The concerns I’ve seen can be summed up like this: practice theorists (e.g. Bourdieu, Giddens, Schatzki) substitute practices for a more differentiated social ontology that recognises the causal significance of structure and agency, and therefore the concept of practices is irredeemably conflationist (Margaret Archer’s term for approaches that collapse social structure and agency together into one). As Mark Carrigan puts it in a recent tweet “the concept of a practice bakes structure into human activity”. Since the distinction between structure and agency is fundamental to realist social theory, this would imply that realists shouldn’t use the concept of practices.

Archer’s accusation that Giddens and Bourdieu collapse structure into agency are largely justified – Giddens tells us that structure consists of rules and resources, and that both exist as mental properties of human individuals, while Bourdieu refers to habitus – the set of dispositions that each individual has, again a mental property – as “structuring structures”. Oddly, both also have structure-like concepts in their theories (systems for Giddens and fields for Bourdieu) but don’t seem to recognise that these might be relevant to debates over structure and agency. One way of looking at this might be to say that even conflationist thinkers can’t really do without social structures, even if they call them something else.

Some of the more recent advocates of practice theory also seem to be on a mission to replace structure and agency with practices. Theodore Schatzki, in his introduction to The Practice Turn in Social Theory, tells us that “practice approaches promulgate a distinct social ontology: the social is a field of embodied, materially interwoven practices centrally organized around shared practical understandings. This conception contrasts with accounts that privilege individuals, (inter)actions, language, signifying systems, the life world, institutions/roles, structures, or systems in defining the social. These phenomena, say practice theorists, can only be analysed via the field of practices” (p. 3).

This kind of “practice imperialism” is just as unacceptable to realists as conflationism: any attempt to accord explanatory primacy in every case to one particular type of force is bound to fail. Having said that, I agree with Schatzki that practices are enormously important. Every aspect of social life is enacted through practices. But there are many reasons to say that this cannot be the only focus of sociological work.

One reason is related to the point that every case in which a practice is implemented is different from every other. My action of writing this blog post has some similarities to other cases of blogging but it is also unique. No-one else has blogged while sitting in my office wearing the clothes I’m wearing now, for example. However, sitting at one’s desk to work using a computer is itself a practice and wearing conventionally acceptable clothes is another, so a practice theorist might say that this is merely an intersection of multiple practices. That does not solve the problem for practice theory, though, because we need an explanation for the variability and intersectionality in the implementation of practices and these cannot be explained in terms of practices alone, which are by definition similarities between actions. How and why did I choose to blog today, in my office, wearing these clothes, when I could have done something else instead? To add to the problem, there are always aspects of actions that aren’t practices at all: there is no generalisable social practice of ‘writing a blog entry that assess the relation of critical realism to practice theory from Dave Elder-Vass’s perspective’. The uniqueness of particular acts, in other words, cannot be explained purely in terms of practices: we need to recognise individual human agency and the choices that people make about how to implement practices.

Equally, we cannot treat practices as unexplained explainers. Practices themselves are caused, and not just by other practices. Elsewhere I’ve developed detailed explanations of how practices are caused, which attribute the primary causal power involved to social entities (or social structures) I call norm circles (see my 2010 and 2012 books for details). Whether or not they accept my particular explanation, realists would argue that practices are produced by an iterative process in which people act under the influence of social structures and/or cultures (Archer’s morphogenetic cycle). Individuals do not just spontaneously act in similar ways to other individuals, but follow practices because they have been influenced by social groups to do so. I would want to add now that the structures that influence us to follow practices are not only social but also material: my practice of blogging is both enabled and constrained, for example, by the technical characteristics of the Blogger platform.

Practices, then, are important because all of our social activities implement practices, but to make sense of practices we must avoid reducing social theory to practice theory, and avoid reducing social ontology to a purely practice ontology. Practices are always implemented in unique ways by reflexive individuals, and always depend on a structural and material context that conditions us to follow them without forcing us to. But recognising all of this should not lead to us dropping the idea of practices. Rather, we can now see that the concept of practices is a useful one that is consistent with a realist social ontology: we can invoke practices as elements in sociological explanations without collapsing structure into agency – and we should!