In this issue and elsewhere, Hedda Hassel Mørch defends the view, contra premise 3 of the above argument, that conscious states do have certain dispositions essentially and that this fact is introspectively apparent.Footnote 3 This is the specific form of power theory known as ‘the phenomenal powers view.’ It is upon reflection apparent to us, Mørch claims, that there is a necessary causal connection between pleasure/pain and the motivational states associated with them:
Pain and pleasure appear to cause their effects in a distinctively dispositional or powerful way. Intuitively, pain seems to make subjects who experience it try to avoid it, whereas pleasure seems to make them try to pursue it. It is also difficult to conceive of pain and pleasure producing different, and especially opposite effects, i.e., pain making someone try to pursue it, or pleasure making someone try to avoid it. Note that the immediate effects of pain and pleasure are mere volitions rather than successful actions, i.e., mere tryings or attempts at avoidance and pursuit respectively (understood as mental events). Therefore, even if it is fully conceivable that volitions come apart from their effects (i.e., the actions they are aimed at), it may still be inconceivable that motivational properties such as pain and pleasure come apart from their effects (i.e. the mere volitions they motivate).Footnote 4
To be clear, Mørch does not hold that, as a matter of necessity, pain always leads to trying to avoid; obviously people very often endure pain for various reasons. Mørch claims, quite plausibly, that the necessary connection between a disposition and its manifestation is ceteris paribus: pain causes its bearer to try to avoid it in the absence of interference.
There is certainly something strange with the idea that pain should (ceteris paribus) cause its bearer to pursue it (or pleasure cause its bearer to avoid it), and this is something I think the opponent of the phenomenal powers view is obliged to explain (I will endeavor to do this below). But there doesn’t seem anything incoherent in the idea. It seems at least coherently conceivable that there could be certain odd creatures whose pain produces attraction compulsions and whose pleasure produces avoidance compulsions. Indeed, Mørch is willing to concedes this to an extent: she accepts that this seems to be conceivable in so far as we are assuming a Humean or a strong laws view. In the light of this, Mørch proposes that the necessary connection between pain and avoidance is conditional: If pain causes something in virtue of its intrinsic character, then pain causes trying to avoid. It is conceivable in a Humean or strong law world that pain generally causes its bearer to try to pursue it, but it is not conceivable that pain causes its bearer to try to pursue it in virtue of its intrinsic nature.
This conditional power of pain does not, in itself, avoid the threat of epiphenomenalism. We merely have a guarantee that if pain causes something in virtue of its intrinsic nature, then it causes trying to avoid. In order to make sense of pain actually causing something, we need to give an account of what satisfies the antecedent of the conditional, of what makes it the case that pain actually does cause something in virtue of its intrinsic nature. Mørch suggests two possibilities. Firstly, it could be that the possibility of pain not causing anything is only prima facie and not ideally conceivable, that is to say, an ideal reasoner would not find it conceivable that pain not cause the will to avoid. This response is in tension with a Revelation-based argument against physicalism, which I have argued is the core of the consciousness-based challenge to physicalism. The anti-physicalist contends that, given Revelation, if pain had a physical nature, then we’d know about it. But if the anti-physicalist can say that an aspect of pain’s essentially causal nature has not yet been discerned by us even though it is available to ideal reasoners, then why cannot the physicalist make the analogous move by holding that the physical nature of pain is available to ideal reasoners despite the fact that we have not yet discerned it?
Mørch is aware of this worry, and responds to it by saying, ‘…it is coherent to suppose that we are…able to reason ideally about those aspects of phenomenal properties that rule out that they are identifiable with physical properties, but at the same time not…able to reason ideally about those aspects that imply that they are powerful.’Footnote 5 Fair enough, but the onus is surely on the anti-physicalist to say what the difference is between the two cases and it’s not clear what could be said in this regard. In neither case are we dealing with some complex mathematics or conceptual subtlety that might foil the non-ideal reasoner.
Mørch’s alternative proposal is that there might be non-phenomenal features of pain that, as it were, activate its causal capacity. This seems to be a version of the ‘consciousness+’ hypothesis I defend in Goff (2017): the view that a conscious state such as pain is an aspect of a more expansive property—pain+—such that plain+ has both phenomenal and non-phenomenal aspects. In Mørch’s view, it is the non-phenomenal aspects of pain+ that activate the causal powers of pain.
