Realist evaluation, mechanisms and theoretical minimalism

At IACR earlier today I heard two interesting talks about Realist Evaluation. I had previously had a vague idea about what this involved, largely through encountering citations from Pawson in other texts, without ever having really grasped what it was in a concrete sense. Now I have, I’m very interested. All the more so because of the number of people who have told me today about the excitement that realist evaluation has generated in environments that one would have expected to be utterly hostile to critical realism more broadly. So there’s an interesting question about how realist evaluation has proved so amenable to circulation outside of its initial domain. Certainly, some of this must be a matter of networks, in terms of the original generation of those advocating this approach and the patterns of work stemming from their own engagements. I imagine it’s also a matter of analytical value – I was particularly interested to hear how realist evaluation appears to those from an applied background working within a positivistic intellectual culture. It sounds like the conceptual distinctions it draws are crucial – realist evaluation critiques what positivist empiricism does but also explains how its own function corrects these mistakes by going deeper into the case.

However the aspect that interested me most were the questions about intellectual self-presentation. Firstly, in terms of the limitations upon the acceptability of realist evaluation that some people encounter e.g. the approach is becoming fashionable but the results may still be somewhat alien on an intellectual level to funders. Secondly, in terms of what strikes me as the terminological minimalism which characterises the approach and its proponents. Take this extract about mechanism in a paper i just found by Pawson and Tilley:

The concept is best grasped through an illustration. The ‘primary school breakfast club’ is a very popular measure used to boost early education performance, often included within community regeneration initiatives. The key point here is that ‘the measure’ is not the basic unit of analysis for understanding causation. A measure may work in different ways or, in realist parlance, they may trigger different mechanisms (M1, … , Mn). A breakfast club may aid classroom attentiveness by offering the kids a ‘nutritious kick-start’ (M1) to the day, which they might not otherwise get. And/or it may act as a ‘summoning point’ (M2) to prevent kids loitering or absconding or misbehaving in the chaotic period before school. And/or it may act as an ‘energy diffuser’ (M3) to soak up gossip and boisterousness before formalities commence.

And/or it may enable to school to present a more ‘informal face’ (M4) to those uninspired by classroom and book learning. And/or it may act as a ‘pre-assembly’ (M5) enabling teachers to troubleshoot potential problems and seed the day’s schedules. And/or it might give parents and school staff an ‘informal conduit’ (M6) to mix and offer mutual support. Mechanisms also explain a programme’s failure, of course, so to this list we might add some adverse processes. It may act as an  opportunity for ‘messing about’ (M7) if only ancillary staff are on duty; it might provide an unintended ‘den of iniquity’ (M8) for planning the day’s misdeeds: or it might prove a ‘cultural barrier’ (M9) because inappropriate food is served, and so on.

I’ve long thought that mechanism is a powerful concept. In fact encountering the notion of a generative mechanism, in virtue of the operation of which events unfold in the way that they do, played a crucial role in winning me over to critical realism in spite of my initial scepticism. Even the more instrumental conception of mechanism found in analytical sociology appeals to me because once you start to think in terms of mechanisms, it’s hard to understand how anyone could be satisfied by a form of social inquiry entirely absent of them, even if you may disagree with the way in which other people conceptualise them.

But it can also be a hard concept to explain to those who don’t think in these terms. They are also often not written about clearly and, in spite of what I’m suggesting is their analytical pay off, I can understand why this is the case. In my own work I’ve tended to use ‘mechanism’ as a vague concept I employ in my provisional analysis but then articulate in other terms at the point of writing. The reason for this is partly because I don’t have the confidence that I can write clearly in terms of mechanisms while also being accurate. Or sometimes, if I’m honest, it’s because what I’m bestowing the ‘mechanism’ title to actually just reflects a causal hunch. Or my own thinking is much vaguer than I would like it to be.

It’s in terms of this experience that I find the clarity of Pawson and Tilley’s writing about mechanisms so striking. What I see as problematic, using it to designate the fact that I think I’ve identified operative causal power of some form, seems utterly fine when I encounter it in their writing. I find it problematic in my own because of the vast meta-theoretical edifice in virtue of which the concept is meaningful to me. But perhaps this is a feature of my own intellectual biography rather than anything that should exercise normative power in relation to my own writing? Could this be a more general mistake i.e. assuming that the theoretical considerations that led you to come to have accepted a concept should figure into your applications of that concept?

I look forward to reading more of Pawson and Tilley’s writing and I’m interested to develop my understanding of the style they write in. The small amount I’ve looked through in the last hour or so certainly fits with the effusive complements I heard about their style during the conference today. I’m also intuitively of the opinion that this style is, in a very particular way, part of the reason for the success their work has enjoyed. I’m not making the rather trite claim that ‘theory would be more popular if theorists wrote more clearly’. I’m suggesting there’s something very specific about their particular kind of clarity which lends to their work a wider popularity than would otherwise be the case.

My suggestion is that it uses the minimum of terminology necessary to convey the conceptual distinctions which have practical implications. It’s stylistic parsimony. Or at least it tends towards this. Are there many other theorists this is true of? I’m not convinced that there are. I’d like to be one of them though.

Categories: Outflanking Platitudes, Sociological Craft

Tags: , , , , ,

1 reply »

  1. Do mechanisms have simplicity or relative autonomy ?
    Within critical realism, mechanisms belong to the real intransitive domain. But one can postulate that mechanisms belong to all three levels of stratified reality : the intransitive, the actual and the observer (empirical). One is then free to create mechanisms to observe the real as others are free to both create and evaluate their mechanisms. Mechanisms that become widely adopted and institutionalized then become part of the intransitive domain. Simplicity is then not an attribute of mechanisms, as they can become complex and embedded, but (more importantly) simplicity is an attribute of elementary human action that can be explained is clear and simple terms but this can get lost in an overemphasis on the intransitive domain. The key feature of mechanisms is their relative and initial autonomy from the intransitive domain.

Leave a Reply

Your email address will not be published. Required fields are marked *