Monday, 8 December 2014

Some notes on David Papineau’s Durham Emergence talk: Emergent Causation and the Philosophy of Mind

I went off to hear my old Masters degree supervisor David Papineau talking on an aspect of the emergence theme. The abstract ran:

Intuitively, causal relations might seem to depend on impacts between microscopic physical objects. However I shall show that this is a superficial illusion, and that causation is an essentially macroscopic phenomenon. On this basis, I shall draw various conclusions about the causal autonomy of the mental realm.

My notes run as follows:

“How does the mind affect the physical world. Interactionist dualism is an intuitive default view. The mind is separate but has the power to come done and produce physical effects. But it is no longer popular in or outside philosophy. Why not?

Because  it seems plausible that all the laws governing brain and bodily processes involve only physical properties. This threatens to make any separate mental realm merely epiphenomenal. There is no space left for the mind to make any difference. A causal dangler. It is merely an illusion that the mental makes a difference, like a child with a toy steering wheel in a car. Worse than unattractive: it is incoherent. Conscious pain, eg., is never the reason for expressing it by a movement of the lips.

To avoid it, with the rise of materialism, one can instead argue that mental events just are complexes of neural events. The problem of causation is now akin to how political parties ever make a difference given that it is groups of people who produce social effects.

But since it seems to be the only contemporary option, why is it such a modern theory? 100 years ago it would have been a minority view. Answer: the assumption that all the laws governing brain and bodily processes only physical properties is itself a modern view. 150 years ago 'nervous energy' was not merely a metaphor. The conservation of energy would have included sui generis chemical, life, and mental energies. A kind of Newtonian interactionist dualism. But evidence has accumulated that there are no special forces operating within the body that do not also operate outside. Eg familiar electrical processes governing cellular processes.

Libet's work might suggest epiphenomenalism. But in other cases mental choices do have bodily effects. Libet's set up is not a paradigm of choice.

But are mental events just arrangements of neurons? One reason: the multiple realisation of mental states construed as functional, rather than type type physical states. But if so, the problem of epiphenomenalism returns since it is surely the neurons / hardware that moves the body around: epiphobia: the fear of functionalists who fear they are turning into epiphenomenalism. Much materialist literature aimed at resolving this problem. Striking that few people have attempted to explore gaps in the fabric of physical law. (Lowe is an exception.)

[It may be worth adding here that the fear of epiphenomenalism in the modern era is not that particular mental states or events are the epiphenomenal effects of neuronal causation. Since mental states are supposed just to be neuronal states, that cannot be the case. There can be no synchronic causation between what seem to be two states but which are really just the very same state. The worry is that the mental types or properties that particular token mental states possess play no causal role.]

A different route is to reexamine the nature of causation. Most philosophers think of causation as a series of pushes and pulls. But basic physics is symmetric in time whereas causation is not. This suggests that causation is a high level phenomenon akin to entropy. Increase in entropy cannot be explained in terms of basic physics because basic physics is symmetric in time. To explain the increase in entropy requires an addition to physics: two additions.

An analogy with the mixing up that a mix of hotter and colder gases will typically undergo. To explain the general pattern that such systems get mixed up requires throwing away information about each particular system.

On an analogous account, causes are macroscopic states that can be realised in different ways at the level of basic physics and their causal powers hinge on probabilistic facts about the  distribution of their realisers. At the level of basic physics, one loses the probabilistic structures on which causation (and entropy increase ) depend. Although this may not be how we intuitively think of causation that doesn't mean it isn't true.

Causation requires variable realisers.

So sometimes physical effects have mental causes and not specific neuronal causes. I waived for the taxi because I wanted to stop it , not because of specific neuronal arrangements.

Cf Stephen Yablo. The bull is angered by the cape's redness not its more specific crimson ness. Yablo turns out to be talking about real causation and not just explanatory salience. (The causal oomph need not be at the most basic level.)

Some effects result from specific neuronal arrangements such as the specific wys in which I waggle my arm. But in many cases mental states will be causes in their own right.

Once we realise that causation as emergent phenomenon relative to basic physics then we can understand how there can be autonomous mental causation.”

A couple of thoughts after the fact.

First, David Papineau was my first philosophy lecturer and so it is perhaps no surprise that much of the presentation and stage setting seems very familiar: the spatial metaphors for synchronic versus diachronic etc. I found that rather comforting.

Second, there is something admirable about the strategy of the talk. Causation at the microphysical level causes worry for functionalist physicalists about the merely epiphenomenal status of the mental (or mental properties). So the purpose of the talk was to destroy the very idea of causation at the microphysical level. Without it, there is no worrying comparison.

Still I have three immediate qualms.

First, it seems a high price to pay. So there is no causation at any level which does not admit multiple realization. I’m guessing at the physics here but that suggests that no Higgs Boson collision can cause anything. But we might think that it raises the probability of particular events, or is connected in some lawlike relations, or if the collision had not happened then neither would the subsequent event, or, by intervening on the Higgs Boson, one might be able to bring about some other desired event. Each of these is the core of a philosophical account or theory of causation. Papineau’s comment to defuse this seemed merely to be: although this may not be how we intuitively think of causation, that doesn't mean it isn't true. But it would be helpful to have some further argument for thinking it better to preserve causation only at the higher level.

Second, it seems to me that for a mental level causal link there might be a neuronal underpinning and that neurones are surely functional features thus themselves susceptible to multiple realisation. But if so causation continues at that level. And isn’t that how the worry about epiphenomenalism is usually put?

Third, I am not sure that the worry about epiphenomenalism need be couched in causal terms if causal terms are so understood. Suppose that the microphysical level cannot sustain causal relations. Still it might sustain sufficient conditions cashed out in nomological terms. (Some people might think this a matter of causation.) Suppose that that is the case. Would the nomological sufficiency of a microphysical redescription of a mental cause not itself be enough to raise the threat of epiphenomenalism?