Consciousness

Consciousness is the fact that some mental states are experienced from the inside: there is something it is like to see red, feel pain, or think about one's own thoughts. It is studied in philosophy of mind, neuroscience, psychology, and increasingly in artificial intelligence, and it remains one of the central unresolved problems in understanding nature. The word covers several different things, including being awake, being aware of something, and having subjective feeling, so the first task of any account is to separate these senses. The entry follows the history of the question from Plato and Aristotle through René Descartes to modern theories, and then surveys the empirical work on neural correlates, clinical disorders, and the main competing theories, including the 2025 adversarial collaboration that tested two of them directly.

Two problems dominate the field. The first is the "hard problem", the question of why physical processes in the brain should be accompanied by experience at all, which is distinct from the "easy problems" of explaining attention, memory, and reportability. The second is methodological: consciousness is accessible directly only to the one who has it, so researchers must rely on reports, behavior, and physiological markers to study it in others, including animals, patients with brain injury, and machines. Each theory discussed here, from global workspace and integrated information approaches to higher-order, predictive, and quantum proposals, takes a different stance on these two problems, and each faces empirical and conceptual objections.

The entry also addresses the practical stakes of the question. Assessments of covert awareness in patients with disorders of consciousness, the moral status of animals and of possible artificial systems, and the cultural representation of inner life in literature all depend on how consciousness is understood. The entry closes by identifying the questions that remain open, where neither the evidence nor the conceptual framework has yet produced consensus.

Concepts and distinctions

Three senses of "conscious"

The English word "conscious" covers at least three different things. In the first, a person is conscious when awake, as opposed to asleep, anaesthetised or in a coma. This is wakefulness, the state clinicians check by testing responsiveness. In the second, a person is self-aware, able to represent themselves as a subject with thoughts and a body; this is the sense captured by "self-consciousness". In the third, a mental state is conscious because there is something it is like to be in it, as with the seen red of a tomato or the sting of a pinprick. In ordinary life these senses usually coincide, which is why they are easily conflated, but they can separate. A dreamer may be vividly experiencing a scene without any self-awareness that they are dreaming, and a person may be awake yet report nothing. Researchers therefore need to say which sense they mean before drawing conclusions.

Block's two concepts

Ned Block's 1995 distinction separates phenomenal consciousness from access consciousness. Phenomenal consciousness consists of subjective experience and feelings, while access consciousness consists of information that is globally available in the cognitive system for reasoning, speech and high-level action control. Block treats the two as conceptually distinct, so they could in principle come apart. One can have P-consciousness without A-consciousness, and vice versa. The best-known case is his "overflow" argument: phenomenal consciousness overflows conscious access, meaning that one can consciously experience something that they do not have conscious access to. Critics reply that the access notion lacks a clear definition and a clear criterion for separating accessible from inaccessible states, which makes it hard to test whether the two really come apart. Some global-workspace theorists go further and hold that the two are identical. The distinction matters methodologically: if access and experience can separate, then a subject's report cannot be assumed to measure experience directly.

Qualia and the "what it is like" test

Phenomenal consciousness refers to qualia, i.e., first-person experiences. Qualia are the felt, qualitative properties of experience, such as the particular redness of red or the particular quality of a toothache, as they appear to the subject. Many philosophers use "qualia" as shorthand for phenomenal character, while others, including some illusionists, deny that qualia so defined exist. The phrase "what it is like" gained wide currency through Thomas Nagel's 1974 essay "What Is It Like to Be a Bat?" Nagel argued that an organism has conscious mental states just in case there is something it is like to be that organism. Echolocation is so unlike human perception that, Nagel held, we cannot imagine its subjective character even when we know its physical workings. The formulation is best read as a marker of experience, identifying it by its subjective perspective rather than by its function, rather than as a complete definition.

Sentience, awareness, and mind

These terms overlap but do not coincide. "Sentience" usually refers to the capacity for experiences with a positive or negative quality, especially pain and pleasure. It is the term most often used in animal-welfare law, and it is narrower than consciousness in general, since a system could register its surroundings without feeling anything good or bad. "Awareness" is more flexible: cognitive psychology sometimes applies it to information that is registered and used without any experiential character, so the word can sit on either side of Block's divide depending on the author. "Mind" is the broadest term. It covers belief, memory, intention and thought, and many of these states are not conscious at a given moment, for example stored beliefs during dreamless sleep. Consciousness is therefore one part of mind, not a synonym for it. Block has argued that the single word bundles together phenomenal, access, reflective and self-consciousness, which he considers scientifically disparate, and much confusion in the field traces to this bundling.

Historical roots in Western philosophy

Greek accounts of the soul

The Greek word psyche originally named the principle that animates a living being, not an inner space of awareness. In Plato's dialogues the soul is the seat of reason and desire. The Republic divides it into rational, spirited, and appetitive parts. The Theaetetus asks how the soul, using its own faculties, judges what the senses present, so for Plato perceiving is something the soul does and not only something the body undergoes. Aristotle treated the soul in De Anima as the form of a living body, with nutritive, perceptive, and rational capacities layered on one another. In his account, perception takes in the form of an object without its matter. He also proposed a "common sense" (koinē aisthēsis) that unifies the five senses and explains why we can notice that we are seeing or hearing. This faculty of registering one's own perceiving is among the earliest Western analyses of what later came to be called consciousness, although neither philosopher had a single term for it.

Descartes and the cogito

René Descartes made the question central to philosophy. In the Meditations on First Philosophy (1641) he set aside the testimony of the senses and even of mathematics until he reached one proposition that he could not doubt without thinking it: that he exists while he thinks. He then defined thinking broadly, to include doubting, understanding, affirming, willing, imagining, and sensing. The mind, as a thinking substance (res cogitans), is distinct from extended matter (res extensa). For Descartes, thinking and self-awareness were effectively identical, because the mind is transparent to itself in a way that bodies are not. This dualism created the problem of how mind and body interact. In The Passions of the Soul (1649) he suggested that the mind acts on the body through the pineal gland, a proposal few later philosophers accepted.

Locke, Leibniz, and Hume

John Locke, in An Essay Concerning Human Understanding (1690), grounded personal identity in consciousness. He described consciousness as the awareness of what goes on in one's own mind. A person, in his account, persists through time to the extent that the same consciousness can reach back to earlier actions and thoughts, so the self is tied to memory rather than to a soul or a fixed body. Leibniz challenged the assumption that all mental states are conscious. In the New Essays (written around 1704, published 1765) and the Monadology (1714) he introduced petites perceptions, minute perceptions too faint to be noticed one by one. His example of hearing the sea shows how many unnoticed sounds, each imperceptible alone, add up to an audible roar. He distinguished this from apperception, the reflective awareness of one's own states. David Hume pushed the empiricist reasoning further in the Treatise of Human Nature (1739–40). When he looked inward, he reported, he found only a bundle of perceptions and never a self that underlies them. For Hume, personal identity was a construction built from the resemblance and causal connection among perceptions.

Consciousness as an empirical subject

In the nineteenth century, consciousness moved from metaphysics toward the laboratory. Gustav Fechner's Elements of Psychophysics (1860) sought lawful relations between physical stimuli and sensation. Hermann von Helmholtz measured reaction times, and Wilhelm Wundt opened a laboratory at Leipzig in 1879 where trained observers reported their experiences under controlled conditions, a method known as introspection. William James's Principles of Psychology (1890) rejected the picture of thought as a set of discrete mental elements. He argued instead that thought flows continuously, and in the chapter on this topic he introduced the phrase "stream of consciousness." He also stressed the "fringe" of vague awareness around the focus of attention, and the selective role of attention itself. These works made consciousness a question for empirical psychology. The methods that followed, and their limits, are treated in a later section.

Mind and self in Indian, Chinese, and Buddhist thought

Ātman and Brahman in the Upanishads

The Upanishads, composed over several centuries from roughly the first millennium BCE, ask what remains of a person once body, senses, and speech are set aside. Their answer is ātman, the innermost self. The Māṇḍūkya Upanishad treats it as the witness that persists through waking, dreaming, and deep sleep. The Chandogya Upanishad's phrase tat tvam asi ("that you are") states the central claim that this individual self is identical with Brahman, the ultimate ground of reality. Later Advaita Vedānta, associated with Śaṅkara, reads this identity as nondual. Dualist schools of Vedānta deny that the two are the same. In these texts, awareness is not simply a function of the body or the brain. It belongs to the self's own nature, and this is why ātman is often described as pure consciousness.

Non-self and the mind as a stream

The Buddha's doctrine of anattā (non-self) turned against this picture. Early Buddhist analysis divides a person into five aggregates (khandha): form, feeling, perception, mental formations, and consciousness. None of these is a permanent owner of experience. The Buddhist account of mind is therefore a sequence of events rather than a substance. Abhidharma scholars analysed mental occurrences as momentary dharmas. In their model, a single moment of consciousness arises, endures, and passes away, and continuity comes from causal succession, not from a lasting carrier. The Theravāda notion of bhavaṅga, a "life-continuum" that sustains the stream between active moments of cognition, was used to explain how experience persists through sleep and unawareness. Doctrinal scholars stress that anattā denies a permanent self, not the reality of persons or experience.

