Panpsychism

Panpsychism

The subject of Panpsychism has been making the rounds recently. In particular as a result of what seems to be a publicity circuit for a new book by philosopher Philip Goff. Philip Goff is a philosopher of consciousness from Durham University and his book presents an argument for panpsychism. Panpsychism is a philosophy that posits that consciousness is inherent to all matter, as opposed to it being an emergent phenomenon arising in only some objects like brains.

I wanted to talk about this topic for a number of reasons. First, I have always enjoyed the philosophy of mind as a subject ¹. Second, this particular topic has continued to appear on my radar over the past year. In chronological order: Sabine Hossenfelder wrote a blog post on the subject; Sean Caroll had Philip Goff as a guest on the mindscape podcast; Ezra Klein spoke briefly and excitedly about the idea recently in his discussion with Barbara Ehrenreich on his podcast; and then not too long ago Sabine Hossenfelder’s blog post was reprinted as a Nautilus article. These are just the places where it showed up for me, a quick google search reveals others. My final reason for writing about this is that I was once inclined towards believing Panpsychism, but ended up deciding it was not correct. What convinced me was a line of reasoning that I haven’t yet seen clearly argued in the popular media where it has been discussed. To his credit, I think Sean Caroll does come close in his podcast, but the discussion does not really get into the details of it. Also, Dennett very clearly presents a similar critique during a talk Goff gave in 2014. But this is a bit obsure. So since panpsychism seems to be on people’s minds, I wanted to share my thoughts, and the line of reasoning that turned me off of the idea.

So first I’d like to say more about what panpsychism is, and in particular try and present the argument that Goff is making. I should say I have not read his book. I’m not sure how much that disqualifies me from trying to replicate his arguments, but I’ve resolved that in order to keep a blog you can’t be paralysed by the fear of not getting things exactly right. I have read one of his more comprehensive papers on the subject, as well as the encyclopedia entry he had a hand in writing, the talk I mentioned earlier, and I’ve read and listened to some of his responses to criticisms. I will do my best to give him a fair shake here.

Panpsychism aims to explain why we experience consciousness. It does this by claiming that consciousness is a fundamental constituent of reality, in a similar sense to how one might say that an electron is a fudamental constituent of reality. Thus the question of why we experience consciousness admits an explanation via a reduction to these fundamental constituents, and not as an emergent phenomenon appearing in our specific biology. A consequence of this line of reasoning is that we should then expect consciousness to be ubiquitous, and not specific to brains. There are many variants of this idea, and it is not clear that Goff subscribes to any one in particular, although if I were to venture a guess I would say he could be a panexperientalist constitutive micropsychist. So instead of focusing on a particular variant, I am going to focus on an argument he seems to stress frequently, which isn’t specific to a particular variant. Goff argues that there are two aspects to things in the world, that which they do, and then that which they are. Thus, a scientist will tell you what a thing does, but not what a thing is. This is in contrast to something Goff and others calls dispositional essentialism, which is the idea that nature is as nature does. To quote Goff:

Things on this view are not so much beings as doings: if you understand what an electron does you know everything there is to know about its nature.

Goff rejects dispositional essentialism, based I think largely on arguments coming from Bertrand Russel. Essentially the objection to dispositional essentialism is that it leads to a circular regress of definitions. If all things are defined in terms of their dispositions towards other things, then there is no foundation on which these definitions rest, and this is not a bullet Goff wants to bite. Having rejected dispositional essentialism, one is left asking “what is the natural world.” Goff’s take is that panpsychism answers this question by describing what things are. Matter, and other natural phenomenon are consciousness, and science gives an account of what that consciousness does in the world. This strategy allows this variant of panpsychism to avoid the problems plagued by dualistic theories of consciousness, in that it is built in order to avoid contradiction with scientific theories. Any prediction science makes should be accomodated by this view, because science makes no claim about what matter is, only what it does.

