Never Mind: Straw Arguments Against Panpsychism

by Jochen Szangolies

The only thing worse than a good argument contrary to a conviction you hold is a bad argument in its favor. Overcoming a good argument can strengthen your position, while failing to may prompt you to reevaluate it. In either case, you’ve learned something—if perhaps at the expense of a cherished belief.

But with a bad argument, you’re put in an awkward situation. Since you agree with its conclusion, to the extent that you’re interested in spreading belief in this position, engaging it would not seem to be in your best interest. On the other hand, such arguments always carry the necessary raw materials for the construction of future strawmen within them: bundled together, they only need to then be knocked down with great aplomb to try and dissuade belief in the conclusion they purport to further. Hence, I call such arguments ‘straw arguments’.

Take the case of Ernst Haeckel’s supposed ‘biogenetic law’, which posits that the creation of the individual retraces the evolutionary history of the entire species—‘ontogeny recapitulates phylogeny’. Thus, that evolutionary history is manifest in the developmental stages of an individual history, giving direct evidence (of a sort) of this history. The only trouble is, it’s wrong; individual stages of embryonic development do not resemble adult stages of phylogenetic ancestors.

On its own, this would not be much of a problem: scientific theories are superseded as a matter of course, and erroneous ideas left in the dust of history. But Haeckel’s proposal has found a second life in creationist books and websites, where it is knocked over with verve to try and rebut the idea it was originally supposed to support, namely, evolution by natural selection. Those faced with this argument then enter the debate handicapped by the twofold burden of explaining that, while yes, Haeckel’s argument was mistaken, this does not challenge the scientific consensus regarding evolution. Inadvertently, Haeckel provided the straw for creationist strawmen.

Nothing nearly as consequential is at stake presently. My topic here is panpsychism, the idea that there is some form of consciousness inherent in all matter. While this seems a profoundly counterintuitive notion at first, it does have a long and distinguished pedigree in philosophy, going back all the way to Thales, the ‘first philosopher’ living around 600 BCE. Additionally, merely running counter to intuition is no great mark against an idea in a subject as challenging as the philosophy of mind, where following one’s intuition often swiftly leads to inconsistency. To refute it, more than mere appeal to intuition is needed.

Two attempts to supply something more have been put forward by philosopher and computer scientist Bernardo Kastrup and physicist Sabine Hossenfelder. As the preceding intimates, I don’t think that either argument succeeds; however, I think there’s something interesting to be learned from precisely how they fail, which somewhat ameliorates the awkwardness of defending a belief I don’t hold.

Both pieces make a claim that, on its face, is rather astonishing—or would be, if it were right: that not only is panpsychism wrong philosophically, but that it is, in fact, in conflict with established physical principles readily checked experimentally. Thus, their arguments would make panpsychism a metaphysical stance refuted empirically—certainly a rare beast!

Evaluating this claim is then my feeble justification for writing this piece that goes beyond someone merely being wrong on the internet. But first, let’s try and get clear on what is typically meant by panpsychism in a modern context.

What Is It Like To Be An Electron?

With panpsychism having been around for some time, it is probably no surprise that the idea has taken different forms and expressions over the years. However, a core insight of modern-day panpsychism is most often traced back to the mathematician and philosopher Bertrand Russell, and the physicist Arthur Eddington: namely, that physical theories are inherently structural or relational.

This view is most clearly epitomized by Eddington’s felicitous phrasing that science is silent on the ‘inner un-get-atable nature’ of matter. Thus, I take panpsychism to be essentially Eddington’s view when he writes (in Space, Time, and Gravitation):

Physics is the knowledge of structural form, and not knowledge of content. All through the physical world runs that unknown content, which must surely be the stuff of our consciousness.

That is, all that is visible of the material world to physics (and science more generally) is a system of relations: whenever we measure, we probe a certain relation between a system and a measurement device that provides us with a reference. Take a measurement of the length of a piece of string: what it establishes is a particular relationship between the piece of string and the yardstick. But such relationships fall short of getting at the ‘content’, the ‘inner nature’ of the string—the relata grounding the relations. (See here for some more discussion on that topic.)

Panpsychism is then basically the thesis that these relata are themselves essentially mental. Importantly, it does not say anything about the nature of this mental content, as such—but a reasonable and widespread assumption is that the simpler the entity, the simpler that mental content. Thus, we might suppose that the conscious experience of an electron, for example, is about as reduced as is imaginable—perhaps analogous to the unchanging experience of seeing a white, featureless wall, or perhaps whatever sort of experience goes along with merely existing.

