such an approach to issues of accountability has the merit that it recognizes anthropology and ethnography as knowledge practices with different empirical groundings, and accepts that there is therefore no single evidentiary protocol that ‘fits all’.