Yogācāra and the storehouse consciousness

The Yogācāra school, associated with Asaṅga and Vasubandhu and active from roughly the fourth to fifth centuries CE, held that what appears to us is a representation of consciousness (vijñapti-mātra). It expanded the usual six consciousnesses into eight. These comprise the five sense consciousnesses, mental consciousness (mano-vijñāna), manas, and the storehouse consciousness (ālaya-vijñāna). The storehouse holds karmic "seeds" (bīja) and matures them into future experience. This explains how karmic effects persist across time, even though no self is involved. Manas is the seventh consciousness, a reflexive process that grasps the storehouse as "I". Some accounts describe the storehouse as only faintly aware and without purposive activity. Its status as either an unconscious substrate or a form of consciousness was debated within the school itself.

Heart-mind in Confucian and Mencian thought

In early Chinese thought, xin 心 denotes the physical heart and also the seat of cognition and feeling. Because the term covers both, translators often render it as "heart-mind". Mencius (c. 372–289 BCE) held that the heart-mind's office is to think. Thinking yields understanding, and failing to think leaves one without it. He also argued that four "sprouts" (compassion, shame, deference, and the sense of right and wrong) are innate predispositions of the heart-mind. Cultivating these sprouts develops the virtues of rén, yì, lǐ, and zhì. Mencius's notion of "seeking the lost heart-mind" describes learning as attentive recovery of one's own moral awareness. Later Neo-Confucian thinkers such as Zhu Xi and Wang Yangming developed these ideas further. Unlike Cartesian substance dualism, the Chinese term joins emotion, perception, and reasoning in a single organ.

Daoist awareness and attention

Daoist texts approach awareness through emptiness and stillness rather than through the ordering of mental contents. The Laozi (ch. 16) recommends reaching utmost emptiness and holding firmly to stillness. The Zhuangzi describes a "fasting of the heart-mind" (xīn zhāi), in which the perceiver receives things without preconception. Another passage, which compares the ideal person's mind to a mirror, pictures responsiveness without grasping or retaining what is reflected. The Zhuangzi also speaks of "sitting and forgetting" (zuòwàng), in which self-conscious thought dissolves. These passages are less a theory of a metaphysical soul than a set of practices for undivided attention. They offer a contrast with both the Upanishadic model of a self and the Buddhist analysis of momentary mental events.

Methods of study and their limits

From introspection to observable behavior

Consciousness became a laboratory subject in the 1870s and 1880s. Wilhelm Wundt set up his psychology laboratory at Leipzig in 1879, and his observers were trained to report their immediate sensory experience under repeatable conditions. Wundt's method was narrower than later accounts suggest. Its purpose was to measure reaction times and the judgments subjects made about external stimuli. His student Edward Titchener took a more analytic version to Cornell and named it structuralism, which aimed to break experience into elementary sensations and feelings. The method proved hard to standardize. Trained observers disagreed with one another, and laboratories produced conflicting results. The "imageless thought" controversy in the years around 1900 deepened doubts about whether introspective reports could be trusted as data.

In the early twentieth century the field turned toward observable behavior. John B. Watson argued that psychology should study overt behavior, which could be quantified, and that the mind was too vague to be a productive scientific object. Later research has partly reversed this position. Modern work does not abandon first-person reports, but it treats them as evidence that must be checked against other measures, and it uses behavioral and neural methods to study consciousness directly.

The problem of other minds

Consciousness is accessible in the first person, so any external study must infer experience from indirect signs. This is the problem of other minds in its experimental form. Three main kinds of evidence are used. Verbal reports ask subjects to say what they saw or felt. Behavior, such as button presses, eye movements, or accuracy on a task, shows what information was used. Physiological markers, such as pupil size, eye movements, and brain activity measured by EEG, fMRI, or intracranial recordings, can be compared across conditions. Each kind of evidence has a characteristic weakness. A report is only as good as the subject's ability to introspect and the researcher's ability to interpret it. Behavior can be produced by unconscious processing. Physiological signals may track attention, arousal, or the motor preparation for a response rather than experience itself. For this reason, most consciousness research compares matched conditions in which the stimulus is held constant and only the subject's awareness changes.

Report, forced choice, and no-report paradigms

Verbal report is the most direct method, but it has a well-known confound. Reporting requires the subject to attend to the stimulus, to retrieve a judgment, and to produce a response, and each of these processes may be linked to brain activity that has nothing to do with the experience itself. Researchers have therefore used forced-choice tasks. In these tasks a subject must guess which of several options was presented, even when they say they saw nothing. Above-chance accuracy without awareness is taken as evidence of unconscious processing. Forced choice avoids some of the problems of subjective reports, but it introduces its own: a subject who reports no experience may still have a partial, unreported awareness, and a low score may reflect a cautious criterion rather than the absence of information. Many studies therefore pair the discrimination task with a separate measure of confidence or perceptual awareness.

No-report paradigms were developed to remove the report altogether. In a typical design, researchers infer when a percept has switched from measures such as eye movements, rather than asking subjects to press a button. Supporters argue that this design separates the neural correlates of experience from the neural correlates of reporting. Critics argue that the method does not remove the problem. Attention, metacognitive reflection, and the mind's spontaneous shifts between percepts can still be included in the measured activity, and the no-report condition may itself contain confounds, such as differences in the stimulus. The debate is unresolved, and it shows that removing the report does not by itself establish that the measured signal is conscious.

Masking, binocular rivalry, and blindsight

Three experimental tools are used to dissociate conscious from unconscious processing, by holding the physical input roughly constant and changing whether it is seen. In backward masking, a briefly shown target is followed by a mask, which can make the target unseen while leaving it able to influence later responses. Continuous flash suppression and other rivalry-based methods present different images to the two eyes. Only one image is seen at a time, so perception alternates while the retinal input remains the same. Binocular rivalry has been widely used to study the neural basis of perception, but its underlying mechanisms are still not fully understood. Some researchers have argued that information about suppressed images is processed in a more fragile way than information about seen images, which is one proposed explanation for the differences found.

Blindsight provides a clinical version of the same dissociation. Patients with damage to the primary visual cortex report no visual experience in part of the visual field, yet they can often point to, discriminate, or respond to stimuli in that region when forced to guess. Studies of such patients show that visual information can guide behavior without reaching awareness, although they do not show exactly which pathways are responsible. Taken together, masking, rivalry, and blindsight provide the main evidence that experience and processing can come apart. Each still depends on the report or behavior that defines the awareness condition, so each inherits the methodological limits discussed above.

Neural correlates of consciousness

The search for the neural correlates of consciousness (NCC) asks which brain events are needed for a particular experience. The NCC are the minimum neural mechanisms sufficient for any one specific conscious percept. Koch and colleagues also separate these mechanisms from the processes that precede, accompany or follow an experience, which is a necessary distinction, since much brain activity around a percept is not the percept itself.

Francis Crick and Christof Koch made the search concrete. Instead of proposing a grand unified theory, they undertook a very specific research program focusing on the visual system, asking how light striking the retina becomes an integrated visual experience. Their early hypothesis drew on work in cat visual cortex and suggested that gamma-frequency oscillations might play a key role in the emergence of awareness. In 1995 they argued that neurons in V1 are not the neural correlates of visual consciousness. Later evidence has kept a narrower version of this point. Some information represented in V1, such as which eye a stimulus is presented to, is not available to consciousness, so V1 activity is necessary but not sufficient for conscious perception. By contrast, activity in extrastriate visual cortex correlates more closely with visual perception, and damage to these areas can selectively impair the ability to perceive particular features of stimuli. Crick and Koch also suggested that an NCC might involve a very small number of neurons (perhaps in the thousands) with certain distinctive properties. David Chalmers has treated this "start small" strategy as a sensible working method, though he notes that broad candidate systems and narrow ones can be hard to distinguish experimentally.

The cortical location of the NCC is contested. Proponents of posterior theories hold that consciousness depends mainly on posterior cortex, in what is called the "posterior hot-zone," which includes roughly the entire cortex minus the insula and the PFC. Prefrontal theories, including the global workspace and higher-order accounts, hold that some neural mechanisms in the PFC are responsible for rendering unconscious contents conscious. A central argument against the prefrontal side is that neural activity in the prefrontal cortex does not correlate with consciousness but with subjective reports. Supporting this, in one study almost no BOLD activity remained in the PFC in the no-report condition. Defenders reply that PFC activity during paradigms investigating conscious perception may be conflated with post-perceptual processes associated with reporting the contents of consciousness. Lesion evidence also cuts against a strict prefrontal requirement: reports of conscious patients after bilateral frontal lobectomy demonstrate that the frontal cortex is not necessary for consciousness. Other data point the other way. Stimulation of prefrontal cortex can jump-start consciousness in patients under anesthesia, so the question remains open.

