Supporters and detractors of a leading theory of how consciousness arises are stuck in an increasingly bitter debate. Opponents suggest that integrated information theory (IIT), which claims that consciousness can be defined on a mathematical spectrum, is pseudoscience that could be misused to influence sensitive debates around abortion and the sentience of artificial intelligences 鈥� while supporters say the detractors are just jealous.
快猫短视频s have long sought to explain how the brain gives rise to conscious experience, but two prominent ideas have recently come to the fore: IIT and global neuronal workspace theory (GNWT).
The former identifies five essential features of consciousness and asks what kind of physical system could exhibit them. It suggests consciousness arises from any system whose components exchange information in a mathematically defined way and can be quantified using a measure known as phi. The more integrated information there is, the higher phi鈥檚 value and is. IIT links consciousness with posterior brain areas, which have neuroanatomical properties suited to producing high phi values.
Advertisement
GNWT, by contrast, suggests that information competes for attention in cortical and thalamic brain regions. If one signal outcompetes another, it is broadcast across the brain, becoming conscious. This transmission happens at the beginning and end of an experience and is tied to the prefrontal cortex, at the front of the brain.
To test both ideas, at the Max Planck Institute for Empirical Aesthetics in Germany and her colleagues, who included people favouring either IIT or GNWT, scanned the brains of 256 people while they looked at objects such as faces and letters, identifying certain aspects of them. The study, which took seven years, examined which brain regions processed the images, how long they were active and how synchronised the activity was.
The results, which were first presented at a conference in 2023 and are formally published today in Nature, were inconclusive. Some findings leaned towards IIT. For instance, visual features like category and orientation were decoded in posterior brain regions, and brain activity during conscious perception appeared sustained. However, patterns of synchronicity matched expectations predicted by GNWT.
While some researchers have welcomed the work as a vital step towards understanding consciousness, others have been increasingly vocal in rejecting IIT since those initial results in 2023. That same year, 124 neuroscientists signed an open letter branding it as pseudoscience, arguing that, according to IIT, inactive computer logic gates, plants and early human fetuses are likely to be conscious. In a , published last month, 100 scientists reiterated that IIT鈥檚 core claims are untestable and fall short of the empirical basis required for scientific enquiry.
at the University of London, a co-author of both critiques, said he and his colleagues had a variety of concerns, from IIT being inconsistent with physical laws to its vague assumptions and its lack of clear, testable predictions. Some of the researchers fear that giving credibility to IIT, or giving the public the idea that IIT is a leading explanation of consciousness, could affect ethical decisions, he says.
鈥淚n the cases of coma patients, AI sentience and abortions, we are asking 鈥榟ow do we know if the patient, the fetus or the AI are conscious?鈥�. We cannot yet use brain activity to answer this question adequately and it would be dangerous, at this stage, to base our answers on any theory that was not empirically validated,鈥� says Frith.
at University College London warns that IIT鈥檚 universalism, allowing it to be applied to any system with the right properties, risks detaching the notion of consciousness from the human brain. 鈥淭he worry is that we might prematurely redistribute ethical concern to such systems,鈥� he says.
But supporters say that many theories of consciousness might affect ethical debates, not just IIT. at the Allen Institute for Brain Science in Seattle, a co-author of Melloni鈥檚 study, says the backlash partially stems from jealousy.
鈥淚IT was perceived to be sexier than other theories, getting more attention and money,鈥� he says. 鈥淭he writers [of the open letter] assembled a grab bag of pseudo-arguments 鈥� 鈥業IT receives funding from right-wing philanthropists, IIT is anti-abortion, IIT has panpsychist implications鈥�.鈥� None of these is a reason to dismiss the idea, he says. 鈥淎ny inferences or implications of a theory ought to be irrelevant to the question of whether or not it is correct.鈥�
at the University of Sussex, UK, who wasn鈥檛 involved with Melloni鈥檚 study or the anti-IIT commentaries, points out that there is a long history of scientific ideas being rejected due to their implications, from the idea that Earth isn鈥檛 the centre of the universe to Charles Darwin鈥檚 theory of natural selection. 鈥淐onsequentialism is not a valid reason to reject a theory as unscientific.鈥�
Melloni says the backlash is due to IIT being so different to other explanations of consciousness, but that doesn鈥檛 make it dangerous. 鈥淚f a theory suggests AI or fetuses may have some level of consciousness, that doesn鈥檛 mean it鈥檚 anything like ours,鈥� she says. Seth shares this view. 鈥淚IT鈥檚 prediction about consciousness in a set of inactive logic gates is actually not all that clear 鈥� it might predict a very simple contentless form of consciousness, which would not necessarily carry much significance,鈥� he says.
Ultimately, Melloni says branding theories as pseudoscience achieves little. 鈥淚t鈥檚 just a soap opera. What we need is more data, not more letters.鈥� Her team plans to release its data, which she hopes researchers will use to test all potential theories of consciousness. 鈥淪cience isn鈥檛 about being right, it鈥檚 about getting it right.鈥�
Nature
