That title is funny (just in case you hadn’t noticed). This has been in draft for ages…publishing now to stop myself from looking at it again :-). One of the things I’ve grown interested in is the (philosophical underpinnings of) Discursive Psychology and related approaches (although I’m less keen on the sort of conversation analysis it sometimes focuses on). I’ve particularly been interested in the work of Derek Edwards and John Potter, and the relationship of DP with sociocultural theories of learning (e.g. Edwards co-authored a book ‘Common Knowledge’ with my previous supervisor Neil Mercer, which neatly outlines Neil’s perspective on language and its function). One of the things DP has been particularly critical of is the use of qualitative analysis in various problematic ways. In reading this, and thinking about it myself, I wonder if we can mark particular sorts of fallacy in qualitative work, for example these three: 1. The constructs are reified in the methods (the circular definition issue) – “what epistemic beliefs do you have?”, “ah well, discussing this shows…” 2. The constructs are reified by the method (the surveyor’s dilemma) – “do you like x or y?” (when neither is the case) 3. The constructs are reified by environment/genetics/neurological factors, etc. – the naturalistic fallacy – “x and y are the case, thus they are immutably so” The first of these is an issue with coming up with coding schemes or interviews which try to probe a construct, but will almost inevitably create those constructs in so doing even if they would not ordinarily be “deployed”. So the sequence goes: 1) create some scheme/schedule from the literature to probe some issue; 2) implement it; 3) find that your data reflects the presence of your construct. This is problematic because it’s utterly unsurprising – the method to probe the construct reifies that construct as an object of enquiry; but it doesn’t mean the construct ‘actually’ exists, or that it’s salient to real world action, etc. This issue is particularly about a failing to recognise the discursive properties of the method. The second of these is another version of the first, in which survey writers (or interviewers, etc.) create well defined tightly conceptualised (non-discursive) constructs into which all responses must fall…I’m reminded of Tim Minchin’s great line in his 9 minute beat poem Storm **“a pigeonhole starts to form. And is immediately filled with pigeon“. **The most obvious cases of this are in psychometric approaches such as the introversion/extraversion scales, but actually they exist in lots of more ‘talk’ based methods too. While newer approaches to scaling (which don’t rely on binary responses) address this issue partly,most approaches still seek to create ‘averages’ – when actually, the anomalies are the most interesting element (and its these that are discarded as not fitting into a construct). This issue is particularly about a failing to recognise the discursive properties of the construct. The third of these is interesting because it’s common in lots of methods and their approach to culture. The criticism applies to lots of methods/approaches – neuroscientists point to neurological evidence of some thing or other, failing to acknowledge the influence of environment on development (and on the way that development is played out in the real world – the bio-social element); medical professionals point to something as problematic, failing to acknowledge the role of interpretation and medicalisation in the creation of disorders, and so on. In this case, my concern is that sometimes we find “things” out – uncover constructs, explore how particular groups interpret their place in society, etc. and then that’s it, there’s no challenge to the normative issue regarding why people might value certain things over another and the role of wider society in these interpretations. This issue is particularly about a failing to recognise the discursive properties of society at large. I see a lot of material being produced regarding prospective student’s understanding of their financial situation, and their likelihood of going to university, often people suggesting that enrollment won’t go down because of the raised fees – but this fundamentally misses the point that just because people will go to university, doesn’t justify the fee increase – it is a statement of fact not an argument of right/wrong. Anyway, I’m sure there are other cases too, and it’s the first one which I’m particularly interested in (design an experiment based on literature/concept, find that concept, claim support for that concepts presence), and of course, a) it’s hard to get out of and b) it doesn’t preclude this sort of research being useful (particularly if we artificially use such things in other contexts like exams) and being good proxies for useful discursive practices. I do, however, think people (including me) should read more discursive psychology :-). Thoughts @sjgknight or in the paragraph level comments