The thalamus and brainstem are a separate line of evidence, drawn mainly from coma and arousal rather than from specific percepts. Damage to the brainstem tegmentum is known to cause coma, the most radical disturbance of consciousness. Classical models placed the ascending reticular activating system in the brainstem reticular formation, with discrete nuclei involved in maintaining wakefulness. The thalamus has been harder to pin down. In a 2018 stroke study, four patients with severely impaired arousal all had damage that extended beyond the thalamus into the hypothalamus and brainstem, while none of the patients with damage confined purely to the thalamus experienced severe impairment of arousal. A 2025 preprint, which has not undergone peer review, reported that concurrent lesions to both the brainstem and thalamus were necessary and sufficient for coma, and that cerebellar and cortical lesions, even when extensive, did not induce coma in the absence of the dual-brainstem and thalamic lesions. Researchers have proposed that the thalamus may matter more for attention and awareness than for basic arousal, a distinction that is still being tested. A proposed mesocircuit hypothesis offers a mechanistic model in which disrupting brainstem input and thalamic output together suppresses arousal and awareness. These data show which structures are required for a state of wakefulness, but they do not fully establish which structures generate the contents of experience.

Disorders of consciousness and clinical assessment

Coma, unresponsive wakefulness, and minimally conscious states

Clinicians separate two dimensions: alertness (wakefulness) and awareness. Coma is the state in which the patient is totally unresponsive, with the eyes staying closed. Within about two weeks of injury, many patients open their eyes, yet if they still don't respond to stimuli on purpose, they're considered to be in a vegetative state. The term "vegetative state" was introduced in 1972 and is now often replaced by unresponsive wakefulness syndrome (UWS). The diagnostic criterion is that UWS requires the presence of a sleep-wake cycle; eyes are open during wakefulness, but there is no evidence of conscious behavior. The minimally conscious state (MCS), added in 2002, requires intermittent evidence of behavioral awareness. Emergence from MCS requires functional interactive communication, or the ability to use objects appropriately, or both.

These distinctions carry prognostic weight. Detecting MCS matters because functional outcome is dramatically better for MCS patients. Yet the boundary is hard to draw at the bedside: upwards of ~40% of patients with a DOC can be misdiagnosed when specifically designed behavioral scales are not employed or improperly administered. Even with the Coma Recovery Scale-Revised, a unique assessment remains associated with a high frequency of diagnostic error.

Covert consciousness

The concept of covert consciousness describes patients who cannot produce behavioral responses but show neural evidence of awareness. The landmark case was reported by Adrian Owen and colleagues in 2006. A woman diagnosed as vegetative was scanned with fMRI five months after her accident while imagining scenarios such as playing tennis or walking through her home, tasks known to activate particular brain areas. Her activation was indistinguishable from that of a normal, healthy, awake individual. Owen himself stressed that the finding was at the earliest days and was unlikely to generalize to most long-standing vegetative cases. Later studies have tested the paradigm more broadly. One 2014 comparison at two scanner field strengths found that command-following activity could be detected in most participants at both 3T and 1.5T, with similar reliability, but the authors' own conclusion that fMRI does not provide a viable means of assessing covert awareness on a routine basis reflects the practical limits. A 2024 meta-analysis of ten studies reported pooled sensitivity and specificity of 0.71 for each, which is useful but far from definitive.

Perturbational measures

The perturbational complexity index (PCI) takes a different approach: it perturbs the cortex and measures the response, rather than waiting for voluntary behavior. The cortex is stimulated with transcranial magnetic stimulation, and the resulting spatiotemporal EEG responses are compressed to measure their algorithmic complexity, based on Lempel-Ziv compression. The index was introduced in 2013 by Marcello Massimini and colleagues as a practical application of principles from Integrated Information Theory.

Reported cutoffs differ between studies. One review gives values of 0.40 or higher as generally representing consciousness, whereas another reports when PCI < 0.3, participants were in slow wave sleep, under anesthesia or in UWS. In a sample of 24 patients, the individual maximum PCI value fell within the consciousness range in 11 MCS patients, yielding a sensitivity of 92%. Among UWS patients in the same study, four patients (36%) provided high-complexity PCI values, which might suggest a covert capacity for consciousness. A separate study of eight subacute patients found that they also showed high PCI values, which are compatible with a capacity for consciousness, although not expressed in behavior, and within less than three months, 7 patients recovered responsiveness either behaviourally or through active fMRI paradigms.

Clinical adoption is still limited. The method is time consuming and expensive to acquire, and advocates note that more such tools need to be incorporated in standard clinical practice. For now, PCI is best understood as a promising research and diagnostic-support tool rather than a settled clinical standard.

Ethical and legal questions

Uncertain diagnoses create difficult questions in both directions. A patient who is aware but classified as unaware may be denied rehabilitation or may have decisions made for them without their voice; a patient classified as aware who is not may be given false hope or exposed to inappropriate interventions. Diagnostic errors and false positives are therefore central concerns, which is why researchers have developed strategies for assessing false positive rates in covert consciousness testing. Decisions about life-sustaining treatment, surrogate decision-making, and the enrollment of nonresponsive patients in research all depend on how confidently a diagnosis can be made, and they raise the problem of consent when the patient cannot communicate. Legal frameworks differ across jurisdictions, and imaging evidence has not yet been fully incorporated into standard legal criteria. These questions connect clinical neuroscience with medicine and philosophy-of-mind, because the definition of "awareness" that clinicians use is itself a contested theoretical choice.

Global workspace theory

Global workspace theory (GWT) began as a cognitive model rather than a neuroscientific one. In A Cognitive Theory of Consciousness (1988), cognitive scientist Bernard Baars proposed that the mind contains many specialized processors that work in parallel and mostly unconsciously. A limited-capacity "workspace" sits among them. Content that wins access to this workspace is broadcast to the whole system, making it available to perception, memory, evaluation, reasoning and action. Baars illustrated the idea with a theater: consciousness is the spotlight on stage, and the audience of unconscious specialists receives whatever the spotlight picks out. The model drew on cognitive architectures from artificial intelligence research.

From cognitive model to neuronal theory

In 1998 Stanislas Dehaene, Claude Kerszberg and Jean-Pierre Changeux translated the model into a neural hypothesis, the global neuronal workspace (GNW). Their central claim was that a stimulus becomes conscious when it is made globally available to multiple brain systems through a network of neurons with long-range axons, densely distributed in prefrontal, parietal and cingulate cortices. Dehaene and colleagues divided local processors into categories such as perception, motor control, memory, value and attention, and proposed that these processors can be dynamically recruited into the workspace. Work by Dehaene, Jean-Pierre Changeux and Lionel Naccache over the following decade, and the 2011 review by Dehaene and Changeux, developed the model into the version most often called the neuronal global workspace theory, or GNWT.

Ignition and broadcast

The key mechanism in GNWT is "ignition". According to the theory, a representation that is weak or merely local fades out, whereas a sufficiently strong sensory input triggers a nonlinear, all-or-nothing transition. Activity suddenly spreads through recurrent loops linking the frontal and parietal cortex, and this amplification is what makes the content reportable. In the simulations of Dehaene and colleagues, ignition showed up as increased gamma-band oscillations in corticothalamic circuits and their synchrony across areas. Ignition can be triggered by an external stimulus, as part of a task, or it can occur spontaneously. Simulations of the resting brain produce a stream of such spontaneous ignitions, which the authors related to the wandering of thought and to phenomena such as inattentional blindness, where heavy stimulation by one event prevents access to another.

The theory's predicted signatures include a late, non-linear amplification of stimulus-specific activity, about 300 ms after onset, and a late positive EEG component known as P3b. These markers are said to distinguish accessed from unaccessed stimuli, and they are one reason the theory is considered testable. Its proponents have also linked conscious access to top-down attention, which amplifies bottom-up signals enough to sustain the workspace.

Limited capacity and flexible access

GWT accounts for two features of conscious experience that often puzzle other theories. The first is limited capacity: only one or a few coherent contents can dominate the workspace at a given moment, which is why people can usually attend to only one of several competing tasks. The second is flexible access: once content is broadcast, it can be used for verbal report, deliberate reasoning, memory encoding and voluntary action, regardless of its sensory modality. In this sense the theory aims to explain what philosophers call access consciousness, rather than the subjective quality of experience itself.

The disputed role of the prefrontal cortex

Early versions of GNWT put the prefrontal cortex at the core of the workspace, which made it a natural target for testing. Critics have argued that frontal activity may reflect report, rather than experience itself, and that posterior sensory areas can support conscious perception without frontal involvement. Proponents have responded that frontal activity is a necessary part of the mechanism that allows content to be used flexibly. The 2025 Cogitate study, described in a neighboring section, sharpened this dispute: the team could decode some conscious content from prefrontal activity but not all of it, and this result was taken as a difficulty for the theory's central claim. Researchers also reported that the expected ignition at stimulus onset was not clearly found, although the theory had predicted such activity at both the start and the end of a stimulus.

Overall, GWT remains one of the most influential scientific theories of consciousness, but its exact anatomical commitments, and the status of ignition as a necessary condition for experience, are still under debate.

Integrated information theory

Integrated information theory (IIT) was first proposed by Giulio Tononi in 2004. It holds that consciousness is identical to a certain kind of information, and that this information must be physically integrated, not merely processed in a functionally integrated way. It can be measured mathematically according to the phi metric. Unlike most theories, which begin from neural activity or behavior, IIT begins from the properties of experience itself. It starts from experience, identifies its essential properties (axioms), and then infers what kind of properties physical systems must have to account for them (postulates).

Axioms and postulates

The axiom list has changed across versions of the theory. The three axioms most often used to motivate phi are information, integration, and exclusion. For current purposes, we can focus on three axioms that primarily motivate Φ as a measure of consciousness, Information, Integration and Exclusion. The information axiom says that each experience distinguishes itself from countless other possible experiences. An experience specifies; it is specific to certain things, distinct from others. The integration axiom says that an experience forms a unified whole, which implies that the physical system must be irreducible: it cannot be split into independent parts without loss. Each experience is integrated, what philosophers call unitary. Whatever I am conscious of, I am conscious of as a whole. The exclusion axiom says that an experience has a definite boundary. Its contents are either within it or outside it, with nothing in between. Consciousness is exclusive: the 'canvas' has an exact border, and any individual quality is either part of that canvas or not. Tononi also treats the existence of experience, along with its intrinsic character, as self-evident, a point that critics later contested.

The postulates are the step from these phenomenal features to physical conditions. Each axiom is accounted for in terms of cause–effect power, translating a phenomenal property into an essential property of the physical substrate of consciousness. Because experience is unified, its substrate must be physically integrated. Because experience is exclusive, the conscious system must have a sharp physical boundary, which is fixed by the maximum of phi. The axioms do not deductively imply the postulates, but the postulates have more than mere statistical inductive support. The argument therefore rests on inference to the best explanation.

Structure rather than computation

A central and distinctive claim is that what matters is the physical causal structure of a system, not the function it performs. A system's input-output behavior, or the algorithm it implements, does not settle whether it is conscious. Its properties can only be accounted for by substrates with a suitable internal organization, regardless of their input-output behaviors or functions. This is why IIT is often contrasted with computational functionalism. In IIT's terms, the canonical elements of a network are logic gates, and the theory says nothing about signals or broadcasting between them. Canonically, the system elements are just logic gates. Crucially, IIT also says nothing about signals between elements.

Phi itself is defined as the information a system generates as a whole beyond what its parts generate separately. The quality of experience is then associated with the set of informational relationships generated within that complex. In principle, the theory is meant to say whether a given system has consciousness, how much it has, and which particular experience it is having. Based on the postulates, it permits in principle to derive, for any particular system of elements in a state, whether it has consciousness, how much, and which particular experience it is having. The theory predicts that phi tracks the level of consciousness, so that it falls in deep anesthesia and dreamless sleep. This prediction is indirectly supported by recent neuroimaging experiments that combine transcranial magnetic stimulation with electroencephalography.

Counterintuitive implications

The most debated consequences concern simple systems. Expander grids are two-dimensional lattices of logic gates, each connected to its nearest neighbours. An expander grid on which all of the logic gates are inactive has an unboundedly high Φ as the number of logic gates increases. Scott Aaronson used this kind of case to argue that the theory has an absurd consequence and concluded that it was untenable, while acknowledging its mathematical precision. After demonstrating that IIT implies the possibility that large grids could be highly conscious, he concluded that the theory was therefore untenable. Defenders have replied that high phi need not mean more consciousness. On a multidimensional reading, the inactive grid has only minimal consciousness, and its contents never change while it stays inactive. Taking the alternative multi-dimensional view leads to the less problematic conclusion that there is in fact minimal consciousness. The theory also implies that very simple devices may have some experience, and critics link this to a broader panpsychism. According to IIT, even simple systems can have some minimal form of consciousness, provided they have a certain degree of information integration.

Testability and criticism

Critics raise several objections. The first concerns measurement. Φ cannot be easily computed from large-scale data, making it difficult to test the theory. It can only be measured in simple model systems. The second concerns predictions. Some predictions have been checked only in an indirect and approximate manner, while others are in principle testable but technically demanding. Predictions about silent neurons illustrate the difficulty. The disabling prediction and the SB prediction are not testable. The third concerns the axioms. Cerullo argues that the main theoretical argument for the relevance of integrated information to consciousness is the principle of information exclusion. Yet, no justification is given to support this principle. Taken together, these objections describe a theory that is mathematically precise but hard to check against experiment, and whose foundational premises are contested.

Higher-order, recurrent, and predictive accounts

Higher-order theories

Higher-order theories all rest on the common observation that a mental state is conscious only if one is in some way aware of being in that state. The idea was developed independently in the mid-twentieth century by David Armstrong, David Rosenthal, and Peter Carruthers. Rosenthal's version, the higher-order-thought (HOT) theory, holds that the higher-order states responsible for conscious experience are distinct states from the lower-order states they are about. On this view, a first-order representation, such as a visual signal in sensory cortex, is nonconscious by definition and becomes conscious only when a second state takes it as its object. The HOT itself need not be conscious. That feature lets the theory explain why conscious experience seems direct: the mediating state is never itself noticed.

The account also changes how appearance and reality are understood. Critics have argued that consciousness cannot separate what seems to be the case from what is the case. Higher-order theorists reply that the higher-order state constitutes the reality of consciousness, while its content supplies the subjective appearance, so the distinction survives. Rosenthal's framework has been criticized for the regress worry (why the higher-order state is not itself in need of a further higher-order state) and for its reliance on conceptual capacities that many animals and infants may lack.

Recurrent processing theory

Victor Lamme proposes a different locus. He argues that the initial feedforward sweep of activity through the visual hierarchy can support sophisticated unconscious processing, but is insufficient for consciousness by itself. Experience appears only when feedback loops return signals to earlier areas and produce sustained, reverberating activity. In Lamme's version, the key feature is local recurrence within sensory cortex, so that localized feedback processing within perceptual cortices is taken to suffice, without long-range loops.

Lamme's graded model separates two kinds of consciousness. Local recurrent processing within sensory regions yields phenomenal awareness, while widespread recurrent processing involving frontoparietal networks yields access consciousness and reportability. This places Lamme in direct opposition to global workspace accounts, which hold that conscious access requires a more extensive architecture of reverberant loops, including frontal-parietal regions. Lamme suggests that frontal and parietal activity may serve mainly for post-experiential operations such as planning and verbal report. Critics respond that local recurrence may not be enough without some background conditions of the brain as a whole.

Predictive processing

Predictive processing reverses the usual picture of perception as bottom-up reception. The brain is treated as a hierarchical generator of hypotheses about the causes of its sensory input. The brain continuously generates predictions about the causes of sensory input, and incoming signals primarily serve to update prediction error. Perception is then the hypothesis that best reduces this error, weighted by how reliable each signal is taken to be.

Two prominent versions differ in emphasis. Andy Clark's work, developed at length in Surfing Uncertainty (2016), concentrates on perception of the external world and on cognition in general, treating action as a way of making predictions come true. Anil Seth, by contrast, turns inward, to the body, to emotion, and to consciousness itself. Seth's slogan of perception as a "controlled hallucination" rests on the claim that experience is generated from the inside out and only constrained by sensory input. In his later "beast machine" framework, the experience of being a self emerges from the brain's continuous effort to predict and control internal bodily states. Interoception, the brain's sense of heartbeat, breathing, and visceral states, becomes central to what it is to feel like a subject.

Embodied and enactive challenges

Enactive approaches, rooted in the work of Francisco Varela, Evan Thompson, and Eleanor Rosch in The Embodied Mind (1991), question whether consciousness can be fully explained by internal brain states. In this view, cognition arises through the organism's sensorimotor coupling with its environment, and experience is shaped by how the body acts and moves. Varela's later "neurophenomenology" (1996) proposed that careful first-person description should be treated as a source of data alongside neural recordings, rather than as a mere report to be explained. Enactivists argue that brain-centered theories treat the cortex as a representing machine whose outputs are later interpreted as experience, whereas experience may depend on the dynamics of the whole living system in its environment. Predictive processing theorists have drawn on enactive ideas, but they still generally place the explanatory weight on the brain's internal models.

Taken together, these accounts locate the difference between conscious and nonconscious processing in different places: a second-order representation, local feedback in sensory cortex, the structure of predictive inference, or the coupling of body and world. They also make different predictions about which neural signatures should appear when experience occurs, and about whether the frontal cortex is needed for experience or only for report.

Adversarial testing and empirical evidence

The Cogitate Consortium's 2025 study is the first large test of this kind, and it was designed to pit two theories against each other rather than add more evidence to each one separately. Its authors described it as an open science adversarial collaboration directly juxtaposing integrated information theory (IIT) and global neuronal workspace theory (GNWT) via a theory-neutral consortium. The proponents of both theories took part in fixing the experimental design, the divergent predictions, the expected outcomes and how the results would be read, and this was done in advance and preregistered. The experiment recruited 256 human participants who viewed suprathreshold stimuli for variable durations, while activity was recorded with fMRI, MEG and intracranial EEG. Preliminary results had first circulated as a preprint, and the final paper was published in Nature in 2025.

The two theories were turned into contrasting, testable forecasts. For conscious content, IIT expected maximal decoding of conscious content in posterior brain areas, whereas GNWT stressed a necessary role for the PFC. For the maintenance of experience over time, IIT held that content is actively sustained in the posterior cortex, while GNWT predicted brief content-specific ignition (approximately 0.3–0.5 s) in the PFC at stimulus onset and offset, with content stored in a non-conscious silent state between these events. The "ignition" idea is treated in detail in the global workspace section of this entry, and the posterior "hot zone" account belongs to the IIT section.

The data gave partial support to each theory. 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. Frontal regions carried some of the relevant information, but researchers were not able to decode all aspects of experience from frontal areas. For IIT, the sustained activity it anticipated appeared in some form, but the study found no sustained synchronization between early and mid-level visual areas located in the posterior part of the brain, which was the specific pattern the theory required.

Because the predictions were registered beforehand, the results were not easy to explain away, and they counted against both theories. The authors summarized the outcome by saying that these results align with some predictions of IIT and GNWT, while substantially challenging key tenets of both theories. For GNWT, the main problem was that the general lack of ignition at stimulus offset was not seen, and the prefrontal representation of some conscious dimensions was weaker than expected. For IIT, a lack of sustained synchronization within the posterior cortex contradicts the claim that network connectivity specifies consciousness. The IIT proponents' commentary on the study reads the same data more favourably, arguing that the findings corroborate IIT's overall claim that posterior cortical areas are sufficient for consciousness, and neither the involvement of PFC nor global broadcasting are necessary. The disagreement over how to interpret a failed prediction is one of the main lessons of the study.

The study also highlights methodological issues. The first achievement was getting rival theorists to agree on what the theories would predict, and this was difficult because both the global workspace and integrated information theories are framed in very abstract terms. Preregistration made the outcome conditions explicit: the proponents of IIT and GNW agreed in advance what findings would falsify their own theories. Even so, neither theory was simply refuted. Those who designed the experiment have to decide which measurements, thresholds and anatomical regions count as a fair test, and the results show how much depends on these choices. The episode suggests that large, theory-neutral adversarial collaborations can make consciousness theories more precise and more open to disconfirmation, but that a single failed prediction seldom ends a research programme on its own.

The hard problem and philosophical positions

David Chalmers introduced the distinction in a 1994 talk in Tucson and developed it in his 1995 paper "Facing Up to the Problem of Consciousness" and his 1996 book The Conscious Mind. He divided the questions about mind into two groups. The "easy" problems concern functions: how a system discriminates and categorizes stimuli, integrates information, makes its states reportable, focuses attention, controls behavior, and shifts between wakefulness and sleep. Each can in principle be solved by identifying a mechanism that performs the function. The hard problem asks why any of this processing is accompanied by experience at all, that is, why there is something it is like to be in a given state. Chalmers held that even a complete functional account would leave an explanatory gap between physical description and experience. The phrase itself is usually credited to Joseph Levine.

Positions on the mind–body relation

Physicalism holds that mental states are physical or wholly grounded in the physical, and it remains the default strategy in philosophy-of-mind. On this view the easy problems look tractable, and the hard problem is either a misdescription or a challenge still awaiting an answer. Substance dualism, most closely associated with Descartes, treats mind and matter as two distinct kinds of substance. Property dualism keeps a single physical substance but insists that phenomenal properties cannot be reduced to physical ones. Chalmers has defended a version of this position, which is why he is often described as a dualist. Panpsychism takes a different route. It treats some form of experience, or a proto-experiential property, as a basic and widespread feature of matter, so the gap is not crossed but avoided, because experience is never generated from non-experiential parts. Each answer has a recognized cost. Physicalists must either explain away the apparent gap or show how it closes. Dualists must say how non-physical properties relate causally to the brain. Panpsychists must explain how micro-experiences combine into a single subject, a problem often called the combination problem.

Illusionism

Illusionism, defended by Daniel Dennett and Keith Frankish, rejects the assumption on which the hard problem rests. Its proponents argue that phenomenal properties, as they seem from the inside (intrinsically private, ineffable, and directly known), are not what they appear to be. Introspection misrepresents mental states, and the conviction that a further fact remains after all functions are explained is itself a cognitive product that the science of the easy problems can account for. If this is right, nothing is left over once the mechanisms are described. Critics reply that illusionism seems to deny the very data it needs to explain, since an illusion of experience already presupposes that something is experienced by someone.

Zombies and Mary's room

Chalmers's zombie argument asks readers to imagine a being physically identical to a person but with no experience at all. If such a being is conceivable, he argues, it is possible, and physicalism is false. Many philosophers accept that the hard problem depends on the first step, the claim that conceivability tracks possibility, and dispute the second. Zombies therefore function as a thought-experiment that tests whether a complete physical description leaves anything out.

Frank Jackson's knowledge argument was first published as "Epiphenomenal Qualia" in 1982 and extended in "What Mary Didn't Know" in 1986. Mary is a scientist who learns every physical fact about color vision while confined to a black-and-white room. When she first sees a red tomato, she seems to learn something new, and if so, the physical facts were not the whole truth. In 1982 Jackson drew a further conclusion, that qualia are epiphenomenal, meaning causally inert, because he accepted the physical completeness of physiology. Responses have taken several forms. The ability hypothesis, associated with Nemirow and Lewis, holds that Mary gains a practical skill, such as recognizing or imagining red, rather than a new fact. The acquaintance hypothesis, defended by Conee, holds that she gains a new mode of acquaintance with a quale while the propositional content stays the same. Others question whether the case is coherent. Jackson himself later moved toward a representational account and away from the anti-physicalist conclusion. The argument continues to test whether a physical explanation of experience can be complete.

Unity, binding, and the self

Experience seems to come as a single field. A visual scene, a sound, a bodily feeling and a thought are present together, and the subject of that experience seems to be one person. Philosophers call this the unity of consciousness. Neuroscience has to explain how distributed processing produces that unity, a problem known as the binding problem. Split-brain studies and theories of the self then ask what the unity is unity of, and whether the sense of being a single subject is itself a product of brain processes.

The binding problem and neural synchrony

The binding problem covers two questions. The first is how features processed in different places, such as colour in one area and motion in another, become one perceived object. The second is how these bound contents join into an integrated experience at all. Researchers often separate the first, which is a question about neural coordination, from the second, which is about why such coordination should come with experience. The problem is open partly because no complete model exists.

The most influential proposal is neural synchrony. Francis Crick and Christof Koch suggested that the coordinated firing of neurons in visual cortex, in the range of roughly 40 to 70 Hz, could link the features of a single object. Early experiments in cats and monkeys reported such synchronized oscillations, but critics later noted that similar synchrony appears in anaesthetized as well as awake animals, and that it may not sustain binding for stationary stimuli. Others argue that synchrony may be necessary for conscious unity without being sufficient, or that it is a correlate of binding rather than its mechanism. Attention also appears to matter. When attention is overloaded, people may detect the right features, such as red and blue, yet mis-assign them to the right shapes, suggesting that the binding of features depends on attentional selection as well as oscillatory coordination.

Split-brain studies

To treat severe epilepsy, some patients had the corpus callosum, the main bundle of fibres joining the two cerebral hemispheres, surgically cut. In the 1960s, Roger Sperry, who received the Nobel Prize in 1981, and his student Michael Gazzaniga showed that stimuli presented to one visual half-field could be processed by one hemisphere without the other gaining access to them. When a patient's speaking left hemisphere had no information about a picture shown to the right hemisphere, the patient could still choose a related object with the left hand, and then produce a fluent explanation for that choice. Gazzaniga proposed that the left hemisphere contains an "interpreter" that constructs plausible causal narratives from the information available to it, and he argued that this mechanism operates in intact brains too, where interhemispheric communication normally hides its errors.

These findings bear on unity and self-report in two ways. First, split-brain patients usually report an unbroken sense of self, which suggests that a single subjective perspective can be maintained by a verbal system that has access to only part of the information. Second, the confident explanations given by patients show that verbal reports about one's own mental causes can be confabulated. The interpretation is contested. Later work showed that the two hemispheres are not fully independent, that some information transfer persists through subcortical pathways, and that the "interpreter" account goes beyond what the original experiments established. Even so, the studies remain a central case in which the apparent unity of consciousness and the apparent reliability of self-report come apart.

