I have just come back from a couple of conferences with a focus on qualitative research. This time, however, I was particularly struck by the frequent use of the phrase ‘discourse of’. In this post, I want to comment on it. As ever, I want to stress that these are only my impressions and they do not pretend to be the truth.

I wrote about the use of the phrase, when I wrote that not everything is a discourse. This time, however, I felt that the use of the phrase went further than before. I felt that if an informant had mentioned loving someone, we glimpsed a ‘discourse of love’, if they had said they were a patient, we obviously dealt with a ‘discourse of patienthood’ (these are invented examples, as I do want to keep my comments outside particular papers). If you encounter something slightly symbolic or abstract, you can immediately claim its discourse. That also means that it is unlikely that there would be discourse of the table, even if the informant sat and had a party at one. Tables, it seems, are too mundane. Who knows what awaits us, though!

I wish I had counted how many ‘discourses of’ I heard of, yet, I think there were at least two dozen.  And I want to make three points with regard to such claims about the existence of these various discourses.

My first point is that I have not heard evidence for their existence. Inevitably, when a particular word or a particular interpretation would appear in the paper and, hey, presto, we’ve got a discourse of something or other. Now, in some cases, the patterns observed by speakers were indeed claimable and the argument behind them was solid. The problem is that a pattern, whether linguistic, narrative (I’m not certain there’s a difference here, but I don’t care at the moment) or based only on contents of what is said/written, is not enough to claim existence of ‘a discourse’. This is because such claims must, I think, be more general and based on more than a pattern or two in a corpus or a few interviews.

The absence of evidence, I think, is at least partly to do with the fact that speakers do not stop to consider what exactly they claim such a discourse to be and I am fairly sure that there were papers in which the word ‘discourse’ was used in a number of ways within one presentation. For example, I keep being surprised that people who claim to do text-based discourse analysis (such as CDA, for example) use a nebulous, macro-scale Foucauldian concept of discourse which is fairly difficult to pin down empirically if you’re basing your arguments on micro-discourse analysis.

That said, I think that the most frequent understanding of discourse is implied to be an ‘order of discourse’, which can be defined as something like the totality of discursive practices, and the relationships between them in a particular sphere of life (I roughly follow Norman Fairclough here). The problem is that in order to make such claims, you need considerably more data than, for example, a dozen interviews. So, just because a number of interviewees say things in a particular way, however consistently, this still doesn’t constitute ‘a discourse’.

My second point is that claims about the existence of a ‘discourse of’ tend to be made in the singular. And yet, to offer a common example, I doubt very much that there is one, homogeneous, one-fits-all ‘discourse of psychiatry’. Yes, there is probably a dominant one (though I am not entirely certain what it is and how exactly to identify it and no, using references to nosological diagnoses is very unlikely to constitute it), but there are also a number of others. This is because those speaking of a ‘discourse of psychiatry’ tend to speak of a few discursive practices in the emanations of institutional psychiatry such as diagnostic manuals. And, yet, it is very easy to argue that discourses of psychiatry are also forged in psychiatrists’ surgeries (and it is extremely unlikely that all shrinks speak in the same way), on social media, in the press, let alone by patients and their groups. Isn’t a ‘discourse of anti-psychiatry’ also part of a ‘discourse of psychiatry’? I would argue that it is. It is therefore extremely unlikely that there is one homogeneous discourse of psychiatry (and it’s to do with diagnosis).

Incidentally,  we can also usefully ask the question of what exactly do people talking about a ‘discourse of psychiatry’ talk about? And I must say that I have no idea how to answer such a question.

Now, the final point I want to make is the most important one, I think. You see, I keep wondering why qualitative researchers talk about these singular ‘discourses of…’. And I suggest that, more or less unwittingly, they want to make a claim outside the scope of qualitative research.

I think, and I stress the weakness of my claim, we, the story tellers, keep being forced to defend the results of our research. The number crunchers keep accusing us of coming up with anecdotal evidence which doesn’t go beyond the data we acquired. And claims about a ‘discourse of’ provide us with a lovely possibility for a bit of sleight of hand.

As we, story tellers, we cannot generalise from our data, nothing explicitly stops us from speaking of a ‘discourse of’. And such claims offer us a backdoor claim to generalisability. No, we are not saying that our research offers insight into how all/most psychiatrists speak, we are just saying we discovered a ‘discourse of psychiatry’, or mental illness, or diagnosis, or masculinity, or patienthood or something yet different. Unfortunately, such postulates invest our data with the power they don’t have.

I would like to stress that we do not need to avail ourselves of such strategies. We do offer valuable insight into people’s experiences, insight which is important in its own right, or can then be explored further also by number crunching. Just to show a social phenomenon is really enough, we do not need to claim that the phenomenon has the status of more or less universal human experience. And do let’s not.

I want to end this post with two minor points.

First, it’s a point about the relationship between thinking and discourse. The philosopher Tim Thornton suggested to me recently that I underestimate the strength of the relationship, and I think he has a point. I also agree that more attention should be directed to unpicking it. Yet I think even Wittgenstein, on whose shoulders such arguments are made, didn’t make claims as to discourse and thinking. And I do wish I didn’t hear arguments that discourse influences our thinking made with ease and gusto. Does it? In what way does it? Surely, if this ‘discourse of psychiatry’ influenced our thinking, we wouldn’t be able to challenge it, would we?

The second point I would like to make is two-fold. I’ve done discourse analysis for quite some time and, while I am not au fait with all that’s happened in DA in, say, the last couple of decades, I am bewildered by the proliferation of discourse analyses. More and more adjectives are used to qualify the phrase ‘discourse analysis’ and I am completely at loss as to what those adjectives refer to. For me one such ‘discourse analysis’ is the so-called Foucauldian DA (I flag this one up, as I have made such points before), which I discovered a year or so ago. And I still think it tends to be quite thin on what exactly it analyses and how it goes about it. And I wish things were clearer not only in theoretical manifestos, but, more importantly, the data on which its arguments are based.

The other point is about claims that a researcher does a well-established form of discourse analysis, such as Critical Discourse Analysis, which seems to be fairly popular at conferences I attend. I have heard too many presentations apparently anchored in CDA which had nothing to do any form of CDA I am aware of. CDA is a text-based DA which focuses heavily on the lexico-grammatical form of text.  And yet, time and again, I see no references to any such form in papers purporting to do CDA. And I would like to make the point again – rephrasing what the analysed text says is not a form of discourse analysis. Rather, it is, unsurprisingly, rephrasing what the text says. Finally, I would like to suggest that discourse analysis doesn’t reveal any hidden or secret messages in the text.

I keep saying that we, qualitative researchers, are our own greatest enemy. I still think that. I think we should challenge ourselves and each other to produce, interrogate, and reflect on the evidence we produce in our research, as well as, the claims it allows us to make. As ‘we’ fight the number crunchers, feeling under siege, I think we often forget that our primary responsibility is to produce excellent research.