In seeking to naturalize consciousness, the most conservative approach consistent with settled science is to assume its physicality: there is nothing we need to add to a physicalist ontology to eventually integrate our understanding of consciousness into the scientific image. Mainstream science has not yet strayed from the physicalism of accepted physics, in which what fundamentally, mind-independently, and uncontroversially exists are the fields and associated particles specified by quantum theory and general relativity. All higher-level phenomena, including biological beings like ourselves, partake of such constituents, and of nothing else. There are no categorically mental particulars that participate in the natural order, near as science can tell thus far. Why expect anything different as science proceeds?
One physicalist hypothesis about consciousness, panpsychism, says that conscious experience somehow inheres in, or is an aspect or property of, the most elementary constituents of matter. In addition to the confirmed properties of microphysical entities like electrons – their charge, spin, mass, etc. – is a phenomenal or proto-phenomenal property, call it microphenomenality, which somehow gets amalgamated to constitute one’s personal consciousness. The difficulty is that there’s no observational evidence to support this hypothesis, either for the existence of microphenomenality or for its assembling into something like our phenomenal experience. Instead, the evidence thus far strongly suggests that consciousness only arises for mind systems like ourselves, and perhaps far simpler creatures or the AIs to come, when certain sorts of sensory-perceptual and cognitive processes are up and running. Progress is being made in pinning down which neural subsystems are most correlated with reports of experience – the neural correlates of consciousness – and what behavior-guiding and system-maintaining functions those correlates perform. Conscious experience, along with other sorts of commonsensically mental phenomena (thoughts, beliefs, desires, hopes, etc.), thus seems to be a higher-level system property, not an attribute of the microphysical components of mind systems. That we don’t yet understand the how and why of consciousness as a system property isn’t justification for positing microphenomenality, at least not until some observational evidence for that posit comes along.
Another physicalist hypothesis is that consciousness emerges from a purely physical basis as a higher-level property or properties. Sean Carroll (2021a), an avowed physicalist, takes this line in his convincing critique of Philip Goff’s book on panpsychism, Galileo’s Error:
The basic idea is that of weak emergence. There exists a description of the system at a fine-grained level – in this case that of quantum field theory – that is complete and accurate on its own terms. And there is coarse-grained level of description, complete and more or less accurate on its own terms, which might involve entirely distinct ontological categories from those of the fine-grained theory. But the descriptions are compatible, in the sense that there is a well-defined map (typically many-to-one) from appropriate states in the fine-grained theory to those in the coarse-grained theory. In this picture, there is a set of processes undergone by certain states in the microphysical ontology, all of which correspond in the macroscopic description to “a person experiencing the redness of red.” (original emphasis)
Carroll intends that the ontological categories that arise in the emergence of consciousness are such that we need not add anything mental to our fundamental ontology, which is of course physical on his view. But what are these emergent categories? To establish that consciousness arises as a higher-level property or properties from suitably organized microphysical processes, we first have to characterize consciousness.
Carroll seems to be a realist (not an illusionist) about consciousness understood as phenomenal experience, or as he puts it, “consciousness as we experience it.”  Conscious experience ordinarily involves a host of qualitative (phenomenal) particulars, e.g., the redness of red, the painfulness of pain, often called qualia or the “what it’s likeness” of red, pain, cinnamon, etc. Such qualitative particulars are usually integrated into a constantly changing but unified phenomenal gestalt, such as what you’re experiencing right now. And indeed, Carroll mentions “qualitative experience” and “qualitative features,” that is, phenomenality, when adverting to consciousness. Moreover, he recognizes that consciousness is subjective; it involves. as he puts it, the “first-person experiences characteristic of human consciousness” (my emphasis). So as a first cut, and consistent with Carroll’s fairly standard conception of consciousness, we can take qualitativeness and subjectivity as two of its central defining properties (as does Goff in his book). The question then becomes how these properties emerge as distinct, higher-level phenomena from the workings of the brain; and more basically, we can ask if they emerge at all.