Metzinger's self-model theory

Thomas Metzinger has argued that the self is not a substance that has experiences but a content of experience. In his Self-Model Theory of Subjectivity, the brain constructs a phenomenal self-model (PSM), an organism-centred representation that is experienced from the inside as "me". Metzinger identifies three phenomenal properties that such a model must have for the self to appear. One is mineness, the feeling that particular states belong to me. A second is a transparency constraint: the model is experienced as direct reality rather than as a representation, because the processing stages that produced it are not themselves available to attention. Because of this transparency, the self seems to be a thing rather than a model.

Metzinger's collaborator Olaf Blanke and he later defined a more basic form, minimal phenomenal selfhood. Its three conditions are global identification with the body as a whole, spatiotemporal self-location, and a weak first-person perspective. The account draws on phenomena such as out-of-body experiences and body-ownership illusions, in which people can be made to feel that a rubber hand or a virtual body is their own. These cases suggest that body ownership and self-location are constructed and can be dissociated. Critics respond that the theory risks describing the self-model while leaving open why there is a subject to whom the model appears, and that it can slide from explaining the sense of self to denying the self.

The sense of agency

The sense of agency is the experience of being the author of one's own actions. It is usually studied separately from the sense of ownership of the body, though the two are related. Experiments since the 1980s have examined this feeling. Benjamin Libet's studies of readiness potentials and reported timing of conscious intention prompted debate about when the experience of choosing occurs relative to neural preparation. Later work on intentional binding suggests that voluntary actions and their outcomes are perceived as closer in time than they are, which some take as a marker of agency. Researchers also distinguish the feeling of control in a particular action from broader judgements of responsibility. How these experiments should be interpreted, and whether they bear on questions of free will, remains disputed, and the empirical findings can be described without settling that dispute.

Sleep, dreaming, anesthesia, and altered states

REM sleep and reportable dreaming

Before the 1950s, sleep was widely treated as a passive, coma-like state. That view changed in 1953, when Eugene Aserinsky and Nathaniel Kleitman published their observation that the eyes of sleeping adults made rapid, jerky movements in recurring periods. Aserinsky had first noticed the pattern while recording the eyes of sleepers, and the pair then woke subjects during these periods and found that they often reported dreams. Kleitman's later student William Dement, working with Kleitman, linked the eye movements to specific EEG patterns and to dream reports, which gave dream research an objective marker. The stage came to be called REM (rapid eye movement) sleep. It is accompanied by irregular breathing and heart rate and by a cortical EEG pattern resembling waking, which is why sleep came to be seen as a set of distinct stages rather than one uniform state. Dreams are not confined to REM sleep, and dreams are most reliably and vividly reported when subjects are woken from it. Its exact function and the status of dreaming as a form of experience remain debated.

Anesthesia and the loss of consciousness

General anesthesia is one of the most practically important tools for studying consciousness, but the way anesthetic drugs switch experience off is still only partly understood. Different agents act on different molecular targets. Many common drugs, such as propofol and volatile agents, enhance inhibitory GABA-A receptors, whereas ketamine acts mainly on NMDA receptors and can produce vivid, dream-like experiences while the patient remains outwardly unresponsive. Recent work has mapped binding sites on receptors and ion channels and has shown that loss of consciousness depends on network-level changes, not just on one molecular effect. Proposed mechanisms include the disruption of thalamocortical loops, the suppression of arousal systems such as the ascending reticular activating system, and reduced long-range connectivity in cortex. A dose-dependent account suggests that higher-order associative cortex is affected before the arousal systems, which could produce a state of "wakeful unawareness." Studies of propofol-induced unconsciousness have found that auditory responses are suppressed beyond primary auditory cortex, which challenges simple accounts that attribute the effect only to thalamic gating. Emergence from anesthesia is studied in parallel, because the return of consciousness may require more neural resources than maintaining it.

Psychedelic states and meditation as sources of data

Psychedelic drugs, especially psilocybin and LSD, produce large changes in perception, thought, and sense of self, and they allow researchers to study how the content and structure of experience can vary within the same brain. Neuroimaging work led by Robin Carhart-Harris proposed the entropic brain hypothesis in 2014, which holds that the richness of conscious content tracks the entropy of spontaneous brain activity, with normal waking consciousness constrained below a critical point and psychedelics raising entropy. The theory also linked psychedelic states to reduced organization in the default mode network. Later reviews have found support for elevated entropy under psychedelics but have also questioned whether brain entropy is a reliable biomarker of the psychedelic state, so the idea remains a hypothesis. Meditation provides a complementary source of data. Practitioners in Buddhist and other contemplative traditions have trained attention, monitoring, and the observation of mental events, and neuroscientists have used such training to study attention, self-referential processing, and the variability of awareness. Reports from these states are valuable but are shaped by training, expectation, and the difficulty of verifying introspective claims.

Dreams, lucid dreaming, and the boundaries of experience

Dreams and lucid dreams are useful because they test where experience begins and ends. In lucid dreaming, the dreamer becomes aware that they are dreaming while still asleep. Researchers have used pre-arranged voluntary eye movements to signal from within REM sleep, which shows that some reflective awareness can persist during sleep and can be communicated to the outside. This offers a test case for theories that link experience to frontal monitoring or to reportability. Dreamless sleep, sleep inertia upon waking, and the states reported in anesthesia are all cases where reports and brain measures may diverge. Some anesthetic states appear to involve experience without perception of the environment, and such experiences are rarely remembered afterward, so their absence from memory does not show their absence from experience. These cases make it harder to decide when consciousness is present, especially in infants, patients, and other people who cannot give reports. They also show why theories must specify what counts as evidence for experience, rather than relying only on what people later say about it.

Evolution and function of consciousness

The evolutionary question is simple to state and hard to answer: if an organism can detect, categorize, and act on information without any felt experience, what does experience add? Much of human perception already works this way. Face recognition, including judgments of familiarity, emotion, and gaze direction, can run without awareness, so we feel in charge even when the nonconscious does most of the work. The puzzle is therefore not whether unconscious processing is enough, but why natural selection would ever have produced a second, experiential layer on top of it. Evolutionary accounts usually start from this point. One view holds that consciousness has a function that raises fitness. The rival view holds that consciousness itself has no function and is a byproduct of other brain properties that do have an evolutionary function.

Functional accounts

The most common answer is that consciousness makes behavior more flexible. A review of the field notes that there is widespread agreement that consciousness serves to 'increase flexibility' of cognition, a shared position sometimes called the "integration consensus." Other proposals are more specific. Some hold that consciousness creates a unified, coherent representation of incoming information, others that it enables the learning of sensory and cognitive representations, and others that it supports complex, flexible decisions. A further line stresses timing: rather than broad flexibility, consciousness may decide between competing options in the moment, depending on current context and current needs. Feeling-based accounts add a mechanism. Felt states can work as a common currency for weighing various options and as motivation for carrying out the chosen action. An important caveat is that conscious processing could in principle evolve without feelings, but it would still need some advantage, such as the weighing of alternatives, to be selected.

Nicholas Humphrey and social cognition

Nicholas Humphrey's work is one of the best-known attempts to tie consciousness to social life. In his 1976 essay "The Social Function of Intellect," he argued that the advanced cognitive abilities of primates evolved primarily to deal with complex social environments rather than to solve purely ecological problems. In later work he extended the idea from intelligence to phenomenal experience itself. A review of his book Soul Dust reports his claim that felt sensation was shaped by natural selection, not simply tolerated by it, because it enriched how organisms understand themselves and others. In a 2024 special issue of the Philosophical Transactions of the Royal Society B he offered a more mechanistic model, in which reflexive responses are internalized and "virtualized" through efferent copies. On this model, phenomenal consciousness is restricted to homeothermic vertebrates, namely birds and mammals. A related proposal, presented in the same special issue, is that when behavioural flexibility emerges, consciousness serves the social function of identifying and responding to the behaviour of social partners. These accounts differ in detail, but they share the idea that the demands of living with others created the pressure for experience.

The epiphenomenalism objection

The most serious challenge to all functional accounts is epiphenomenalism, the view that experience exists but has no causal effect on behavior. If this were true, natural selection couldn't favour consciousness directly, and experience would only be a side effect of neural processes that were selected for other reasons. Defenders of functional accounts reply that if consciousness evolved through selection, it must have had effects that selection could act on. Critics answer that our sense of having a function may itself be misleading. The question is open because the causal role of experience is exactly what remains disputed in the philosophy of mind.

Limits of evolutionary explanation

Evolutionary explanations face several practical limits. First, the fossil and comparative record contains no direct trace of experience, so claims about when it appeared depend on inference from behavior and neuroanatomy. One model dates the first feature-like precursors to early amniotes roughly 320 million years ago, with feelings introduced as a strategy for behavioral decisions. Second, comparing species is difficult because we tend to look for advanced performance when deciding whether another animal is conscious. Critics point out that cephalopods have highly sophisticated behavior, but it is no more sophisticated than what our nonconscious brains handle, so behavioral sophistication cannot settle the question alone. Third, a trait can be adaptive without being the target of selection, and some traits persist because they are not harmful, which means a clear fitness advantage is not guaranteed. Evolutionary stories are therefore best treated as hypotheses that must be tested against independent evidence from animal studies and from neuroscience, rather than as settled explanations of why experience exists.

