Tom Nagel on Panpsychism vs Neutral Monism

Last month I was involved in a fantastic conference: the 26th annual meeting of the Association for the Scientific Study of Consciousness. I think it went really well, and some parts got a fair amount of mainstream coverage, with varying degrees of accuracy. 

One highlight for me was a keynote talk by Tom Nagel, a philosopher who has had a huge influence on my thinking over the years. In this talk (video available here), building on themes of his work for the last few decades, he said a lot that I would enthusiastically agree with, but also drew some contrasts I don’t agree with between the view he’s attracted to (what he called neutral monism) and panpsychism. So I’m writing this post because, hopefully, unpacking what his distinction comes to, and why I think Nagel should really embrace the “panpsychist” label, might be illuminating for other people interested in exploring worldviews that are non-physicalist but still monistic.

So first, some things I would enthusiastically agree with:

• First, that consciousness is irreducible: no amount of behavioural or physiological facts can explain why there’s any subjective experience in any creature.
• Second, that all scientific progress points towards our being nothing but material structures, made up of the same stuff as everything around us, just more intricately arranged – there’s no evidence of anything strange and nonphysical in us, beyond our physical brains and bodies.
• Third, that the first of the above points is fatal to physicalism, and the second to dualism, thus pushing us towards some form of ‘psychophysical monism’: the world contains only one type of stuff, but rather than being purely and exclusively physical, that stuff has a fundamental aspect that relates somehow to the mental or subjective.

I think these three ideas define one of the basic orientations that you can have in contemporary philosophy of mind – not physicalist (because it can’t explain consciousness), not dualist (because the world seems too unified for that), but some sort of synthesis (compare Dave Chalmers’ Hegelian synthesis argument for panpsychism).

A few more of Nagel’s key claims also strike me as pretty plausible:

• Fourth, this sort of monism is really hard to think about, because “the conceptual scheme of our natural language is dualistic”. A satisfying theory that would clearly show mind and matter to be two sides of the same coin is thus more of a long-term aspiration or regulative ideal than something to demand immediately.
• Fifth, “the most promising place to start on this quest is with experiential affect, positive or negative”, as opposed to treating perceptual qualities as the central paradigm of consciousness.
• Sixth, a key challenge for the development of such a theory will be “a completely new theory of composition of mental or psychophysical parts and wholes.” (Hey! Hey! Psst! Over here!)

So as I said, I found a lot to agree with in this talk; what I was a bit less fond of was the following remark around the middle:

“This position is not equivalent to panpsychism: panpsychism is in effect dualism all the way down. This is monism all the way down.”

(Nagel didn’t say too much more about what this means and why he’s not a panpsychist, though there was perhaps some elaboration in an exchange with Andrew Lee during the Q&A, which I’ll discuss below.)

I have two things I want to say here, one fairly straightforward and one a bit deeper. First, this is a pretty unfair characterisation of panpsychism: although it does have versions that are, technically, dualist, the most widely defended forms are thoroughly monistic. Second, the very distinction between panpsychism, neutral monism, and other nearby views (including tongue-twisters like ‘panprotopsychism’ and ‘panexperientialism’) is actually pretty murky, and not something we can draw with much confidence at the outset of inquiry – which is where Nagel thinks we still are.

For readers who are scratching their heads at all these ‘isms’, let’s first step back. What is neutral monism? The standard definition is easy enough to give: it’s the view that neither mental nor physical properties are fundamental, because they’re both explained by some sort of ‘neutral’ properties that are the real bedrock of reality. What are ‘neutral’ properties? Good question! Neutral monists often struggle to characterise them in a way that is both informative and clearly non-mental, and Nagel in particular is explicit that he thinks we’ll need to go through some major conceptual transformations before we can really say what they are. So for now they’re sort of just a ‘secret third thing‘.

Ok, and what is panspychism? It’s usually defined as the idea that mental properties (usually phenomenal consciousness specifically) are everywhere – every last bit of matter has some sort of experience, even if it’s too faint and rudimentary for us to imagine. Our minds somehow emerge from or are built up out of this basic mentality that’s inherent in matter.