The specific version of the consciousness+ view I formulate in Goff (2017) is a bit different. I defend the view that there are no necessary causal connections between pain itself—considered in isolation from any other properties—and its causal effects, but nonetheless that the whole pain+ property, which subsumes pain as an aspect, is necessarily causally connected to its effects. As it were, when God looks down at pain+, it is inconceivable that pain+ not be disposed to produce certain outcomes, including presumably trying to avoid. There is a temptation to interpret this as the view that the phenomenal aspect of pain+ is epiphenomenal, that it is really the non-phenomenal aspects that are doing all the work, dragging the phenomenal feel of pain along for the ride. But it is coherent to suppose that the phenomenal feel of pain does make a causal contribution, but only as part of the whole consciousness+ property.
On this view, pain, considered in itself, is not directly connected to trying to avoid; but there is an indirect necessary connection between these two things. Pain is essentially disposed such that, in combination with the non-phenomenal aspects of pain+, it causes trying to avoid. But a reasonable construal of Revelation is consistent with this disposition of pain not being introspectively accessible. According to Luke Roelofs’ definition, for example, revelatory knowledge of a property X yields knowledge only of those of X’s essential properties which don’t also involve another property Y, where Y is not itself grasped.Footnote 6 If the causal capacities of pain+ essentially involve both its phenomenal and non-phenomenal aspects, then there is no reason to think that mere knowledge of its phenomenal aspects should allow one to deduce the potency of the combination.
Hence, the proponent of the consciousness+ hypothesis can respond to the above revelation argument for epiphenomenalism by denying premise 1:
and replacing it with Premise 1*:
Premise 1*—If Revelation is true and our conscious states, considered in isolation from all other states, have some dispositions essentially, then introspection would reveal that our conscious states have some dispositions essentially.
With the argument thus modified, we can infer:
Conclusion 1*—Therefore, it is not the case that that our conscious states, considered in isolation from all other properties, have some dispositions essentially.
However, we cannot infer epiphenomenalism from Conclusion 1*. So long as conscious states make a casual contribution, perhaps as part of consciousness+ properties, epiphenomenalism is avoided.
Whether one goes with Mørch’s version or my version of the consciousness+ hypothesis depends on whether you believe that the dispositional nature of pain is at all introspectively apparent. And whether you think this may depend on what seems to you conceivable. Contra Mørch, it seems to me perfectly conceivable not only that pain should cause subjects to try to pursue it, but that pain should do this in virtue of its intrinsic nature. When I reflect on how pain feels, I can find no necessary causal connection with any kind of behavioural effects.
Having said that, there does seem something very weird about the idea of pain causing attraction volition. People on my side of the debate are obliged to give some account of the ‘gut feeling’ of absurdity attached to the idea of pain being separated from trying to avoid (or pleasure being separated from trying to pursue). My view is that there is a necessary connection between pain (considered in isolation from pain+) and trying to avoid, but that it is a normative rather than a causal connection. It is rationally appropriate, all things being equal, for pained creatures to try to avoid their pain. The absurdity we feel in the idea of a creature whose pain causes attraction and whose pleasure causes avoidance is not the absurdity of incoherence but the absurdity of deep irrationality. We may also be implicitly assuming when we imagine such a scenario that the conscious subject in question is a rational agent, in which case they will be disposed, all things being equal, to avoid pain and pursue pleasure. But to suppose this is already to suppose the efficacy of the mental, which is the very thing we are here trying to explain.
One point in favour of this account of the situation is that the sense of absurdity we feel when imagining a pain/pleasure invert (i.e. an imaginary creature whose pleasure plays the causal role that pain plays in a normal human being, and whose pain plays the causal role that pleasure plays in a normal human being) is not present when we reflect on inversion cases involving conscious states that don’t have such strong normative implications, such as colour experiences. The sense of absurdity we feel in the case of the pleasure/pain invert is entirely lacking in reflecting on the case of a colour experience invert. The best explanation of this, I suggest, is that the sense of ill-fit in the case of pleasure/pain invert is rooted in the normative implications of the conscious states involved, which is why we don’t have the same sense of ill-fit when reflecting upon the case of the colour invert. Colour experiences presumably have no less causal power than pleasure/pain experiences, so if the ill-fit was rooted in the causal powers, then why is it not equally apparent in colour invert cases?
On this basis, I suggest that the anti-physicalist should respond to the Revelation-based threat of epiphenomenalism by accepting that our conscious states, conceived of in isolation, lack essential causal powers, whilst at the same time holding that our conscious states do make a causal contribution as part of our consciousness+ states. This is not an entirely satisfying view, as we seem to have no positive grip on the non-phenomenal aspects of our consciousness+ states. However, we have known since Hume that the ground of natural necessity is hidden from us. The benefit of being a proponent of Revelation is that at least one aspect of concrete reality is known to us: the reality of consciousness.