Many physicists instinctively rankle at this idea, because I think many of them are dispositional essentialists (whether they know it or not). Take for example Hossenfelder’s post. Her argument is that electrons for example can not be conscious, because if they were this would require additional degrees of freedom, and these additional degrees of freedom would be observable in the behaviour of the electrons. I do not think this alone actually engages with the premise being put forward by Goff, since it doesn’t address this is/does distinction. If you examine the comments of the post, you can see Goff make this point under the username “artuncut”:

physics tells us what spin does not what it is. You haven’t given any provided any argument contrary to that claim.

And Hossenfelder’s response:

Science is about describing observations. In science we construct models that we use to make predictions. A spin is a property of a wavefunction, which is a mathematical object in our best current model. We know exactly what spin is in this model. It has a definition that you can look up in any textbook.

If you are asking anything besides that, you are asking a question that’s unanswerable. Of course you cannot answer this question either. Saying that spin is “consciousness” does not explain anything. I could equally well say “spin is tokohila” which explains equally little.

Here we see what I think is the real crux of their dispute, which is that Hossenfelder sees the distinction Goff is trying to make as nonsensical. I wont try and get into the details of their exchange. This is meant merely to illustrate that to properly critique Goffs version of panpsychism, one needs to engage with this is/does distinction that he makes. This is what I want to try and critique here.

As I said earlier, I was once sympathetic to panpsychism. So why did I change my mind? A favourite rhetorical technique of philosophers of consciousness, especially if they want to emphasize that it is a hard problem – and one potentially outside the scope of science – is to insist that they are intimately familiar with their consciousness. They really see red in an intimate and visceral way that goes beyond any dry scientific account. To quote Goff on the Mindscape podcast:

what grounds my knowledge of consciousness is my immediate awareness of my own pain

This contention seems to function as a kind of foundational principal on which many philosophers of consciousness motivate the rest of their arguments. Indeed, if we were not convinced of our own experience of consciousness, then what would it be that needs explaining? In fact I think that this common contention, that we know that we are conscious, is precisely what undermines Goff’s premise.

If you really know that you are conscious, then when you tell people that you are conscious you should also have good reason to insist that it is a true claim. The question then becomes, how do you connect the truth of your claim – that you are conscious – to your intimate experience of that consciousness? If you believe, as Goff argues, that consciousness is what matter is, and not what matter does, then you are going to run into a problem for the following reason. Consider that any instantiation of the statement “I am conscious” is physical matter behaving in a particular way. If you wish to make Goff’s move, to say that science is responsible for explaining what matter does, then you are forced to concede that the story of how this statement came into existence is a story about what matter does, and thus is a scientific story. That is unless what matter is has some causal influence on what it does, but this is precisely what Goff is trying to avoid. So what matter is can not be responsible for the existence of the statement “I am conscious”. Consequently, if consciousness is what matter is, then neither can consciousness be responsible for the existence of the statement. How then can you claim that the statement be true, if it were not at all informed by the fact that you are conscious?

I think this argument applies more generally in support of a kind of epistemic version of dispositional essentialism. For the purposes of claims of knowledge, one must be a dispositional essentialist, because knowledge is as matter does. Knowledge is physically instantiated as the behaviour of matter. At least the knowledge we share with one another. Even if matter has a more essential character, there does not seem to be a way to justify any claim to know it. So if you claim to know you are conscious, then this thing that you claim to know must be something nature is doing. I suspect this is the intuition that many scientists take when engaging with this topic, and I am curious to know what Goff’s take on this is.

I said earlier that Carroll had begun to broach this argument in his conversation with Goff, but I feel as though it got deflected or misunderstood. To quote Carroll:

[Caroll] So the zombie is David Chalmers’ idea of a physical thing that acts exactly like things in the real world, a physical collection of atoms and so forth, but that does not have consciousness. And my point is if it acts exactly the same, that includes when you ask it, “Do you have consciousness or are you experiencing things?” It says, “Yes.” And if you ask it, “Are you lying?” It says, “No.” So clearly it honestly thinks it’s conscious but it’s not by hypothesis. So by that therefore, none of us knows that we’re not a zombie ’cause zombies think that they’re conscious just like we do and therefore you’ve really added nothing.

This has a similar character to the argument I have presented. The problem with zombies is that they rely on the concievability of their existince. In fact I think Caroll actually meant this argument in a similar fashion to mine. But we can really drop the zombies, since they are a kind of red herring I think. In any case, Goff responds

[Goff] coming to your second argument as I’m understanding it, well, doesn’t this imply, none of us would know we were conscious. There are tricky issues here actually, just as a preliminary thought, about the relationship between thought and consciousness. So the dominant view in my philosophical tradition has been that thought has nothing to do with consciousness, and you can see this because the dominant theories of thought in the 20th century such as Jerry Fodor or Donald Davidson or Dennett, have absolutely nothing to say about consciousness, so they think you can give a complete account of thought without talking about consciousness. […] So on that view a zombie would have thoughts and would think it’s conscious, as you say. But I’m actually one of a growing minority of philosophers who think actually thought is a kind of consciousness, and if that’s true, then the zombie wouldn’t have any thoughts at all. So that’s one way you can go with this, but maybe that’s getting into slightly contentious territory, but the other thing to say is… Yeah, the zombie behaves in all the same ways but I don’t know about my consciousness, it’s not the case that I know about my consciousness by observing myself, I know about it because I’m immediately aware of my own experiences.

[Sean] But that’s just what a zombie would say.

[Goff] It would say that, but it’s not true of the zombie, by stipulation. It doesn’t have this… What justifies, what grounds my knowledge of consciousness is my immediate awareness of my own pain, and by stipulation, that’s what a zombie lacks.

We see here a return to this fundamental contention. Goff knows he is conscious. But this does not address the problem I think. The problem is how, in his conception of nature, can it be possible to claim to know such a thing.

I mentioned earlier that Dennett presented a similar argument here.

[Dennett] the trouble with your postulation of intrinsic properties is that, as a number of people have suggested, an intrinsic property seems to be systematically elusive. Nothing could be evidence for it because whatever evidence you find is just going to be another manifestation of another disposition, and if intrinsic properties are not discovered by what they do then they’re not discoverable at all. So you end up postulating an intrinsic property to conscious states, an intrinsic property to matter, and, talk about theft vs honest toil, the research programme is over! There is nothing more to ask, there’s no questions to ask. Nothing could count as an answer, because an evidence would have to count as an evidence of another dispositional property.

[Goff] I want to say two things. Firstly, whenever I talk about consciousness to Wittgensteinians they always say “but you’re left with skepticism about other minds.” As though a theory that ends up with having that problem must be wrong. And I say well I do! That’s a shame, I think that’s the human epistemic situation. I hear what you’re saying, physical science is great, so it should be able to cover everything. But maybe the worlds not perfect. It would be great if we could, but if we can’t that’s the human situation.

Dennett puts forward a broader form of the argument presented here. But because it’s so broad, I think Goff misses this important consequence. Goff’s own words count as “an evidence of another dispositional property”, and this is a problem not because this evidence fails to convince others of Goff’s consciousness, but because, by his own conception of reality, Goff himself must concede that his own words are completely unrelated to his own consciousness! In the video they go further into the weeds on this, and I think get closer to this point, but I wont go into it here.

I want to say that I really don’t mean to throw Goff under the bus here. Indeed I would not be surprised if he has considered arguments like this one, and may have good responses to them. Also it’s very likely I am not completely understanding his responses. If so I would be very eager to be educated on this point. In the end I primarily want to use this topic, and the argument presented here, as a jumping off point to a different subject. I am very interested in the structure of this argument. In particular I am interested in how it employs this fact that information is physical to make conclusions about what one could possibly claim to know. This is something I would like to explore more.