Saturday, 26 September 2015

Explaining Explaining... and Knowledge: Reflections on Chris Smith's Realist Personalism

Between realists and constructivists there are competing explanations of the world. The nature of explanation itself remains, however, unexplored. In his book "What is a Person", Chris Smith targets the kind of explanation which appears in what he calls 'variables social science'. He says:
"variables social science typically breaks down the complex reality of human social life into 'independent' and 'dependent variables,' whose answer categories are assigned numeric values representing some apparent variation in the world. Dependent and independent variables are then mathematically correlated, usually "net of" the possible effects of other variables, in order to establish independent statistical associations between them. Important independent variables are identified as related to the dependent variable through calculations of statistical significance, and statistical models are produced purporting to represent how certain social processes operate and produce the observed human social world"
Smith's fundamental objection might be succinctly summarised as the problem of the "mereological fallacy", about which Rom HarrĂ© has been arguing with Peter Hacker recently in the journal Philosophy (see Mereological fallacies are explanations of whole phenomena in expressed in terms of parts. Smith's concern in his book is the person, and he is right to point out that explanations of the person gets lost among the variables. He says:
"Persons hardly seem to exist in variables social science - they are rarely actually studied. What are studied instead are variables - which, when it comes to humans, are usually only single aspects or dimensions of persons or human social arrangements."
But then we come to realist explanations. Is this any better? From his Critical Realist standpoint, Smith argues that realism entails a change in perspective on the variable, not a rejection of it. This means to see causation not as the result of event regularities, but as the "operation of often nonobservable yet real powers and mechanisms that naturally exist at different levels of reality and operate (or not) under certain conditions and in particular combinations to tend to produce characteristic results". Variables are not excluded from the investigation of conditions, but a realist perspective would:
"take seriously the fact that variables are not causal actors. Variables do not make things happen in the social world. Human persons do. Persons are shaped by the enabling and constraining influences of their social structures - even as social structures are always emergent entities produced by human activity. And variables can represent aspects of social structures and relevant features of human actors. But variables do not cause outcomes. Nor do variables lead to increases and decreases in the values of other variables. Real persons acting in particular contexts in the real world do." 
Fine. But we still end up with an explanation, and yet there is no consideration of what an explanation is. Moreover, Smith's realism seems concerned with proscribing certain 'erroneous' ways of thinking among social scientists. Without an understanding of explanation itself, it seems that this kind of proscription is a constraint on the imagination where the warrant for the proscription is itself an explanation (i.e. Bhaskar's ontology). This is ironic because the book is about the person, and persons and explanations seem to be deeply entwined. Smith says when criticising evolutionary explanations of persons: "The very nature of human personhood [... ]- its reflexivity, self-transcendence, moral commitments, causal agency, responsibility, and freedom - means that persons can live, move and have their being in ways and for reasons not strictly tied to an evolutionist's monotonic explanation of everything." And yet he implies there is an explanation somehow within the reflexivity, self-transcendence and so on. I suspect there are many. It might be a mistake to believe variables to be causal actors, but equally it may be energizing and generative of powerful new ideas, initiatives, interventions and so on - many of which we are grateful for. Po-faced critical realism seems to lack a sense of fun and freedom: the erroneous imagination can produce remarkable and good things.

Explanations entail some concept of causation: the difference between realists, constructivists and positivists lies in the conception of causation as either being real, inherent in nature and discoverable, or being social constructions in the light of event regularities. But explanation itself is the elephant in the room in the debate between realists and constructivists (a debate which seems to me to be a bit worn-out now anyway).

An explanation is a kind of social constraint. Father Christmas and gravitation are explanations (or as Bateson would put it, "explanatory principles"), and each of them serve to coordinate social behaviour. We might think of "gravity" as having depended on event regularities for its establishment, or think of Santa as manifesting through 'normative behaviour' - although in both cases it is more multi-layered: each case has aspects of constraint materially, psychologically, socially, biologically, politically and so on. Both Father Christmas and gravity, however, are ideas which generate new practices, innovations, new forms of social organisation and so on - whether they are 'right' or not. Many of the ideas that they generate are infeasible. But by the fact that the ideas are there, we gain knowledge about the world in discovering the difference between what we might imagine and what we can bring into reality. We might imagine a world where everyone believes in Santa; but we know there is an age that our children (and we) reach when we realise it's a kind of game.

The confusion is at the interface between explanation and knowledge. It is the fault of the education system for conflating knowledge with the production of explanations. But knowledge usually emerges at the point where we realise our explanations don't work.

No comments: