PLoS ONE
February 10, 2023
Lucia Melloni, Liad Mudrik, Michael Pitts et al.
106 citations
An adversarial collaboration between proponents of Global Neuronal Workspace and Integrated Information Theory devised and preregistered two experiments that test contrasting predictions about the location and timing of correlates of visual consciousness. Six theory-impartial laboratories will follow the study protocol using fMRI, M-EEG, and intracranial EEG, with built-in replications between labs and within datasets. The project aims to provide decisive evidence for or against the two theories and clarify the neural footprints of conscious visual perception, while modeling large-scale, collaborative, and open science practice.
Nature
June 1, 2025
Oscar Ferrante, Urszula Gorska-Klimowska, Simon Henin et al.
94 citations
Different theories explain how subjective experience arises from brain activity. An open science adversarial collaboration directly juxtaposed integrated information theory (IIT) and global neuronal workspace theory (GNWT). Human participants (n = 256) viewed suprathreshold stimuli for variable durations while neural activity was measured with fMRI, MEG, and intracranial EEG. Information about conscious content was found in visual, ventrotemporal, and inferior frontal cortex, with sustained responses in occipital and lateral temporal cortex reflecting stimulus duration, and content-specific synchronization between frontal and early visual areas. These results align with some predictions of IIT and GNWT but substantially challenge key tenets of both theories, including a lack of sustained posterior cortex synchronization for IIT and a lack of ignition at stimulus offset and limited prefrontal representation for GNWT.
Neuroscience and biobehavioral reviews
March 1, 2025
Liad Mudrik, Melanie Boly, Stanislas Dehaene et al.
68 citations
In a structured public debate at the 2022 meeting of the Association for the Scientific Study of Consciousness, proponents of five major theories—Global Neuronal Workspace Theory, Higher-Order Theories, Integrated Information Theory, Recurrent Processing Theory, and Predictive Processing—clarified their theories' core mechanisms, foundational premises, and what each theory aims to explain. The discussion revealed more controversy than agreement, particularly on the most basic questions: what consciousness is, how to identify conscious states, and what any adequate theory must account for. Addressing these foundational disagreements is essential for advancing the field and enabling meaningful comparison of competing theories.
Cognition
September 1, 2023
Yoni Zion Amir, Yaniv Assaf, Yossi Yovel et al.
25 citations
People can have a phenomenal experience—a raw, qualitative feel—without having immediate access to it or being able to report it at the moment. Using a novel paradigm, participants (Experiment 1, N = 40) lacked online access to a stimulus but could later retrospectively judge its phenomenal, qualitative aspects. A second experiment (N = 40) ruled out explanations based on unconscious processing or responses to stimulus offset. The results suggest that phenomenal and access consciousness are not only conceptually distinct but can also be empirically separated, supporting Ned Block's controversial dissociation.
bioRxiv Preprint Server
June 10, 2021
Itay Yaron, Lucia Melloni, Michael Pitts et al.
21 citations
preprint
A bird's-eye analysis of 412 experiments on the neuroscience of consciousness reveals that methodological choices alone can predict which of four leading theories a study will support, regardless of the actual findings. Most studies interpret their results after the fact rather than testing critical predictions in advance. This suggests that the field's theoretical commitments may be driven more by how experiments are designed than by the evidence itself. The authors provide an open-access website for further exploration of these trends.
The Behavioral and brain sciences
July 22, 2025
François Stockart, Maor Schreiber, Pietro Amerio et al.
20 citations
The scope of unconscious processing remains hotly debated, driven by diverse methods for manipulating and measuring perceptual awareness. Through dialogue among researchers with varied theoretical backgrounds, ten recommendations and nine outstanding issues are provided for designing experimental paradigms, analyzing data, and reporting results. These guidelines aim to evoke discussion about norms in studying unconscious processes and help researchers make informed decisions. While some recommendations may not align with existing approaches and will likely evolve, they are intended to foster a more convergent understanding of the extent and limits of unconscious processing.
Neuron
May 15, 2024
Liad Mudrik, Rony Hirschhorn, Uri Korisky
15 citations
The field of consciousness research has developed rigorous experimental methods, but these often rely on artificial paradigms that differ from everyday conscious and unconscious processes, raising concerns about ecological validity. This review argues that adopting more naturalistic approaches, as other cognitive science fields have done, can challenge existing hypotheses, yield stronger effects, and enable new research questions. Three paths are identified: changing stimuli and experimental settings, changing measures, and changing research questions. The authors review studies that have begun implementing such approaches and, while acknowledging challenges, call for increasing ecological validity in consciousness studies.
Trends in cognitive sciences
December 17, 2025
Liad Mudrik, Nathan Faivre, Michael Pitts et al.
5 citations
A major controversy in consciousness science divides sensory and cognitive theories. Reexamining Block's 1995 distinction between phenomenal consciousness (P) and access consciousness (A), the authors argue that P and A are not two different types of consciousness but two necessary conditions for consciousness. This conceptual shift helps resolve unresolved questions about neural mechanisms, functions of consciousness, and its relationship with attention. The proposal motivates selective unification across different classes of theories.
Scientific data
May 23, 2025
Alia Seedat, Alex Lepauvre, Jay Jeschke et al.
5 citations
An intracranial EEG dataset was collected from 38 epilepsy patients across three research centers as part of an adversarial collaboration testing Global Neuronal Workspace Theory and Integrated Information Theory. Participants viewed visual stimuli—faces, objects, letters, and false fonts—in three orientations and for three durations, performing a Go/No-Go target detection task. The dataset includes demographics, clinical information, electrode reconstructions, behavioral performance, and eye-tracking data, all converted to BIDS format. It is intended for reuse in consciousness science and vision neuroscience to investigate stimulus processing, target detection, and task-relevance.