This is grounded in the fact that our complex experience is rooted, ultimately, in the complexity of our brains—which furnishes our capacity for reflection, emotion, and integration of multimodal sensory experience into a unified perceptual field. Thus, any reduction in complexity on the material side seems to be accompanied with a reduction in complexity on the experiential side. To study panpsychism at its base, then, it makes sense to focus on the simplest, or basic, entities—whatever they may be: particles, strings, or something else altogether.

With the picture sketched so far, claims that one could rule out panpsychism by appeal to physical science should immediately seem dubious. After all, the mental properties of basic entities are precisely those that don’t figure in their scientific image. Nevertheless, it would be rash to dismiss such arguments out of hand.

Thought Versus Experience

Hossenfelder headlines her piece with a claim that’s true, but vacuous as regards the discussion at hand: electrons don’t think. Indeed they don’t: but the panpsychist makes no claim that they do. Mental content is not synonymous with thought; experience, specifically phenomenal experience, is quite different from the processes of reasoning usually called ‘thinking’. True, thought is accompanied by phenomenal experience—the experience you have of yourself as thinking certain thoughts—which indeed is not a kind of experience an electron would have. But that doesn’t entail that it can’t have any other experience.

Consider your experience of having a certain perception, say, of seeing your coffee mug on the table. What is it, ultimately, that underlies this experience? On anything but a starkly dualist conception, it is, in whatever way, the state of your visual apparatus: light hitting photoreceptors in your retina, causing photochemical changes in certain proteins, which triggers electrochemical signals traveling along your optical nerve, leading to a particular pattern of neuron firings in your occipital lobe.

This state can, in principle, be expressed in terms of a very (very, very…) long list of numbers: neuron spiking frequencies, molecular configurations, signal intensities, and so on. It is, thus, the stuff of physics: the structural, relational picture of the act of having a particular experience, with the experience itself being its content, its inner un-get-atable nature.

A conflation between the two is the origin of Hossenfelder’s main criticism. Essentially, she posits that, to account for conscious experience in basic entities such as electrons, the list of numbers describing their state would have to be extended by further numbers to account for their experiential properties; but this is at variance with experiment.

Concretely, she claims that a minimal expectation of a conscious particle would be to expect it to be capable of change, so as to account for the possibility of different ‘thoughts’. This is somewhat of a non sequitur: while she claims that ‘it’s hard to have an inner life with just one thought’, she doesn’t really offer much in support of this idea. To me, it doesn’t seem that difficult: just imagine the unchanging experience of seeing a white wall, above. You might claim that that’s a bit boring, but it wouldn’t be to the entity having that sort of experience, because it would preclude the further experience of being bored.

But the real problem with her argument is that she expects the particle’s inner experience to necessitate modification of its physical state. Basically, she argues that the electron’s state is exhaustively defined by a brief list of numbers—its spin, charge, mass, and so on—and that any addition to this list would be experimentally detectable, because it would change the production rate of particles in certain physical experiments. Both of these claims are true; but they tell us nothing whatever regarding the possibility of panpsychism.

The error is in thinking that experiential properties are something ‘added’ to the particle over and above its physical properties; but in fact, on a panpsychist construal as outlined here, they are what grounds the relations that we characterize as ‘physical properties’ (such as the property ‘length’ is really the relation between an object and some standard yardstick).

Perhaps an analogy helps to make this more clear. If Hossenfelder’s argumentation were right, then an electron could also not bear any information, for to do so, we would have to extend its state by whatever information it carries—say, a single bit, one or zero. This, then, should again affect the branching ratio in certain experiments. As nothing of this sort is observed, we infer that electrons can’t, in fact, carry information, and as the whole of modern digital technology uses electrons as carriers of information, we conclude that computers don’t exist.

Seeing as how the device I’m typing this on didn’t spontaneously vanish in a puff of logic on the above argument, something must be wrong about it. The resolution is that electrons, of course, don’t carry information in addition to their properties; they carry it by means of those properties. For example, an electron in the state ‘spin up’ may represent a logical 0, and in the state ‘spin down’, a logical 1. And just as it may be a bearer of information by means of its state, it may be a bearer of experience. Indeed, on several popular panpsychist construals—such as the ‘integrated information theory’ of neuroscientist Giulio Tononi or David Chalmers’ double-aspect theory—the two may be one and the same thing.

Consequently, Hossenfelder’s criticism of panpsychism ultimately boils down to a basic misconstrual of the theory as ‘adding’ further, mental properties to physical systems. We’re not going to find any empirical refutation of metaphysical claims here.