Quantum and other speculative theories

The most developed quantum proposal is the orchestrated objective reduction (Orch OR) model, which Roger Penrose and Stuart Hameroff built in the 1990s. Penrose argued that gravity-related collapse of the quantum wavefunction is a real physical process that cannot be computed, and that it might underlie non-computable aspects of thought. Penrose chose to connect his view of non-computable consciousness with his views on the foundation of quantum mechanics: objective reduction (OR) postulates that quantum gravitational interaction at the Planck scale will inevitably make quantum wavefunctions collapse physically. Hameroff, an anesthesiologist, supplied the biological site. In their account, aspects of quantum coherence and a self-collapse of the wave function occur in cytoskeletal microtubules and other structures within each of the brain's neurons. Coherence is said to persist until the differences in mass-energy distribution among superpositioned tubulin states reaches a threshold related to quantum gravity. Each collapse is then taken to be a discrete event, and sequences of such events create a flow of time, and consciousness. The model also proposes that microtubule-associated proteins "tune" quantum oscillations, and "orchestrate" possible collapse outcomes. Its proposed timing is loosely tied to neural rhythms: the theory identifies discrete conscious moments with quantum computations in microtubules inside brain neurons, e.g., 40/s in concert with gamma synchrony EEG.

The main objection concerns decoherence, the rapid loss of quantum coherence through interaction with a warm environment. The brain is a warm, wet, and noisy environment, which is generally considered hostile to the delicate quantum states required for computation. In 2000, Max Tegmark estimated that orthodox mechanisms of decoherence would collapse a microtubule-associated quantum state on a timescale of the order of 10^(-13) seconds, and that superpositions of neurons firing would last only about 10^(-20) seconds. Relevant cognitive processes unfold over roughly 10^-3 to 10^-1 seconds, so the gap is many orders of magnitude. Proponents responded that Tegmark had not used Orch OR's own assumptions. Hagan, Hameroff and Tuszynski recalculated with a much smaller superposition separation and a dipole rather than charge model, which extended the calculated MT decoherence time to 10^-5 to 10^-4 sec, and they argued that shielding by counter-ions and an actin gel could push this further extends calculated decoherence time to hundreds of milliseconds. Critics also noted that Tegmark's formulation gives decoherence times that increase with temperature, which runs against ordinary physical intuition, and McKemmish and colleagues in 2009 concurred with (Tegmark, 2000) conclusions criticizing the much longer decoherence timescales. A review of microtubule biology has also argued that the tubulins do not possess essential properties required for the Orch OR proposal, as originally proposed, to hold.

Direct experimental support remains thin and contested. A 2024 study in rats by Michael Wiest and colleagues reported that when they gave rats a drug that binds to microtubules, it took the rats significantly longer to fall unconscious under an anesthetic gas. The authors read this as support for a quantum account, because "Since we don't know of another (i.e., classical) way that anesthetic binding to microtubules would generally reduce brain activity and cause unconsciousness," Wiest says, "this finding supports the quantum model of consciousness." Yet the result is open to a classical reading. Anesthetics are currently believed to cause unconsciousness by acting on one or more molecular targets including neural ion channels, receptors, mitochondria, synaptic proteins, and cytoskeletal proteins. Showing that one target is involved does not by itself show that quantum effects are. Other work by Hameroff's group, such as a 2022 preprint on Electronic Energy Migration in Microtubules, has also been presented as evidence, but such findings have not yet been independently replicated to the satisfaction of the wider field.

The interest in quantum interpretations of mind is partly tied to the measurement problem in quantum mechanics. Standard formulations do not fully specify when a superposition of states yields a single definite outcome. Penrose's proposal treats collapse as a physical process, while many-worlds and other interpretations avoid collapse altogether. Tegmark's critique explicitly drew on the latter line, arguing that the original motivations for the modifications to the Schrödinger equation implicit in a reduction postulate are obviated by the Everett relative state picture. Because consciousness is entangled in older discussions of observation, some physicists, notably von Neumann and Wigner, placed the observer's mind at the point of collapse. This historical link is a motivation for quantum theories of mind, but it does not supply evidence that the brain actually exploits quantum effects.

Most researchers treat quantum theories as speculative for several reasons: decoherence estimates still count against them under standard assumptions, the relevant experiments have not isolated quantum effects from classical ones, and the theories are not yet tied to testable predictions that rival models do not also make. Recent work has narrowed the debate rather than closing it. A 2026 analysis argued that Tegmark's Markovian bound no longer provides an upper limit in this regime and that mesoscopic quantum effects cannot be excluded on first principles, while also stating that this does not establish the existence of functional quantum coherence in microtubules. The question therefore remains open, but the burden of showing a functional role for quantum processes in consciousness has not yet been met.

Machine consciousness and artificial intelligence

Alan Turing's 1950 paper "Computing Machinery and Intelligence" began with the question "Can machines think?" and then replaced it with a more tractable one. In his imitation game, an interrogator exchanges typed messages with a hidden human and a hidden machine and tries to tell which is which. Turing proposed that a digital computer which could reliably mislead interrogators would deserve to be called a thinker. He predicted that by about 2000 an average interrogator would have no more than a 70 per cent chance of making the right identification after five minutes of questioning. The test is built on conversational performance, so it deliberately avoids the question of inner experience. It can show that a machine behaves intelligently; it cannot show that the machine feels anything. Turing's response to critics who demanded felt experience was to note that we attribute minds to other people on the basis of their behaviour, and that the same kind of evidence is all we have for them.

John Searle's Chinese room argument, published in 1980, targets the claim that running the right program is sufficient for understanding. Searle asks us to imagine a person who knows no Chinese, sitting in a room with an English rulebook for manipulating Chinese characters. Written questions go in, and the rulebook yields written answers indistinguishable from those of a native speaker. Searle argues that the person still understands nothing, because the rules operate on symbols by their form alone (syntax) and supply no meaning (semantics). The most common reply, which Searle himself calls the systems reply, concedes that the person does not understand Chinese but holds that the whole system, including rulebook, notes and database, does. Searle answers that the person could memorise everything and perform every step in his head, and still would not understand Chinese. The robot reply proposes embedding a program in a body with sensors and actuators; Searle responds that the robot would still be manipulating symbols derived from sensory input without knowing what they represent. Other replies include the brain-simulator reply, the combination reply and the other-minds reply, which argues that if we credit humans with understanding on the basis of their behaviour, consistency requires the same for machines. Searle's rejoinders to these have been contested, and the debate has continued for decades. Whatever its verdict on understanding, the argument is usually read as a challenge to computational functionalism rather than as a proof about consciousness as such.

The most systematic recent attempt to assess machines for consciousness is the 2023 report by Patrick Butlin, Robert Long, Yoshua Bengio and a group of co-authors. Their method has three steps: select neuroscientific theories that merit serious credence, derive from each a set of "indicator properties" stated in computational terms, and then check whether existing or future AI systems have those properties. The theories they drew on included recurrent processing theory, global workspace theory, higher-order theories, predictive processing and attention schema theory. Their conclusion was that no current AI system is conscious, but that there are no obvious technical barriers to building systems that satisfy the indicators, while satisfying them would not guarantee consciousness. A 2025 follow-up in Trends in Cognitive Sciences stressed updating credences under uncertainty rather than declaring detection.

Applying these checklists to large language models is much harder than applying them to the architectures the theories were designed for. Several indicators assume recurrent processing, a broadcast workspace, a body that is modelled in interaction with the world, or agency that persists over time, and a transformer that generates text token by token fits some of these only ambiguously. Behavioural evidence is also compromised. Such models are trained on enormous quantities of human writing about experience, so their reports of feeling are weak evidence on their own, and a system that mirrors human talk about minds may look conscious without being so. Empirical studies of consciousness-related capabilities in language models are possible, but the authors of the indicator work judged that they do not provide strong evidence that current models possess indicator properties. The wider debate remains open. Some researchers hold that functional organisation is what matters and that suitably built systems could have experiences. Others argue that consciousness depends on biological processes or embodiment, or that the concept is too poorly defined to apply to these systems yet.

Animal consciousness and moral status

For most of the twentieth century, the question of whether non-human animals have experiences was argued mainly from philosophical premises. A more formal scientific position emerged in 2012. The Cambridge Declaration on Consciousness held that humans are not unique in possessing the neurological substrates that generate consciousness. It named all mammals and birds, and many other creatures, including octopuses, as also having those substrates. The declaration also stated that the absence of a neocortex does not preclude an organism from experiencing affective states, which shifted attention toward subcortical systems. Its wording is narrower than it is often reported. One critic notes that the declaration seems to say that we have no scientific reason to suspect that nonhuman animals are less conscious than humans, rather than that animals are conscious. Its authors pointed to evidence from birds in particular, where birds appear to offer, in their behavior, neurophysiology, and neuroanatomy a striking case of parallel evolution of consciousness.