Journal of experimental psychology. General
October 1, 2024
Amir Tal, Liad Mudrik
5 citations
A preregistered replication of a 2012 study tested whether people can follow an unseen instruction, integrate it with unseen digits, and perform addition without awareness. Two highly powered experiments failed to reproduce the original finding. The authors conclude that current evidence does not support the claim that arithmetic operations can be flexibly initiated without awareness, consistent with arguments for a more limited scope of unconscious processing.
Philosophy of Science
August 27, 2025
Niccolò Negro, Liad Mudrik
4 citations
Attributing consciousness to nonstandard systems such as animals, neural organoids, and artificial intelligence requires a robust justificatory framework. This article argues that analogical abduction—a reasoning method that combines analogical and abductive elements—is the strongest approach for extrapolating consciousness from humans to these systems. The argument from analogy and inference to the best explanation, taken individually, fail to meet the criteria for successful extrapolations. Analogical abduction offers a promising alternative, though limitations in current consciousness science remain. The article clarifies the argumentative structure behind such attributions, which is important for evaluating their validity.
Neuroscience of consciousness
January 1, 2026
Alicia Franco-Martínez, Ricardo Rey-Sáez, Jesús Adrián-ventura et al.
1 citation
Working memory may operate on unconscious perceptual contents, though it remains linked to conscious perception. A large, multisite replication (19 labs, 531 participants, 720 trials) of Soto et al. (2011) found above-chance accuracy (.55) on a visual discrimination task when participants reported not seeing the subliminal Gabor grating. Performance correlated positively with cue detection sensitivity (r = .228), and the regression intercept was significantly above chance (β₀ = .521). The study provides an open-access dataset and confirms that measures were reliable and valid, supporting the existence of unconscious working memory.
eLife
January 13, 2026
François Stockart, Alexis Robin, Hal Blumenfeld et al.
Identifying neural correlates of perceptual consciousness (NCC) remains difficult because most studies use non-invasive brain measures that lack the spatial or temporal precision to separate consciousness from other cognitive processes. Low sensitivity also makes null findings hard to interpret. This review argues that human intracranial recordings—which offer high spatiotemporal resolution, better signal sensitivity, and broad brain coverage—can advance the search for NCCs. It examines studies at the single-neuron and population level, evaluates their implications for debates between cognitive and sensory theories of consciousness, and discusses current limitations and future directions using emerging technologies and experimental paradigms.
Neuroscience of consciousness
January 1, 2025
Shaked Palgi, Tamara Bester-Arest, Nathan Faivre et al.
Unconscious mental processing may not extend to integrating relationships between objects. In five experiments, participants viewed pairs of objects that were either related (e.g., a lock and key) or unrelated, presented either visibly or invisibly through masking. When pairs were visible, behavioral priming and a larger N400 brain response occurred for unrelated versus related pairs, indicating relational processing. When pairs were invisible—verified by subjective and objective awareness measures—no differences in N400 amplitude were found, and decoding of pair relations from brain activity was no better than chance. The results suggest that consciousness is needed for integrating relationships beyond single objects, supporting theories that emphasize consciousness's role in relational integration.
Scientific data
May 29, 2026
Ling Liu, Oscar Ferrante, Tara Ghafari et al.
A large-scale, multi-center dataset combines MEG, EEG, eye-tracking, and structural MRI recordings from 100 individuals (mean age 22.79, 54 female, all right-handed) across two research centers (UK and China). The data were collected through an adversarial collaboration between advocates of the Global Neuronal Workspace Theory and the Integrated Information Theory of consciousness. Participants performed a non-speeded Go/No-Go target detection task with visual stimuli from four categories (faces, objects, letters, false fonts) at different orientations and durations (0.5, 1.0, 1.5 s) under various task conditions. The dataset follows the Brain Imaging Data Structure (BIDS) and includes extensive metadata to enhance reusability.
Scientific data
May 7, 2026
Aya Khalaf, David Richter, Yamil Vidal et al.
An adversarial collaboration between proponents of the Global Neuronal Workspace theory and the Integrated Information Theory of consciousness produced a functional magnetic resonance imaging dataset from 118 participants. Participants viewed faces, objects, letters, and false fonts presented at three orientations and three durations (0.5, 1.0, 1.5 seconds) while identifying infrequent targets. Two categories were task-relevant and two were task-irrelevant in each block. The simple design makes the data broadly reusable for testing predictions from other theories of consciousness and for examining visual processing. Anonymized data, quality reports, demographics, behavioral performance, and eye-tracking data are publicly accessible.
arXiv Preprint Archive
August 17, 2023
Patrick Butlin, Robert Long, Eric Elmoznino et al.
No current AI systems are conscious, but there are no obvious technical barriers to building ones that might be, according to an analysis grounded in neuroscientific theories of consciousness. The report surveys prominent theories—recurrent processing, global workspace, higher-order, predictive processing, and attention schema—and derives computational indicator properties from them. Applying these indicators to recent AI systems yields no evidence of consciousness, but the authors argue that future systems could potentially implement the necessary properties.
bioRxiv Preprint Server
June 23, 2023
Oscar Ferrante, Urszula Gorska-Klimowska, Simon Henin et al.
preprint
An open science adversarial collaboration directly juxtaposed Integrated Information Theory (IIT) and Global Neuronal Workspace Theory (GNWT) by investigating neural correlates of visual experience. 256 human subjects viewed suprathreshold stimuli for variable durations while neural activity was measured with fMRI, MEG, and ECoG. Information about conscious content was found in visual, ventro-temporal, and inferior frontal cortex, with sustained responses in occipital and lateral temporal cortex reflecting stimulus duration, and content-specific synchronization between frontal and early visual areas.