Oil Into The Fire

While thus Hossenfelder flounders on the philosophy, it’s the physics that are, for the most part, troublesome in Kastrup’s piece. Like Hossenfelder, Kastrup doesn’t think that elementary particles could be conscious as a matter of principle; however, his take is that that’s simply because they don’t exist. Rather, nature is made up of (quantum) fields, and particles are just localized excitations of these fields, like ripples on an otherwise calm pond.

While there’s some truth to that, matters are also rather more complicated than Kastrup lets on. But let us first investigate why, exactly, he thinks that a field-ontology is incompatible with panpsychism, while a particle-ontology might be.

Kastrup argues that what renders my experience distinct from yours is the relative spatial boundedness of the entities giving rise to it, namely, our brains. Because my brain is here, and your brain is there, my experience is mine, and yours is yours.

Particles, then, seem well-suited loci for conscious experience, being themselves well localized spatially. But, Kastrup claims, the world is not made of particles, but of fields; and these fields have no such neat spatial boundaries. Hence, there is nothing to differentiate my experience from yours on panpsychism; but as there clearly is such a difference, panpsychism fails.

But matters are not nearly as clear-cut. First of all, the fields of quantum field theory (QFT) have preciously little in common with fields as we would think of them in an everyday context. Such a field is given by a distribution of quantities across a certain domain: say, the value of the air pressure over a region of land. For each position, the pressure has a certain value, and the totality of these values constitutes the pressure-field.

Quantum fields, on the other hand, are fields of operators—mathematical ‘machines’ that act transformatively on other mathematical entities. Therefore, it’s not clear what might be meant by considering a particle to be a ‘ripple’ in such a field.

Moreover, even supposing that quantum field theory is best interpreted in terms of fields, what matters for our experience are the observations we can make—which are given by so-called ‘observables’, that, in QFT, are represented by explicitly local mathematical objects. (Indeed, their locality is axiomatic in rigorous formulation of QFT.) Otherwise, it would be problematic to derive any local concepts at all. Indeed, why Kastrup thinks it could be a problem for consciousness to be localized in a world composed of fields, while at the same time seeing no issue with our localized bodies and brains emerging from those same fields, seems a bit of a mystery.

Aside from QFT, Kastrup also delves into the notion of quantum entanglement to support his contention that there are no independent particles—or, rather, that ‘the physical properties of a particle depend on what one chooses to measure about another, entangled particle far away’.

Interpreted in the most straight-forward way, this is simply false. Considering only one localized party in such a setup, it will be impossible for them to tell whether they have a singular particle or one of an entangled pair—they will not be able to detect any change on their end, no matter what happens to any putative entangled partner. Otherwise, entanglement could be exploited to send information faster than light, in conflict with the tenets of special relativity, and known to be impossible thanks to the quantum no-communication theorem. It is only once both parties in such a setup get together and compare their measurement results that they may conclude that their results are more strongly correlated than would be possible without entanglement. On their own, neither notices anything amiss.

A hope for the particle picture, Kastrup acknowledges, might be provided for by the Bohmian picture of quantum mechanics, which adds explicit localized particles to the quantum ontology bearing the properties revealed in experiments. He notes that this is difficult to square with special relativity—which is for the most part true, but not as hopeless as he makes it out to be: several candidate relativistic formulations exist.

But moreover, he also claims that Bohmian mechanics has been experimentally refuted years ago! As the theory is, by construction, empirically exactly equivalent to standard quantum mechanics, this would be a shocking development, if true. However, in support of this claim, Kastrup points to an experiment done on a model system devised by Yves Couder, using oil droplets bouncing on a vibrating bath. But it is absolutely no surprise that this system fails to fully account for quantum features, lacking the nonlocal influences needed to make Bohmian mechanics work; and thus, its failure does not tell us anything about the viability of the Bohmian picture.

Kastrup’s criticism, furthermore, seems doubly odd: his central claim is that a conscious ‘field’ could not give rise to the sort of disjunct conscious experience you and I enjoy. But on his own version of idealism, where reality is fundamentally mental rather than material, ultimately, any conscious subject is a dissociated ‘alter’—in the sense of dissociative identity disorder—of the universe. Why could there not be some similar process giving rise to individuated mental experience, starting from an underlying conscious ‘field’ of whatever nature?

In the end, then, neither Kastrup’s nor Hossenfelder’s arguments succeed in bringing metaphysics into the laboratory. Both fail to appropriately engage with their subject matter—albeit from different angles. Both, also, do so with a good measure of scorn directed towards those considering panpsychism a live option. Perhaps, then, the takeaway message is not, as Hossenfelder cautions us, to be weary of philosophers speaking about elementary particles—but rather, of those who would too quickly dismiss difficult matters that have engaged some of the finest minds for centuries.