Examples of physical emergence are not hard to come by. My state-of-the-art non-stick fry pan has the wonderful property of being very easy to clean; properly cared for, virtually nothing sticks to it. Its non-stickiness isn’t possessed by its micro-constituents taken individually, but arises – emerges – as a function of their higher-level organization, the details of which I’m blissfully ignorant. But there’s presumably a transparent story to be told of how lower-level physical properties enable the higher-level physical property. As Carroll might have it, there’s a lower-level description in terms of chemical bonds and, consistent with that, a higher-level, coarse-grained mechanical description in terms of the surface’s resistance to food stuffs. It’s the story going from one level to the other that confirms the consistency; no unexplained transitions are allowed in deriving the higher-level property from the properties, behavior, and organization of lower-level constituents.
To make the case for the emergence of consciousness, we’d like a similar story, getting from the physical constituents of the brain and their organization to the experiential properties of qualitativeness and subjectivity, as for instance in my experience of pain. The problem, though, is that the subjective qualities of conscious experience are not observables in the way that the food-phobic surface of my pan is. The latter is open to inspection as an uncontroversially higher-level physical property, but my experiences of pain, among millions of other experiences I’ve had, are not. If they were, there would be no interesting puzzle about consciousness. So right away there’s a problem in telling the story of emergent experience: the end point of the account – consciousness – is simply missing as far as observational science is concerned. Goff’s Galileo’s Error notwithstanding, Galileo did not err in pointing out that phenomenal qualities are not to be found as objective existents in nature as observed by science. They only attend, only exist for, mind systems such as ourselves as the terms in which the world, including the body, appears to us in conscious experience (Clark, 2019). As much as I’d sometimes like to share my subjective, qualitative pain, I can’t.
Skirting this difficulty, instead of telling a story about emergent properties, Carroll tells us about emergent concepts, vocabularies, and descriptions (emphasis added in the following):
…there are multiple vocabularies we can use to describe the same underlying physical situation, each of which can stand on its own feet without reference to the others (we could profitably talk about conscious states before we knew anything about neurons or electrons), but which are compatible with each other in the sense that there are well-defined maps between them (even if the current state of human knowledge is unable to specify what those maps are).
Emergent concepts, even in this weakly-emergent sense, capture something true and real; they are neither illusory nor arbitrary... Conscious states, in particular, describe real phenomena, and play causal and predictive roles in the world. If we are told “she was conscious of being watched,” we immediately have somewhat reliable expectations for how she will behave in response to future events. Subsequent identification of corresponding microstates in a more fine-grained description doesn’t affect the reality of the emergent concepts in an appropriate domain of applicability. And the recognition of these causal powers doesn’t imply that emergent concepts need to be accounted for by enriching the underlying ontology; rather, it counts strongly against any purported explanation that separates the concepts from their causal role.
We do indeed have the concept “conscious states” and most of us, like Carroll, are realists (see note 3) about the referent of that concept, namely the qualitative episodes characteristic of consciousness. We also have concepts and vocabularies that describe the complex organization of the underlying physical constituents of persons hosting conscious experiences. But those seeking an explanation of consciousness would like a transparent explanation of how and why qualitativeness and subjectivity emerge as higher-level properties – not concepts – from their neural correlates, which again, are not themselves concepts but physical existents. The apparent non-objectivity of conscious experience, that it isn’t a public object open to observation, makes coming up with such an explanation problematic, and indeed no such account currently exists. That it doesn’t puts pressure (or should, I suggest) on the assumption that consciousness is straightforwardly physical.
