Friday, 15 August 2014

Meillassoux's Challenge to Critical Realist Natural Necessity

Among the younger philosophers who loosely became associated with the term "speculative realism" (although many disown the association) Quentin Meillassoux is the most impressive. Closely allied to the work of Alain Badiou, who is something of a father figure lurking in the background of the directions of travel of speculative realism, Meillassoux wrote an important book called "Apres la Finitude" in 2006 (English translation, "After Finitude", 2009).  It's a difficult short book which I've been poring over for some months now - and only slowing coming to an understanding of. Philosophy is like this! It is fascinating to me how intuition is often the only guide to persevering with text that on first appearance seems intractable. Where does the intuition come from? It's really quite mysterious.

The issue of the mystery of intuition isn't beyond the scope of Meillassoux's topic - although not central to it -  it is clearly important territory for Badiou - but I'll write about that some other time. Meillassoux takes on the issue of 'natural necessity' - the attitude to which is one of the principal areas of dispute in the philosophy of science - particularly in the philosophies of Hume and Kant. It is also the fundamental point upon which Critical Realist philosophy reacts to Hume and turns back towards Aristotle.

To briefly summarise the Critical Realist attitude, it is that Hume's scepticism about causality which demanded that causes must be human constructs in the light of regular successions of events can't be right. The reason why it can't be right is that causal laws discovered by science apply in situations which are beyond the confines of closed-system experiments, and so this suggests that those laws are not entirely human constructs, but real operational things - Bhaskar calls them mechanisms - actually at work in the world which have been discovered. Bhaskar uses a transcendental argument - rather like Kant's approach to subjectivity - to argue for an ontology of causal mechanisms: those mechanisms that must exist independently of human agency (intransitive mechanisms) and those mechanisms which must exist through human agency (transitive mechanisms). I think it's interesting to note that this transcendental move is rather similar to Kant's move, except to say that Kant transcendentalises the subject in the form of categories rather than transcendentalising the world.

Meillassoux's position runs counter to critical realism and is more closely associated with Hume's original insight that there can be no necessary laws of nature. However, Meillassoux, like Bhaskar, is aware of the problem of closed system experiments - which he refers to as the 'problem of ancestrality' (ancestrality refers to how we reason about the evidence we are surrounded by of the world before there were humans in it - fossil records, geology, etc) but takes this as a philosophical problem which demands a solution that is in keeping with the proposition that there are no necessary laws of nature.

Meillassoux considers the proposition that if there is no natural necessity, then the laws of nature could theoretically change all the time. How come they don't appear to? He starts by arguing that the Hume argument can be interpreted as a probabilistic argument: The determination of causes rests on a probabilistic judgement about probable events - so it is unlikely that light will suddenly turn round corners - which in turn, argues Meillassoux, rests on a presupposition that the totality of possibilities about the behaviour of light might be known, and its probable behaviour calculated. I recently argued at the conference on Ontology of Organisations that Hume's argument looked rather like a theory of 'redundant expectations' - which I guess is amounting the same kind of thing - so I'm with Meillassoux on this.

Meillassoux then argues against the possibility of grasping the totality of possibilities. In an argument drawing on Badiou's thinking about "multiples", and particularly Badiou's use of Zermelo-Fraenko set theory (which conceptualises the 'unencompassable pluralization of infinite quantities') and Cantor's diagonal argument about orders of infinity, Meillassoux presents the 'non-totalizability' of possibilities as a key foundation for thinking about causation. This is interesting because, from dealing with the same problem as Critical Realism, it articulates a solution which is effectively in the opposite direction to the natural necessity argument of Bhaskar. Meillassoux pinpoints a contradiction in Critical Realism's position (without identifying CR). He says:
"although we have not positively demonstrated that the possible is untotalizable, we have identified an alternative between two options - viz, the possible either does or does not constitute a totality - with regard to which we have every reason to opt for the second - every reason, since it is precisely the second option that allows us to follow what reason indicates - viz, that there is no necessity to physical laws - without wasting further energy trying to resolve the enigmas inherent in the first option. For whoever totalizes the possible legitimates the frequential [probabilistic] implication, and thereby the source of the belief in real necessity, the reason for which no-one will ever be able to understand - thus, whoever does so must maintain both that physical laws are necessary and that no one can know why it is these laws, rather than others which necessarily exist"

Meillassoux's criticism of those who uphold natural necessity is basically that that it leaves the door open to woolly metaphysical arm-waving about the nature of the world. In response to questions like "Where do we come from?", "Why do we exist?" he points out that philosophers will generally "find the easiest way to shrug their shoulders" and find a way of arguing that the question is somehow flawed. For critical realists, the answer to such questions has always been along the lines of 'social emancipation', and the dovetailing of political arguments with a materialist upholding of natural necessity is perhaps the most significant aspect of Bhaskar's work. Meillassoux (and Badiou) reach a similar point. However, in place of Bhaskar's somewhat dogmatic assertions of "the real", "properties", "tendencies", etc (often using the language of Marxism), Meillassoux (and Badiou) use apriori mathematical reasoning. Bhaskar might accuse them of "ontological monovalence" (at least Alan Norrie has done this in his "Dialectic and Difference") - but I think this criticism doesn't really stand given the richness and depth of the work: indeed, Meillassoux and Badiou are fundamentally polyvalent in their thinking, and themselves complain (effectively) of what is effectively "monovalence" in the work of Deleuze and Foucault.

The tendency towards dogmatism in Critical Realism is its real weakness. It exists partly because of a failure for it to effectively address theory-practice gaps. Despite inspiring better methodological approaches in the social sciences (like Realistic Evaluation), there is still something missing. Of course, Realistic Evaluation is more sensible than Grounded theory (which is remarkably ungrounded in my opinion!), but it does not seem  to have ushered in a more naturalistic social science, nor does it show any signs that it might. The lack of a naturalistic social science from critical realism is due, I think, to a fundamental mystification that surrounds the description of natural necessity in CR. If we can deal with that mystification - as Meillassoux attempts - then coordinated approaches to closing theory-practice gaps become something perhaps a bit more realistic. 


Dai Griffiths said...

Thanks, very interesting.

It's not obvious to me (on first and ill-informed reading) that explanations based on probability are incompatible with natural laws. Can't laws be seen as descriptions of probability? The law of Brownian motion provides a formula which makes it possible to make statistical predictions about probability. Quantum physics is famously about probability. Is this fundamentally different from what Meillassoux is asking for?

Does it resolve Meillassoux's critique if we shift our position from claiming that we have found a law which is a discoverable and discrete entity in a realm beyond the human, and rather say that we have posited a mechanism which has predictive power in explaining what we observe?

In other words, is it the idealisation of the laws which is the problem, so that the laws are seen to cause a phenomena rather than to describe it?

Mark Johnson said...

Hi Dai,

Meillassoux's argument is that Hume's theory is implicitly probabilistic. I agree. Hume is actually very close to the idea of positing laws with explanatory power: he rejects natural necessity. Natural necessity is Kant's baby really (which is why Meillassoux labels him as 'correlationist').

There is a fundamental question about probability: the quantification of totalities. Following Cantor, we know this is impossible - except to say that totalities exhibit an order (aleph null, etc). So how can we have probabilistic reasoning about regularities if we can't know totalities? But if we can't have this, and yet the world doesn't continually change around us (there appear to be regularities), and we reject natural necessity, what on earth is going on? Deep down, the problem is in the maths.

It's really very good!