Anil Seth’s book, Being you. A New Science of Consciousness, Faber & Faber, London 2021, proposes an intriguing theory of consciousness that seems quite convincing, although it explicitly refuses to confront the hard problem of consciousness, as David Chalmers calls it, and so evades the crucial question in my opinion.
What, then, are the main reasons for the volume’s interest?
The basic interpretation of the consciousness that Seth provides is convincing: it would be a system that, based on a predictive approach (“bayesian”), “hallucinates” a “functional” reality, not necessarily similar to the “true reality” (Seth himself in the epilogue recalls the Kantian distinction between phenomenon and noumenon and alludes to the possibility that even the three-dimensional fabric of space – unfortunately he does not speak of time, which is very close to my heart, – can be a “perceptive effect”).
Within certain limits it also convinces the “naturalistic” point of view adopted by Seth, who assigns more relevance to the “real problem” of consciousness, as he calls it, than to the “hard problem”. I intend this as follows: any interpretation of the functioning of consciousness must be consistent with empirical data: although consciousness, as a subjective experience, is not the brain, nothing prevents (indeed everything suggests) that there is a close correlation between consciousness and brain (or body), between “inside” and “outside”. So it is right to “naturalize” the problem of the consciousness not assuming only a “transcendental” perspective (This “naturalization” does not exclude such a trascendental perspective: phenomenological analysis, simply, cannot be incongruous with the empirical data: e.g. it cannot be that I perceive a cat if my brain sends the waves typical of deep sleep).
It must be said that, although Seth is a neuroscientist, his theory appears more speculative than “scientific”. Seth certainly takes inspiration from observations and experiments of his own and of others, but what gives meaning to the volume is an overall interpretation that is certainly authorized by empirical data, but is not the only possible.
In my point of view a problem, as mentioned, is that Seth, despite his speculative “drift”, claims to evade the “hard problem”.
He claims to distinguish what he calls real problem (the problem of explaining “scientifically” consciousness) from “easy problems”, as Chalmers calls them, which would concern the “functioning” of consciousness. Yet, in the end, Seth also explains consciousness functionally (with his Bayesian theory) taking for granted the phenomenology of consciousness. But this phenemology is exactly what distinguishes consciousness as first-person experience. This should have been a matter of explaining. But Seth doesn’t follow this way.
Seth ultimately fails to “justify” the very existence of consciousness. Why couldn’t its effective Bayesian functioning be implemented by an unconscious system, as are, according to Seth, some different forms of artificial intelligence?
The reductionist approach, in general, to nature is useful but not exclusive (as Seth presents it with epistemological fallacy in my point of view).
Let’s consider the reductionist approach to life in modern biology, which Seth considers paradigmatic. It merely shifts the “hard problem” (what life is and why it exists), doesn’t dissolve it into a cloud of metaphysical smoke, as Seth thinks.
The same applies to the problem of the existence of consciousness (Chalmers’ hard problem, in fact). If I am a self-regulating system whose goal is to survive, why do I need to be conscious? You can build systems that self-regulate themselves based on the inputs they receive without being alive and conscious.
The right distinction that Seth makes, in the last chapter before the epilogue, between consciousness and intelligence suggests (cf. the pages on which he guesses unlikely that a machine capable of “predictive processing” is aware) that this “predictive processing”, which according to him characterizes consciousness, can characterize also simply an intelligent system capable of self-regulation, but unconscious.
Seth also discusses the interesting paradox of teleportation: if a certain Eva entered, here on Earth, into a teleportation machine that would disintegrate her and then rebuild her elsewhere, e.g. on Mars, could you say that Eva is always Eva? Would her consciousness be the same? And if, due to a machine failure, the “terrestrial” Eva survives and a second Eva, identical to the first, is reproduced on Mars, which one would Eva be? (Seth also evokes somewhere the “split brain” and the interesting real case of the Siamese twins united by the brain).
In all these cases, in my opinion, we continue to escape the problem. It is not enough to say “both women are Eva”. You who have entered the machine, will you find yourself on Mars or on Earth? My answer is: on Earth . On Mars there is a clone (with your memory). It cannot be physical identity that determines the continuity of consciousness. If anything, the opposite. Everyone who enters a transporter machine dies instantly even if no one notices.
Likewise the representation of a “diffused” consciousness in the octopus, which Seth presents in the last chapter, does not seem justified to me. Seth can recognize that parts of the octopus move with great autonomy and intelligence (after all, our heart and our immune system do too). But a consciousness that is not “one” is unthinkable. Whatever it was, it would not be what we mean by consciousness.
Attention: I don’t’ claim that whoever is conscious (someone in particular or the universe itself) must conceive of himself or of herself necessarily as one and as a body. This may well be a hallucination. One can identify himself or herself with this or that, but who identifies himself or herself with this or that – suppose erroneously – cannot but have one consciousness if he or she has consciousness. It makes no sense to say that he or she can have two or more “consciousnesses” or a “diffused” consciousness, whatever this means. If two people had the same consciousness, exactly as two parts of an animal, they can simultaneously perceive themselves suppose both in Rome and in London: they must have only one consciousness with two or more perceptions (as I now see more colors). If they have this “consciousness” at different times, they still has only one consciousness that simply moves and becomes filled in time with different contents.
Seth roots consciousness in life rather than intelligence. Interesting and plausible. But how to prove it?
It’s just a conjecture.
Many living activities, even human and even very complex ones, take place unconsciously. Perhaps the function of consciousness is linked to the possibility of experiencing pleasure and pain to orient oneself in the world between opportunities and dangers? But even an unconscious mechanism emulated by a robot could do it…
About function of plesure and pain see my recent post.
We come, finally, to the epistemologically pivotal issue, in my opinion.
If everything is controlled hallucination, including external objects, even the limbic system or the cerebellum or the neurocortex are hallucinations, even the living body is a hallucination that seems to us to occupy a three-dimensional space, even space and time themselves: Seth assigns them a role just because he needs to believe that they exist as external objects in order to survive (as a neuroscientist rather than as a mystic?). It is the “mise en abime” that distinguishes every radical naturalism (taken to its extreme consequences) that ends up making it an idealism.
For the truth, on page 272 Seth seems to apply his theory of consciousness to itself: as consciousness works as predictive, not “objective”, so also a theory of consciousness works if it obeys a Bayesian epistemology. But it is like saying that his theory can safely be also false! In short, it does not fall into the paradox of the liar?
In this “idealistic” perspective I can agree with Seth that the self is a hallucination, not unlike other “objects”: maybe the One (God, Shiva etc.) identifies himself erroneously with me and with you (and with Seth and so on).
But Seth seems to think that “who” is wrong in these identifications is not the One, but my or your or his “living body”, This is because, in my opinion, he ends up committing the same mistake that he unmasked, assigning to a phantom living body a fundamental reality, while on the basis of its own criteria and results it must be considered a hallucination, a construct functional to life.
In general, however, this book is really very interesting and inspiring, everyone should read it.
But with a warning: as mentioned above, many of Seth’s theories (e.g. that consciousness is linked to life or is predictive), as many other intriguing theories that Seth cites (e.g. Friston’s Free Energy Theory), are not immediately derived from experiments but seem speculative interpretations of experiments (which is fine to me, I find it inevitable) and provoke further speculative hypotheses.
With a pun you could say that the theory of predictive consciousness is not itself… predictive (controllable by experiments that could falsify it). Too much Bayesianism, after all, can inebriate, as Popper argued!