Carroll’s approach to consciousness in what I’ve quoted above is an example of what, in his book The Big Picture and in a related Nautilus article, he calls poetic naturalism: that there are different “ways of talking” about a single, all-physical reality: there are different concepts, vocabularies, and descriptions that are each appropriate to the situation at hand, whether the situation involves fundamental particles, chemical reactions, biological organisms, or conscious experiences. The reality of the entities and properties referred to in these situations is certified by the utility of the descriptions in allowing us to predict and control the phenomena in question. In speaking of conscious experiences, we are, as in his example above, better able to predict the behavior of a person to whom we attribute a particular experience, and the utility of talking this way certifies the reality of conscious experiences, what we’re referring to by such talk. But on Carroll’s physicalism, the ultimate referent here is “the same underlying physical situation” (see first paragraph of the quote above) that could also be talked about in terms of physics, chemistry, biology, or neuroscience. To talk about experience is, under poetic naturalism, to talk about something that’s physical, not a distinctly mental property or substance.
Carroll says that the most fundamental, ontologically basic description of a situation, for example a person and her experience, is provided by the quantum mechanical “Core Theory”, which consists of “both the Standard Model of particle physics and the weak-field limit of general relativity.” It’s to the description at this level he wants to resist adding anything fundamentally mental:
Within its domain of applicability, the Core Theory is what we might label causally comprehensive. If we give a complete specification of the quantum state of the Core Theory fields within that regime, there is a specific equation that unambiguously predicts how it will evolve over time. This equation is sufficient to describe everything human beings generally do, unless they jump into a black hole or stick their hand inside the beam of a high-energy particle accelerator. There are no ambiguities or loose ends. The fact that brains are big, complex things is irrelevant. The Core Theory makes specific predictions for how any particular brain will behave; our choice is to either accept that prediction, or modify the theory in some way. (original emphasis)
Even if the Core Theory can, at least in principle, predict my brain and my behavior, Carroll’s poetic naturalism about conscious experiences is silent on how subjective qualitative episodes emerge as higher-level properties of mind systems, if indeed they emerge at all. In his Nautilus article, Carroll says:
There are certain processes that can transpire within the neurons and synapses of my brain, such that when they occur, I say, ‘I am experiencing redness.’ This is a useful thing to say, since it correlates in predictable ways with other features of the universe.
From this it doesn’t seem that any new distinct property emerges under poetic naturalism about consciousness, but only that certain physical processes get talked about differently. Nor would it seem that experiences themselves play distinct causal roles in guiding action, since again there is nothing distinctly experiential to play such a role, only a different way of talking. Of course, if we assume physicalism about consciousness, then the causal role of pain must boil down to the causal role of certain processes in the brain and body, whatever those might be. But should we assume physicalism? It’s obviously a convenient working hypothesis favored by perhaps most of those theorizing about consciousness, but we can’t take it as a foregone conclusion. We can assert that talk of experiences is, ultimately, just a useful way of talking about the underlying physical situation as described by the Core Theory, a claim which makes consciousness ontologically physical. But that assertion – poetic naturalism concerning conscious experience – has to be proven as the conclusion of a settled theory of consciousness, for instance by telling a coherent story about the emergence of subjective qualities from brain processes.
Since that story isn’t yet in hand, poetic naturalism, which amounts to the claim that physicalism is universal and thus encompasses consciousness, isn’t necessarily the case. Instead, we can safely say, and should say, that talk of conscious experiences is justified by the fact that we have them, that we really undergo qualitative episodes. Such realism is independent of, and prior to, the predictive utility of talk about consciousness, and it leaves open the question of its physicality. More generally, it leaves open the question of whether to naturalize a phenomenon is necessarily to physicalize it.
Physicalism is the default approach to naturalization since science proceeds empirically, taking its ontology from what’s intersubjectively available to observation and measurement, the hallmark of ascertainable physicality. This is good epistemic practice, but as we’ve seen, the categorical subjectivity of consciousness, its non-objectivity, makes it difficult to naturalize in this fashion. Hence its persistence as the target of the so-called “hard problem.” We can reiterate Carroll’s basic question:
Can we conceive of a world where the ontology consists of nothing other than some notion of physical “stuff” (or the specific quantum fields in the Core Theory) without any inherently mental aspects, but which nevertheless accounts for consciousness as we experience it?