What’s worth noticing is that already, just looking at these two standard definitions, we can see two slightly different issues at stake – the issue of what explains what, and the issue of how widespread certain properties are. As a result, neutral monism and panpsychism aren’t actually incompatible alternatives, unless you build in some extra assumptions.

Neutral monism is defined by a claim about what explains what: neither the mental nor the physical explains the other, nor are they both explanatorily fundamental, but rather both are explained by the secret third thing. Neutral monists often pair this with the idea that mental properties (and perhaps also physical properties) are not everywhere, but only appear under specific circumstances (when the secret third thing is arranged just right), but that doesn’t automatically follow from the definition. So a neutral monist might also be a panpsychist – they might think that every single bit of matter has both a physical and a mental aspect, but that these two aspects are both explained by its secret-third-thing properties.

Panpsychism, by contrast, is defined by a claim about how widespread mental properties are, and thus is compatible with a range of views about explanatory priority. Technically you could be a physicalist panpsychist, holding that everything has a mind but its mind is explained by its physical properties, but such a position is rare: usually the motivation for thinking mind is inherent in matter is thinking that there’s no way to explain it purely in terms of non-mental stuff. Typically, panpsychists treat the mental as fundamental (see, e.g., here).

The easiest way to do that is to add it on to matter, alongside and in addition to matter’s physical properties, linked to them by a special sort of natural law (a ‘psychophysical law’, as opposed to the ‘physical laws’ of physics). This sort of view might naturally be termed “dualism all the way down”, but it’s not the only form of panpsychism available, nor the most popular. We just noted the possibility of physicalist panpsychism, and before that the possibility of neutral monist panpsychism, but the most widely defended form is what’s often called ‘Russellian’ panpsychism (one form of ‘Russellian monism’), inspired by certain insights about physics developed by Bertrand Russell (among others). The key Russellian insight is that what we think of as a thing’s ‘physical properties’ turn out, under scrutiny, to be just a very elaborate summary of its dispositions to interact causally with other things – saying that a particle has mass and charge is saying what it does, without saying anything about what it is. Russellian panpsychism conjectures that what it is is some sort of experience, or something otherwise mental. Rather than adding extra psychophysical laws, it says that the laws of physics are simply an abstract description of how simple experiences like this interact with each other. It’s not entirely clear how best to classify this view (people sometimes suggest it’s a sort of idealism, since ultimately all there is are mental things interacting; I have no objection to that construal), but it’s pretty clearly not any sort of dualism.

So I think Nagel is just wrong to call panpsychism “dualism all the way down”: although dualistic versions of panpsychism exist, the dominant Russellian form (the form I’ve defended) is impeccably monistic.

But so far this response to Nagel’s talk might seem a bit pedantic: ok, he oversimplified a rival theory, is that really the most interesting thing I can think of to say? Fortunately, no. Part of the reason for going over these definitional points is as set-up to say something a bit more interesting, namely the following: even though we define these options in terms of the ‘mental’ and ‘physical’, it’s pretty up in the air what those terms actually mean. It’s not just that we often use the words in sloppy or ambiguous ways (though we do), it’s moreover that when we’re trying to understand the fundamental level of reality, a lot of familiar constraints and parameters will no longer apply, and it’s not at all clear when something should still count as ‘physical’ or ‘mental’

Here’s a fairly simple example: the way we’re used to thinking of physical objects, they occupy a defined portion of 3-dimensional space. For some of the Early Moderns, this was the essence of being physical: physical objects might lack colour, odour, and taste, might even (according to Malebranche) lack real causal powers, but they had to take up some bounded volume of 3D space – otherwise you’re just not talking about physical stuff any more. That seems in retrospect to have been a mistake: in modern physics, it’s not clear that anything really has a bounded volume (particles seem like points, fields seem to extend indefinitely, etc.) – and on some theories space might have more than three dimensions. If we still believe that there’s a ‘physical’ world, we must be working with a different idea of what’s essential to being physical (though it’s not easy to say exactly what).

