Psychologist logo
Cognition and perception

Grand unified theories of consciousness

An extract from ‘Models of the Mind: How Physics, Engineering and Mathematics Have Shaped Our Understanding of the Brain’ by Grace Lindsay, with kind permission from the publisher Bloomsbury Sigma.

09 April 2021

By some accounts, no theory of the brain could be complete without explaining its biggest and most enduring mystery: consciousness. The C-word can be a tough subject for scientists, laden as it is with centuries of philosophical baggage. Yet, in the eyes of certain researchers, a precise scientific definition that could be used to not just identify but also quantify consciousness anywhere it may exist is the holy grail of their work. It is also the promise of ‘integrated information theory’.

Integrated information theory (or IIT) is an attempt to define consciousness with an equation. It was originally put forth by Italian neuroscientist Giulio Tononi in 2004 and has been iterated on by him and others ever since. IIT is designed to measure consciousness in anything: in computers, rocks and aliens as easily as in brains. By making a universal claim of what consciousness is, it differs from the more biology-centred theories devised by some neuroscientists.

IIT is able to free itself from the specific physical features of the brain because its inspiration comes from another source entirely: introspection. By reflecting on the first-person conscious experience, Tononi came up with five important traits fundamental to consciousness; these are the ‘axioms’ on which IIT is built. The first axiom is the basic fact that consciousness exists. Others include the observation that a conscious experience is composed of multiple distinct sensations, the experience is specific, it appears to us as an integrated whole and it is uniquely what it is – no more or less.

Tononi considered what kinds of information-processing systems could give rise to these axioms of experience. Through this, he was able to map the axioms to mathematical terms. The end result is a unified measure of what is called ‘integrated information’, a value Tononi symbolises with the Greek letter phi. In total, phi indicates just how intermixed the information in a system is. The right kind of intermixing is supposed to give rise to the richness and wholeness of experience. According to IIT, the higher the phi that a system has, the more conscious it is.

As it turns out, calculating the phi for a system with any reasonable amount of complexity is nearly impossible. For the human brain, it would first require conducting a near endless amount of experiments in order to probe how the different sub-structures of the brain interact. Even if that could be done, a long and gruelling series of computations would then begin. To overcome this hurdle, multiple approximations to phi have been devised. Through this, it is possible to make educated guesses about the phi in a system. This has been used to explain why certain brain states lead to more conscious experience than others. For example, during sleep, the ability of neurons to communicate effectively is interrupted. This makes the brain less able to integrate information, resulting in lower phi. According to Tononi’s theory, similar reasoning can explain the unconsciousness that comes with seizures as well.

The theory also makes some, perhaps surprising, predictions. For example, the phi of an average thermostat is small, but still not zero. This implies that the device regulating your room temperature has some amount of conscious experience. What’s more, some very simple devices – if built just right – can actually have a value of phi much higher than the estimated phi of the human brain. These counterintuitive conclusions make some scientists and philosophers sceptical of IIT.

Another critique of the theory is targeted at its axiomatic basis. According to this argument, the axioms Tononi chose aren’t the only ones that a theory of consciousness could be built on. And his way of mapping these axioms to mathematics isn’t obviously the only, or best, way either. The problem is: if the foundations of IIT are arbitrary, then how can we trust the conclusions that spring from them, especially when they surprise us?

An informal survey of consciousness scientists conducted in 2018 revealed that IIT was not the favoured theory among experts (it came in fourth after two other theories and the catch-all category of ‘other’). But the same survey found that IIT fared better among nonexperts: in fact, it was rated first among the subset of non-experts who felt they had enough knowledge to respond. Some of the survey authors suspect this may be a result of IIT’s PR. From the outside, the theory looks well founded if only because it has the authority of hard math behind it. And more than most scientific theories of consciousness, IIT has been featured in the popular press. This includes writings by Christof Koch, a prominent neuroscientist who has become a collaborator of Tononi’s and a public advocate of IIT. In his book, Consciousness: Confessions of a Romantic Reductionist, Koch describes his personal journey through the scientific study of consciousness, including work he did with Nobel Prize winner Francis Crick, and his views on IIT.* Such popular accounts may be effective in getting the theory out to a broader audience, but don’t necessarily convince scientists in the know.

Even scientists who lack faith in the power of IIT still tend to applaud the attempt. Consciousness is a notoriously difficult concept to tame, which makes IIT’s effort to submit it to orderly scientific inquiry still a step in the right direction. As vocal critic of IIT physicist Scott Aaronson wrote on his blog: ‘The fact that integrated information theory is wrong — demonstrably wrong, for reasons that go to its core – puts it in something like the top 2 per cent of all mathematical theories of consciousness ever proposed. Almost all competing theories of consciousness, it seems to me, have been so vague, fluffy and malleable that they can only aspire to wrongness.’

* Tononi himself has also written a book aiming to explain his theory to a broader audience. In Phi: A Voyage from the Brain to the Soul, Tononi tells a fictional tale of seventeenth-century scientist Galileo Galilei exploring notions of consciousness through interactions with characters inspired by Charles Darwin, Alan Turing and Crick.

-        This extract is taken from Models of the Mind: How Physics, Engineering and Mathematics Have Shaped Our Understanding of the Brain by Grace Lindsay (Bloomsbury Sigma), out now in hardback and ebook. Available at the Bloomsbury website and at all good bookshops. For 20 per cent off, use the code MODELS20 at https://www.bloomsbury.com/uk/models-of-the-mind-9781472966421/