The stuff of physicality is usually observed in 4-dimensional spacetime, or inferred to be there on the basis of observations interpreted within an accepted theory: the existence of invisible dark matter is inferred from the rotation of galaxies, the Higgs boson from secondary particle decay products of proton collisions. From a physicalist perspective, such stuff, however removed it might be from our everyday perceptual vantage point, is the essence of fundamental reality. As noted above, however, on the face of it there’s no observable stuff of consciousness, unless we assume physicalism, which I’ve argued we shouldn’t. What we can safely say is that what we see in brains are the physical correlates of consciousness, and consciousness might or might not turn out to be physically “stuffy,” as we might put it, depending on how the settled theory of consciousness pans out. I’ve suggested elsewhere that conscious experience might be perspicuously construed as a species of representational content, and content isn’t something we see (observe, measure) out in the world as represented in terms of that content. Whether and how content gets naturalized, perhaps as a representational, not straightforwardly physical, phenomenon, is a philo-scientific research project now underway that might bear on explaining consciousness.
But whatever the case about consciousness, an interesting twist is that what’s physical at the fundamental level may not be particularly stuffy as we commonsensically conceive of it – a conception, we might note, that originates in the everyday tactile conscious experience of the heft and resistance to deformation of human-scale physical objects extended in space. In exploring what he thinks might be the fundamental ontology of the physical, of what’s ultimately, mind-independently there, Carroll (2021b) ends up with a rather rarefied, even abstract entity:
…the fundamental ontology of the world is completely and exactly represented by a vector in an abstract Hilbert space, evolving in time according to unitary Schrodinger dynamics. Everything else, from particles and fields to space itself, is rightly thought of as emergent from that austere set of ingredients. (p. 2)
There is no further concrete physical specification of this ontology to be had, according to Carroll, nothing that would immediately connect it to “our direct experience of the world” (2). Nevertheless, he suggests that we can recover (derive) all the higher-level physical phenomena we observe in everyday life and in physics experiments as emergent from such an austere basis. How we precisely recover it, however, is the problem:
A vector in Hilbert space contains no direct specification of what the physical content of such a state is supposed to be; there is no mention of space, configuration space, particles, fields, or any such familiar notions. Presumably all that is going to somehow emerge from the dynamics, which seems like a tall order. (4)
Later, on the same theme:
The world of our experience doesn't seem like a vector in Hilbert space, evolving according to a list of energy eigenvalues. It seems like there is space, and objects located in space, and those objects interact with each other, and so forth. How in the world is all of that supposed to come from a description as abstract and featureless as a vector evolving through Hilbert space? (6)
And the question posed again:
The search for emergent levels is precisely the search for higher-level, non-fundamental descriptions that approximately capture some of the relevant dynamics, perhaps on the basis of incomplete information about the fundamental state. The question is whether we can recover the patterns and phenomena of our experience (space, objects, interactions) from the behavior of our fundamental ontology (6).
He goes on to answer this question in the affirmative, albeit tentatively and speculatively. The details needn’t concern us (how convenient!) since the point I want to raise is that the fundamental ontology in play here, like consciousness, is not an observable. We can’t see or measure or detect in 4-dimensional spacetime a vector evolving in Hilbert space since that space is what spacetime, fields, particles and other observables emerge from as higher-level phenomena, according to Carroll. The referent of the austere formalism of the Schrodinger equation and Hilbert space vector, what it represents to exist, isn’t physically specifiable – it’s abstract and featureless. This leaves open the question of what sort of physical stuff, if any, lies at the root of reality as described by science.
If, as Carroll seems to suggest, there’s no observable physical stuff the formalism of Hilbert space refers to, then we might conceive of its referent as proto-physical, that is, as the non-stuff out of which arises the measurable, stuffy phenomena of higher-level physics and everyday life. “Ways of talking” at various levels of description within science (and thus naturalism) – the concepts and vocabularies we use to refer to mind-independent phenomena – perhaps needn’t presume physicality at the most fundamental level, at least not as observable, measurable stuff.