(A related question: above I briefly described Russellian panpsychism as a world made entirely of minds interacting with each other according to the laws of physics: does that still count as a physical world? If ‘physical’ means ‘things causally interact in a structure accurately described by the laws of physics’, then yes. If ‘physical’ means ‘the fundamental level is non-mental‘, then no.)

Let me bring this back to Nagel’s talk. The contrast between panpsychism and neutral monism is about whether the fundamental properties of the world are mental or not, but what would it mean for them to be mental? A view like Bishop Berkeley’s, where there’s just human minds, the mind of a personal, loving, God, and their perceptual images, would clearly qualify, because the fundamental things there are familiar minds like ours, or greater but still personal minds like God’s. But panpsychists don’t think that: they recognise that even if fundamental particles, or inanimate objects, or the indifferent cosmos we seem to inhabit, have some sort of mental aspect, it will be vastly different from a human mind, just as the physical aspects of these things are vastly different from a human brain. Saying that we’re interested in phenomenal consciousness specifically is helpful, but doesn’t resolve the question entirely. We can still ask: how much of what we associate with our conscious experiences can we strip away and still call what’s left ‘consciousness’?

This is partly a question for panpsychists: by what criterion do they call these fundamental properties ‘mental’ at all? But it’s equally a question for neutral monists: by what criterion do they call the secret third thing ‘non-mental’?

Nagel says some things in his talk that I think shed light on this. In characterising our common-sense conceptual scheme, he writes that:

“Mental phenomena appear within the point of view of an individual subject, and cannot be perceived from outside by another subject, whereas physical objects and physical phenomena don’t have a point of view.”

And later, in responding to a question from Andrew Lee about the possible role of introspection in helping us grasp the fundamental neutral properties, he says:

“The trouble is that I think introspection seems to show us a domain of subjective experience that is completely independent, at least metaphysically, from anything physical […] I mean, this was Descartes’ very persuasive thought that in introspection I have a complete grasp of what is going on in my mind, and I can see that it can’t include anything physical.”

If Nagel is saying that because introspection doesn’t reveal the whole story about the fundamental level, it can’t reveal any part of it, I think I just disagree with that inference. The rationale might be that introspection comes with this promise of revealing everything, and so if we reject that we should regard it as misleading all the way through. But there’s no need to be that drastic: both Andrew and I have published papers about how we can simultaneously take introspection to reveal something important about the nature of consciousness, while also thinking there are facets to our own experiences that we can’t introspectively discern.

Here’s a slightly different reading of what Nagel might be saying here: he’s identifying a certain structure as essential to our everyday sense of what ‘mental’ means. The structure in question is that of an individual with a single unified perspective, the contents of which are both perfectly known to that individual, and also completely inaccessible to all other individuals. Anything that doesn’t fit into that structure just doesn’t count as mental.

Here’s the thing: if we treat this structure, the omniscient perspective of the exclusive, isolated, integrated individual, as definitive of the ‘mental’, then I’d have to call myself a neutral monist. Because the view I’ve been defending for years, and referring to as ‘constitutive Russellian panpsychism’, systematically deconstructs the exclusive, isolated, integrated, individual, treating it as a superficial structure within a more fundamental world of overlapping, fluid, fuzzily-bounded consciousness. And I agree with Nagel that this requires a lot of conceptual re-thinking of our everyday assumptions about consciousness: for example, I’ve argued the following in various places:

• That we need to re-think the privacy of experience, and accept that multiple overlapping subjects might share their experiences (see this forthcoming paper with Philip Goff, and this related paper about overlap in everyday situations);
• That the most plausible version of the ‘revelation thesis’ (that having an experience shows you directly what that experience is like) is compatible with our experiences having a lot of inaccessible structure (see this paper again, as well as these two shorter ones);
• That the same ‘phenomenal unity’ or ‘co-consciousness’ relation that ties together all of your present experiences could in principle (and likely in actual fact) tie them together with other people’s experiences as well (see this paper specifically, but also pages 91-100 of Combining Minds);
• That one mind, unified the way we are, could be gradually changed into two minds, or vice versa, while remaining fully conscious of the process and without any sharp break or sudden discontinuity (see chapter 8 of Combining Minds as a whole);
• That our awareness of ourselves as singular is not due to our actually being singular, but due to the smooth co-operation of the many overlapping minds that compose us (see particularly pages 244-253 of Combining Minds, as well as this older paper).

