1. Why Think About Composite Subjectivity?

Combining Minds aims to persuade you that minds can indeed combine. But first, we need to work out what that claim even means and why you might think they can’t.

So here’s a little bit of intellectual autobiography on that topic. In the last few decades, philosophers interested in the hard problem of consciousness have been paying more attention to ‘panpsychism’, the surprising idea that consciousness is pervasive in nature (e.g. Nagel 1979, Chalmers 1996, Seager 2006, Strawson 2006, Goff 2017). A common objection is called the ‘combination problem’ – not the only objection, but the one that most directly strikes at panpsychism’s ambitions (e.g. Seager 1995, Carruthers & Schechter 2006, Goff 2006, Goff 2009, Coleman 2012, 2014, Mørch 2014, Chalmers 2017). Panpsychists say that to explain human consciousness, we should accept some form of consciousness as a basic ingredient of nature: the combination problem says that even if we do that, it won’t help. It won’t help because ‘minds don’t combine’: many simple conscious minds can’t add up to a single complex one, so consciousness inherent in all matter still won’t explain human consciousness.

When I came across this claim back in 2010, I was puzzled by it: ‘minds don’t combine.’ People seemed awfully sure of it, but I wasn’t even sure of what it meant. Was it really so obvious? So I started looking outside the debate over panpsychism, to see what philosophers thought about this idea of minds ‘not combining’. Eventually I became convined of the following:

  1. There is indeed a widespread intuitive idea that something about minds stops them entering into compositional relations with each other;
  2. This idea manifests in several, otherwise unconnected, areas of philosophy – in thinking about collective intentionality, mental conflict, brain structure, the problem of the many, and other areas;
  3. This idea is rarely stated explicitly, but often assumed in passing;
  4. The reasons given for this idea, when reasons are given, are often not very convincing (at least, I thought so);
  5. This is partly because people often have very different ideas of what it would mean for minds to combine, and why they can’t.

I call this widespread intuitive idea ‘anti-combination’, and the opposing idea, that minds can combine, ‘combinationism’. More precisely (from p.6):

Anti-combination: The experiential properties of a conscious subject cannot be mere combinations of the experiential properties of other subjects which compose it.

Combination: One property-token P is a “combination” of other property-tokens p1 , p2 , p3 . . . , if and only if (i) P is fully grounded in the ps, and the real relations obtaining among them, and (ii) P is fully explained by the ps, and the relations obtaining among them.

Combinationism: The experiential properties of a conscious subject are sometimes mere combinations of the experiential properties of other subjects which compose it. (Such cases are what I call ‘composite subjectivity.’)

In Combining Minds, I tried to articulate as clearly as possible what combinationism would mean, what arguments there are for Anti-Combination, and how a combinationist might answer them.

I actually ended up sketching three distinct combinationist theories: ‘panpsychist combinationism’ (in chapters 3-4), ‘functionalist combinationism’ (in chapters 5-6), and ‘psychological combinationism’ (in chapters 7-8). That’s because one of the things I found was how much things depend on people’s different understandings there were of what ‘minds’ are, of what ‘combining’ is, and of what sort of mental combination is interesting or problematic. As a result, the arguments for and against mental combination can’t be satisfyingly resolved in a fully general way. Arguments that were powerful against combining one sort of ‘mind’ fell flat against another, and responses that worked for defending one sort of ‘combining’ missed the point for defending another.

For example, it matters whether consciousness is thought of as a fundamental property, or as something reductively explained by some kind of physical or functional structure. Similarly, it matters whether minds (or, as I’ll tend to say, ‘subjects of experience’)  are thought of as constructed out of psychological structures within consciousness, or as the underlying system or structure within which consciousness arises.

Attending to these differences made me think that I needed to consider at least three distinct contexts in which questions about composite subjectivity could arise:

  • There’s the above-mentioned debate about panpsychism, where the assumption is consciousness is a fundamental property, and the relevant comparison is with the way that fundamental physical properties work. I called the view I defend in this context ‘panpsychist combinationism’.
  • Then there are discussions about the structure of the human brain, its parts, and wholes that it might enter into. Here the assumption is that consciousness is tied specifically to systems with the right kind of structure. I called the view I defend in this context ‘functionalist combinationism’.
  • And then there are discussions about the dynamics of individual psyches, independent of their physical realisation, with a focus on their cleavages and conflicts, and how these relate to high-level phenomena like agency and self-awareness. I called the view I defend in this context ‘psychological combinationism’.

I’m going to talk a bit about each of these in the coming three posts. They’re meant to be separable, so that different readers might accept one while rejecting others. But they’re also compatible – in fact, personally, I think they’re all true, because even though they assume different conceptions of what minds are, those conceptions all pick out important aspects of reality.

