Global workspace theory (GWT)
Proposed by psychologist Bernard Baars in 1988 and developed into the global neuronal workspace by Stanislas Dehaene, Jean-Pierre Changeux and colleagues. The brain contains many specialised processors working in parallel and unconsciously. Information becomes conscious when it enters a limited-capacity "workspace" and is broadcast widely, so that memory, planning, language and action can all use it at once. A signature prediction is "ignition": a sudden, widespread surge of activity, involving prefrontal and parietal areas, when a stimulus becomes conscious.
Integrated information theory (IIT)
Developed by neuroscientist Giulio Tononi from 2004, with the latest formal version (IIT 4.0) published in 2023. IIT starts from the properties experience seems to have, such as being unified and specific, and asks what physical systems must be like to have them. Its central quantity, Φ (phi), measures how much a system's causal structure is integrated: how much the whole does that its parts cannot do independently. IIT predicts that consciousness depends on physical causal structure, centred on posterior regions of the cortex, and that standard digital computers may have very little of it even when running brain-like software.
Φ ≈ how much information the whole system specifies, over and above its parts
Recurrent processing theory (RPT)
Associated with Victor Lamme. When you see something, activity first sweeps forward from the eyes through the visual hierarchy. According to RPT, this feedforward sweep is not enough. Experience arises when higher areas send signals back to lower ones, creating recurrent loops within sensory cortex. RPT suggests consciousness may not need the frontal "broadcast" that GWT emphasises.
Higher-order theories (HOT)
Defended by philosophers such as David Rosenthal and neuroscientists including Hakwan Lau. A mental state is conscious when the system has a suitable representation of that state, a kind of inner awareness that "I am seeing this". These theories link consciousness to metacognition and self-monitoring, and often point to the prefrontal cortex.
Predictive processing and the attention schema
Predictive processing views the brain as a prediction machine that constantly guesses the causes of its sensory input and corrects itself using errors. Anil Seth and others argue that experience, including the sense of self and emotion, is built from these predictions, especially predictions about the body's internal state. Michael Graziano's attention schema theory proposes that the brain builds a simplified model of its own attention, and that this model is what we describe as awareness.
Putting theories to the test
Because theories make different predictions, researchers have started running adversarial collaborations, where proponents of rival theories agree in advance on experiments that could count against them. The Cogitate Consortium tested predictions of global neuronal workspace theory and IIT with hundreds of participants, publishing results in Nature in 2025. The findings challenged key predictions of both theories rather than crowning a winner, a useful reminder of how open the field is.
Our framework draws one mechanism from several theories rather than betting on one: information integration from IIT, recurrent processing from RPT, broadcast workspaces from GWT, and self-representation from higher-order and attention schema views. We then test which mechanisms make a measurable causal difference in artificial systems.
Quick answers
What is the leading theory of consciousness?
There is no agreed leader. Global workspace theory and integrated information theory are the most widely discussed, alongside recurrent processing, higher-order and predictive processing theories.
What is Φ (phi) in integrated information theory?
Φ is IIT's measure of integrated information: how much a system's causal structure specifies as a whole, beyond what its parts specify independently. IIT holds that a system's consciousness corresponds to this integrated structure.
Further reading
- Baars, B. (1988). A Cognitive Theory of Consciousness. Cambridge University Press.
- Dehaene, S. & Naccache, L. (2001). Towards a cognitive neuroscience of consciousness. Cognition, 79.
- Albantakis, L. et al. (2023). Integrated information theory (IIT) 4.0. PLOS Computational Biology.
- Lamme, V. (2006). Towards a true neural stance on consciousness. Trends in Cognitive Sciences, 10(11).
- Brown, R., Lau, H. & LeDoux, J. (2019). Understanding the higher-order approach to consciousness. Trends in Cognitive Sciences, 23(9).
- Cogitate Consortium (2025). Adversarial testing of global neuronal workspace and integrated information theories of consciousness. Nature.