In 1998, at the conference of the Association for the Scientific Study of Consciousness (ASSC), the neuroscientist Christof Koch made a bet with the philosopher David Chalmers: by 2023, science would be able to explain how the brain’s tangle of neurons gives rise to the phenomenon we call consciousness. The winner would get a case of wine.
Koch was a professor of cognitive biology who helped pioneer the mechanistic study of the “neural correlates of consciousness,” which maps the relationship between brain activity and subjective experiences. He believed that consciousness was fundamentally measurable and that it was only a matter of time before science identified how it arose in the brain.
Chalmers was both a philosopher and cognitive scientist who was skeptical that science would be able to build explanatory bridges between neural correlates in the brain and the subjective experience of consciousness. Famously, he called consciousness “the hard problem,” which he believed was sufficiently challenging to keep any explanation of consciousness at bay for at least a quarter of a century.
At the 26th ASSC conference this past weekend, 25 years after the initial wager, the results were declared: Koch lost. Despite years of scientific effort — a time during which the science of consciousness shifted from the fringe to a mainstream, reputable, even exciting area of study — we still can’t say how or why the experience of consciousness arises.
Galileo split consciousness away from science 400 years ago
While the Western science of consciousness only grew into a reputable field over the past few decades, part of the reason answers remain so elusive may be buried in the deep structure of scientific inquiry itself, reaching back to the 1600s.
The Italian astronomer Galileo Galilei is widely credited with inventing the scientific method. As the philosopher Philip Goff recounts in his 2019 book Galileo’s Error, in order to formalize the study of objective qualities like size, shape, location, and motion, Galileo bracketed out the fuzzier domain of conscious experiences. The modern scientific endeavor he helped create is a study of the universe shorn of what Galileo called the soul, and what we today might call sensory qualities, the gestalt of what consciousness feels like. The scientific method can explain the electrical activity that sparks in the brain when you jump into a freezing lake, but it can’t explain why a subjective experience of invigoration comes along with it.
“Those sensory qualities have come back to bite us,” Goff writes. “Galileo’s error was to commit us to a theory of nature which entailed that consciousness was essentially and inevitably mysterious.”
In other words, Galileo’s scientific method required walling off the study of consciousness itself, which is why it’s perhaps not surprising that even centuries later, his method’s inheritors still struggle to explain it.
New avenues in the science of consciousness
As the years passed, Chalmers and Koch forgot about their bet, but in 2018, the science journalist Per Snaprud brought it back to their attention. A few years later, as part of a $20 million project supported by the Templeton World Charity Foundation, a series of “adversarial experiments” were designed to pit theories of consciousness against each other, including global workspace theory (GWT) and integrated information theory (IIT).
GWT imagines consciousness as a theater: The brain is populated by a crowd of local information streams, but only what gets broadcast to the whole crowd — put onstage — becomes conscious. IIT identifies consciousness with the degree of, yes, integrated information, represented by the Greek character phi (Φ). The more phi, the more consciousness.
Preliminary findings from one Templeton-sponsored gauntlet comparing GWT and IIT were presented at the recent ASSC conference and ultimately used to settle the Koch/Chalmers bet. Six independent laboratories followed a shared protocol designed to test how well each theory could predict brain activity. IIT fared slightly better than GWT, but neither made entirely accurate predictions. This uncertainty was enough to make Chalmers the victor, while scattering researchers off to update the theories or consider new ones altogether.
Goff’s preferred resolution is to reintroduce consciousness into our understanding of nature by way of a secular version of panpsychism, the theory that consciousness is a fundamental and ubiquitous element of the physical world. In this view, physical sciences à la Galileo describe matter from “the outside.” Consciousness is also a property of matter, but matter as experienced from the inside.
Meanwhile, Karl Friston, the world’s most-cited living neuroscientist, has an idea called the free-energy principle. Stripped of all the math, it suggests that the behavior of all living systems follows a single principle: To remain alive, they try to minimize the difference between their expectations and incoming sensory input. (Other terms for that difference include surprise and free energy.) In this model, human brains minimize surprise by generating internal models that predict the outside world. Here, consciousness is basically the experience of an internally generative model complex enough to imagine states of the world that have not yet happened.
The process by which brains generate these internal models has a theory of its own, known as predictive processing, perhaps most associated with the philosopher Andy Clark. To grasp the idea, think of what’s happening during a dream. You’re lying in bed, eyes closed in a dark room, completely still. But your brain is generating a rich internal dream world that feels entirely convincing (lucid dreams aside). Well, predictive processing claims that the same sort of thing is happening during waking consciousness, with a few caveats.
In other words, the kind of world you’re experiencing when awake is basically the same kind of world you experience in a dream: a hallucination. The difference is that our brains are constantly comparing our waking hallucinations to the sensory input they receive from the outside, fine-tuning the waking dream to keep it in line with what the incoming sensory data suggests is going on beyond our skulls. That’s what the neuroscientist Anil Seth means when he calls consciousness a “controlled hallucination.”
Okay, so living systems want to minimize surprise, and predictive models help brainy creatures do so. But what makes consciousness feel the way it does? How can we explain why some states of consciousness feel so rich and alive while others feel so dreary? One interesting idea hovering on the periphery of consciousness science is the symmetry theory of valence (STV), first proposed by the independent philosopher Michael Johnson and his collaborators at the Qualia Research Institute, a nonprofit focused on the science of consciousness.
The STV starts with the idea that you can map every state of consciousness onto a perfect mathematical representation, like a unique objective signature for each subjective state (it shares this idea with IIT). Next, it claims that the valence, or positive/negative feeling of any given state of consciousness, depends on the symmetry of that representation. In practice, drawing on the work of neuroscientist Selen Atasoy, they use the underlying neural activity as that representation.
Every conscious state has an associated orchestra of neural activity that gives rise to harmonic patterns across the brain. QRI co-founder Andrés Gómez Emilsson figured out how to decompose that activity in a way that deciphers how much consonance exists across the brain harmonics, which works as a proxy for the symmetry. The more symmetry in the brain, the more positive the experience. Inversely, the more dissonance and less symmetry, the more negative the experience. While the STV hasn’t received much mainstream attention, its ideas are beginning to crawl their way into citations on papers at the forefront of the science of consciousness.
So we have a growing constellation of relevant theories, though as the result of the Koch/Chalmers bet suggests, we still lack a definitive, falsifiable explanation. We even lack consensus on whether one may ever exist.
Toward a paradigm of consciousness science — or not
Still, some neuroscientists argue that we are living in the dawn before a theory of consciousness arises, like those who lived in the time shortly before Darwin’s theory of natural selection. This paints the current field as “pre-paradigmatic,” a term developed by the philosopher of science Thomas Kuhn to describe an immature science where competing schools of thought do not share the same basic understanding of their subject. Everything from methodologies to metaphysics can differ in a pre-paradigmatic science of consciousness.
Eventually, in this view, the field might coalesce around a unified theory and the first true paradigm of consciousness science would begin. This is the view Koch continues to hold (despite being down a case of fine Portuguese wine). He doubled down at the recent ASSC conference, renewing the bet on the same 25-year horizon. Chalmers, too, reports plenty of progress, telling Nature that the problem of consciousness “has gradually been transmuting into, if not a ‘scientific’ mystery, at least one that we can get a partial grip on scientifically.”
But there’s no guarantee that some critical mass of correlations between brain states and feelings can ever tell us how or why consciousness happens. Chalmers suspects that at the conclusion of their renewed bet in 2048, despite all the surrounding progress of insight that’s sure to unfold, the mystery may remain as perplexing as ever.