Despite their differences, there are some general features that connect all three – things that I think are probably needed in any viable combinationist theory.

One concerns the ‘privacy’ of experiences, the idea that we have special access to our own experiences that nobody else can have. I think it’s important that combinationism can preserve this idea in some form, but it also has to be qualified. The standard reading of ‘nobody else’ is ‘no non-identical subject’: since my parts are not identical with me, that entails that my parts can’t have first-person access to any of my experiences, nor I to theirs. I think that makes it hard to maintain that my consciousness is fully explained by and grounded in theirs. But this result doesn’t follow if we read ‘nobody else’ as ‘no discrete subject’: then my experiences may be literally shared with my own parts, or with systems that contain me or overlap with me, but not with any separate person ‘outside me’.

Another shared feature is ‘between-subjects unity’: the relations that hold within a single subject’s mind (most notably the various things that have been discussed under the heading of ‘the unity of consciousness’) must be able to obtain between separate subjects (cf. Roelofs 2016). That’s because two parts of me can be separate subjects from each other, and me and you might be parts of a larger mind. How exactly we think about between-subjects unity will depend on how we think about the unity of consciousness in general, but any version of combinationism, I think, ought to say that the same relations can hold within and between subjects.

There are some other common features, which I discuss in chapter 2 of Combining Minds, but that’s enough for today. Tomorrow I’ll talk a bit about functionalist combinationism, then on Wednesday and Thursday I’ll cover psychological and panpsychist combinationism.

References:

Chalmers, D. (1996). The Conscious Mind: In Search of a Fundamental Theory. Oxford University Press.

Chalmers, D. (2017). “The Combination Problem for Panpsychism.” In Bruntrup, G. and Jaskolla, L. (eds.) Panpsychism: Contemporary Perspectives. Oxford University Press: 179-214.

Carruthers, P., and Schechter, E. (2006). “Can panpsychism bridge the explanatory gap?” Journal of Consciousness Studies 13 (10–11): 32–39.

Coleman, S. (2012). “Mental Chemistry: Combination for Panpsychists.” Dialectica 66 (1): 137–166.

Coleman, S. (2014). “The Real Combination Problem: Panpsychism, Micro-Subjects, and Emergence.” Erkenntnis 79 (1): 19–44.

Goff, P. (2006). “Experiences Don’t Sum.” Journal of Consciousness Studies 13 (10–11): 53–61.

Goff, P. (2009). “Can the panpsychist get round the combination problem?” In D. Skrbina (ed.) Mind that Abides: Panpsychism in the New Millennium. John Benjamins: 129-136.

Goff, P. (2017). Consciousness and Fundamental Reality. Oxford University Press.

Mørch, H. H. (2014). Panpsychism and Causation: A New Argument and a Solution to the Combination Problem. Doctoral Dissertation, University of Oslo.

Nagel, T. (1979). Panpsychism. In Mortal Questions, Cambridge University Press: 181-195.

Roelofs, L. (2016). “The Unity of Consciousness, Within and Between Subjects.” Philosophical Studies 173(12): 3199-3221.

Seager, W. (1995). “Consciousness, Information and Panpsychism.” Journal of Consciousness Studies 2-3: 272-88.

Seager, W. (2006). “The ‘Intrinsic Nature’ Argument for Panpsychism.” Journal of Consciousness Studies, 13 (10-11): 129-145.

Strawson, G. (2006). “Realistic Monism – Why Physicalism Entails Panpsychism.” Journal of Consciousness Studies 13 (10-11): 3-31.

2 Comments

  1. There was a common scifi trope of a computer getting more and more complex until it became self-aware. One of the better known examples would be AI from the Terminator movies. Weirdly, self-aware computers usually seem to decide to murder all humans. When I think about it, the trope is obviously based on the panpsychist fallacy. But then after a while we had supercomputers, especially massively parallel computers, and no sign of them being anything other than very fast at floating point operations. It became clear that complexity and or density of computing power does not lead to consciousness. Nor does a brain-like structure produce brain-like behaviour – so far we can’t even model a brain in a way that produces brainlike behaviour, though I believe we are close for C Elegans.

    The trope fell out of use in scifi because it was simply not believable anymore.

    • Luke Roelofs

      Thanks Jayarava, but could you clarify what you mean by “When I think about it, the trope is obviously based on the panpsychist fallacy”? I know the tropes you’re talking about, and you’re right that they haven’t really kept up with progress in AI. But I think I’m missing the point you’re making by bringing them up?

Comments are closed.