In sum: my view is that the individual perspective is a construct, absent from the fundamental level, and so if one thinks (as Nagel seems to suggest) that it’s essential to something being mental, then it follows that on my view, the fundamental level is not mental, but rather proto-mental. It’s something like phenomenal qualities or phenomenal valences, phenomenally manifesting, entering into the sorts of relations that we’re acquainted with in our experience, but not organised into coherent and mutually-exclusive perspectives. Perspectival organisation depends on nervous systems and analogous structures of information-processing and coordination.

So you might ask: why call this panpsychism? Why not neutral monism? I guess I’d say: because this re-thinking we need to do, to understand the non-individualised basic level, feeds back onto how we should think about the everyday level of individual perspectives. If ‘the mind’ essentially means exclusive, omniscient, perfect unities, then there are no minds: there’s just this experiential flow temporarily appearing to itself, in various places, like that sort of ‘mind’. Our experiences belong not just to us but to our parts, and they are connected with events both inside and outside our brains in the same fundamental ways (though to different degrees). But I’d rather not say that there’s nothing really mental: our experiences are paradigmatic examples of mental things, so we should continue to count them as ‘mental’ even if they turn out to have a very different fundamental structure than we thought.

In general, I think it’s pretty common for different panpsychists theories to posit such different stuff at the fundamental level that the ‘fundamental mentality’ of one will count as ‘non-mental’ by the lights of another. A lot turns on which features of our experience are essential to mentality per se. For example, suppose the fundamental level is timeless, with time and space emerging out of it, as suggested by both Advaita Vedanta and some interpretatons of modern physics. Could it still be mental, a sort of ‘timeless awareness’ (as advocated by Miri Albahari here)? Or is temporality essential to mentality, such that a timeless fundamental level is a problem for panpsychism (a worry raised both by Susan Schneider and by Barry Dainton here)? Or consider the way that many, perhaps all, of our experiences involve an awareness of some sort of object, with some sort of quality. Which of these features – awareness, quality, the subject-object distinction – are essential to mentality, if any? This will determine whether Sam Coleman’s view that the fundamental level has phenomenal qualities, but no awareness of them, counts as panpsychism. Or looking a bit further back into history, three figures often viewed as panpsychists have radically different construals of the fundamental core of mentality that’s everywhere, and which other features arise from: Spinoza says it’s “ideas”, Schopenhauer says it’s “will”, Leibniz says it’s both “perceptions” and “appetite”.

The upshot is that it’s probably not very feasible to try to define panpsychism by first specifying some purportedly essential feature of the mental, and then asking whether the fundamental level has that. A better definition, I think, is that panpsychism says that the fundamental level has whatever ultimately turns out to be most basic to mentality as we know it, whatever aspect of our minds remains when we gradually strip away the features that can be gradually stripped away, whatever serves as the irreducible core that other features of our experience turn out to be reducible to. And the process of developing a panpsychist theory will include the process of figuring out what that is.

One advantage of this definition of panpsychism is that it centres its connection to the Hard Problem: lots of people can agree that there’s something about the mind that seems to resist explanation in any non-mental terms, though isolating and articulating its exact nature is harder. Any response to the Hard Problem which proceeds by attributing something to matter that’s supposed to somehow bridge that gap could be called ‘panpsychist’ in the most inclusive sense, because the debate over whether the added ‘something’ qualifies as mental or not is the sort of debate that panpsychists have with each other. And so by that definition, Nagel is a panpsychist.

But I don’t want to pick a fight over words. If we want to refer to a view where separate individual perspectives arise from an underlying level which lacks that structure but is otherwise phenomenal as “neutral monism”, then I’m a neutral monist.