To show that the “warm, welcoming, richly-structured ontology we are used to thinking about in physics” in fact emerges from Hilbert space, and how, would be, as Carroll concludes, “a triumph of unification and simplification.” Equally, to show how consciousness emerges from its correlates located in spacetime would constitute a theoretical triumph, most likely of neuroscience. But just as the possible emergence from non-observable non-stuff to observable stuff at the fundamental physical level needs to be transparently explained (no unexplained transformations), so too must the emergence of consciousness, but in the opposite direction: from the observable to the unobservable. To get from neurons to phenomenality, we’re moving from what’s quantifiable and objective to what’s qualitative and subjective, and thus far the story of physicalist emergence eludes us. We’re perhaps in better shape recovering higher-level physical, chemical, and biological phenomena from Hilbert space (Carroll’s remit) than we are in recovering phenomenal experience from the brain.
Whether consciousness as qualitative subjectivity can be derived as emergent from what the Core Theory says physically exists is an open question. In any case, since the commonsense physicality (observable stuffiness) of fundamental reality as described by physics is at least somewhat suspect, perhaps we should cut consciousness some slack in that regard. If, for example, content turns out to be the key to consciousness, and content is best thought of as fundamentally representational, not physical, then consciousness, albeit natural, might not be best conceived as physical. Physicalism might not be the last word, or the best way of talking, when it comes to naturalizing all reality (see note 5). Stay tuned.
Carroll, S. M. (2017) The Big Picture. Dutton: New York.
Carroll, S. M. (2017) Zombies must be dualists. Nautilus.
Carroll, S. M. (2021a) Consciousness and the laws of physics. Forthcoming, Journal of Consciousness Studies.
Carroll, S. M. (2021b) Reality as a vector in Hilbert space. Invited contribution to Quantum Mechanics and Fundamentality: Naturalizing Quantum Theory Between Scientific Realism and Ontological Indeterminacy; Valia Allori (ed.).
Clark, T.W. (2019) Locating consciousness: why experience can’t be objectified. Journal of Consciousness Studies, V 26, No.11-12, 2019.
Dennett, D. C. (2016) Illusionism as the obvious default theory of consciousness, Journal of Consciousness Studies, 23 (11-12), pp. 65-72
Frankish, K. (2016) Illusionism as a theory of consciousness, Journal of Consciousness Studies, 23 (11-12), pp. 11-39.
Goff, P. (2019) Galileo’s Error: Foundations for a New Science of Consciousness. Random House: New York.
Seth, A. K. (2021) The real problem(s) with panpsychism. Forthcoming, Journal of Consciousness Studies.
 Carroll contrasts weak emergence with “strong emergence – the idea that legitimately new behaviors arise in collective phenomena that cannot be derived in terms of the individual behaviors of constituent parts of the system.” He dismisses strong emergence since it would involve “extravagant deviations from contemporary physics” for which we have no evidence.
 Of course, when we think of the causal role of pain, we’re thinking of the causal role of the felt quality of pain, not its neural instantiation. On any sort of physicalist identity claim about consciousness, it’s hard to find a role for qualities per se given that the physical instantiation is already doing the causal work.
 What then should be our basic criterion of the physical, if not stuff? Perhaps it’s just that which we suppose exists mind-independently on its own, and thus has a self-nature independent of how we represent it. This permits counting as physical even the most abstract, non-spatial sort of phenomena, such as Hilbert space, so long as the claim is based on observation and explanatory and predictive connections to entities in an accepted theory. It’s the requirement to have a mind-independent self-nature that most obviously disqualifies consciousness, as well as things like concepts, propositions, and numbers, from being physical. These are all essentially representational phenomena deployed by mind systems in representing the world, and they aren’t themselves intersubjectively available to observation. But since mind systems are physically instantiated, and thus available to observation, the naturalization of representational phenomena, including consciousness, might be achieved by a mature, observationally based theory of representation. Looking at a system, we could unequivocally deduce from the theory whether or not representational content exists for it, and whether such content is qualitative, thus conscious.