Behavioural tests and their limits

The most famous behavioural test is the mirror test. It was developed in 1970 by American psychologist Gordon Gallup Jr. In its standard form, an animal is anesthetized and then marked on an area of the body the animal normally cannot see. If, when the animal recovers from the anesthetic, it is given access to a mirror. If it subsequently touches or examines the mark on its own body, this behavior is interpreted as evidence that the animal recognizes its reflection as an image of itself. Great apes pass the test, and more recent research suggesting other large-brained mammals, such as dolphins and elephants, may also share this ability.

The test is contested on several grounds. A review of 89 studies on 62 species from 17 orders found conceptual, methodological and taxonomic biases, and noted that large-brained species are overrepresented, limiting the generalizability of results. A failure is also hard to interpret: if the animal does not make any attempts to remove the mark, perhaps owing to species' differences in motivation to interact with mirrors or to remove marks placed on their bodies, the interpretation of an absence of evidence is problematic. Dogs are a case in point. Since dogs and cats are primarily olfactory species, researchers have built scent-based versions of the test, and in one experiment dogs spent significantly more time investigating the altered scent. The fish case is similar: a cleaner wrasse that responded to a mark in a mirror led one researcher to argue that moving your tail before it gets stepped on, or scraping a parasite off your scales, isn't the same as sitting and pondering your place in the universe. Researchers also distinguish the two concepts. Mirror self-recognition only represents basic and not full-blown self-awareness.

For this reason, many assessments rely on a broader pattern of evidence: learning, flexible decisions, and responses that resemble pain. In the 2021 review commissioned for the UK, the criteria included nociceptors, integrative brain regions, self-protective behavior, associative learning, motivational trade-offs, flexible self-protection, and responses to anesthetics or analgesics. The authors reported very strong evidence of sentience in octopods and substantial evidence for other coleoid cephalopods. Evidence for fish has a longer history of dispute. One recent commentary notes that the debate over whether fish feel pain is instructive. Decades of careful behavioral and physiological work have shifted intuitions and policy without settling the underlying question, because that question cannot be settled by third-person evidence alone.

Legal recognition

Policy has moved faster than consensus in some jurisdictions. The review by Jonathan Birch's group led the UK to recognize cephalopods and decapod crustaceans as sentient. The original bill covered only vertebrates, but in response to Birch's report, the Act was amended to include cephalopods and decapods. The resulting Animal Welfare (Sentience) Act 2022 provides legal recognition that vertebrate animals, decapod crustaceans and cephalopod molluscs are sentient beings, with the capacity to experience pain, distress and/or harm. The Act also requires policy makers to consider animal sentience when creating or changing laws. The practical effects are still being worked out. The Home Office has been exploring the implications of this for the Animals (Scientific Procedures) Act, since that statute has covered vertebrates and cephalopods but not decapods. Recognition has also been used in arguments about how crustaceans should be killed. Commentators drew on the London School of Economics review that underpinned the original sentience recognition, which identified humane killing methods for decapods as an argument for regulatory guidance.

Uncertainty and precaution

The evidence is graded rather than binary. The 2021 review did not assign each taxon a fixed measure of sentience. Instead, it explicitly acknowledged uneven evidence across taxa, and true crabs met more criteria with higher confidence than several less-studied decapod groups, so evidence about one taxon should not be generalized mechanically to every crab, lobster, crayfish, prawn, or shrimp species.

Because subjective experience cannot be observed directly, uncertainty is built into every case, including human cases where self-report is available. This raises a precautionary question: how much weight should a plausible but unconfirmed possibility of suffering receive when a decision affects an animal? One answer is that the cost of wrongly denying experience to a sentient being can be severe, while the cost of extending some protection to a creature that turns out not to be sentient is usually smaller. Critics respond that precaution can be applied too widely, and that indicators such as behaviour and neural structure are themselves interpreted through human assumptions. Current policy generally sits between these positions, and law tends to lag behind scientific opinion.

Consciousness in literature and the arts

From psychology to the page

The phrase began in psychology rather than fiction. In The Principles of Psychology (1890), William James argued that thought does not present itself as separate pieces, describing it instead as a continuous current. He described the mind not as a chain of linked ideas, but as a continuous, flowing current. James's own wording, that consciousness does not appear to itself chopped up in bits, captures the break with the earlier associationist model of discrete ideas. Within a few decades, critics were applying the label to fiction that tries to set down a character's mental processes on the page. The word has kept two senses: a psychological one about the continuity of experience, and a literary one about a narrative mode. The literary term only borrowed its fluidity for artistic expression.

Joyce, Woolf, and the novel

The technique was developed in the early twentieth century. Interior monologue was pioneered by James Joyce in Ulysses (1922) and further developed by Virginia Woolf in Mrs Dalloway (1925). In Joyce's novel, the reader is drawn into the flow of a character's thoughts and memories rather than observing them from outside. Woolf is particularly renowned for the stream of consciousness she learned and developed from James Joyce, but her method differs from his, and her interest in Sigmund Freud's ideas about the mind also shaped it. Her aim, as one reading of her work puts it, was to render the psyche or soul of her characters, with the external world mattering mainly insofar as it affects inner life.

Techniques for rendering inner life

Several narrative devices are usually grouped under this heading. In first-person interior monologue, associations of thought are presented from the first-person point of view, as a direct quotation. A third-person variant, sometimes called indirect interior monologue, keeps the character's perspective while the narrator remains in the background.

Free indirect discourse blurs the boundary between the two. It merges a character's voice with the narrator's voice, reporting thoughts or speech without quotation marks and without "she thought." Woolf uses this device to move between minds within a single scene: she constructs the narrative by jumping suddenly from one character's thoughts to another's within the same spatial scene. In Mrs Dalloway, Woolf blurs the distinction between direct and indirect speech, freely alternating between omniscient description, indirect interior monologue, and soliloquy.

Fragmented time is the third device. Mrs Dalloway covers a single day in London in June 1923, and the story moves between past memories and present impressions. Joyce's A Portrait of the Artist as a Young Man takes a different route, with prose whose style tracks the perceptions of a growing child.

Influence on vocabulary

Literary depictions of consciousness have fed back into the everyday and academic vocabulary of the mind. Dictionaries now list the phrase with both a psychological and a literary sense, which means the novelists' usage has become part of the standard way people talk about thought. Phrases like "a stream of thoughts" now serve ordinary speech as well as psychology. The influence runs both ways, since the novelists drew on James and Freud, and their practice gave critics new terms for describing mental life.

Limits as evidence

Literary renderings are authored constructs, and they should not be taken as direct data about experience. Woolf's own narration includes authorial interjections to guide the reader and shape the narrative, so the apparent immediacy of a mind on the page is produced by choices. Critical descriptions such as an "all-embracing consciousness" that merges the internal and external worlds of her characters are interpretations, not measurements. A novel can suggest hypotheses about how thought moves, such as its shifts, gaps, and associative leaps, which James himself described. It cannot confirm them, because a convincing fictional mind is not evidence that it resembles a real one, and literary form also reflects the psychological theories of its era. For these reasons, literature is most useful to the philosophy of mind and psychology as a source of vocabulary and questions, not as an independent test of them.

Open questions and future research

Unresolved foundations

The most basic question is whether the hard problem must eventually be solved, meaning explained why any physical process is accompanied by experience, or whether it can be dissolved by showing that it rests on a misframed question, as illusionists claim. Neither outcome has been established, and the answer also shapes which kinds of evidence count. A related problem is theory choice. A 2022 review co-written by Anil Seth catalogued 22 theories grounded in brain biology, while a broader survey in 2024 reached more than 200 candidates, In 2022, British neuroscientist Anil Seth and I published a review listing 22 theories based in the biology of the brain. In 2024, operating with a less restrictive scope, US public intellectual Robert Kuhn counted more than 200. Such a spread suggests that the field lacks a shared criterion for deciding among rival accounts, and that many theories remain too loosely specified to be refuted by a single experiment. This is a question of falsifiability as much as of data.

Measures that can be trusted

Progress also depends on valid and reliable measures of consciousness that work where ordinary report is unavailable. Infants, non-verbal patients, animals and artificial systems all lack the self-report on which much of human consciousness research relies. Measures validated in healthy adults who can describe their experience, such as neural markers of awareness or perturbation-based indices, must therefore be checked against cases where the ground truth is known, including sleep, anaesthesia and clearly reported perception. Without such calibration, a positive reading in a patient or animal can reflect a genuine experience, a different cognitive process, or a feature of the measurement itself. Reliable measures would also let researchers compare theories across species and conditions, rather than only within the human adults who usually serve as the test population.

Adversarial collaboration and indicator-based assessment

Future theories may be shaped by two methodological developments. The first is adversarial collaboration. The 2025 Cogitate study, published in Nature, brought together proponents of two leading theories of consciousness in an ambitious "adversarial collaboration" designed to rigorously test their predictions against empirical data. Its outcome was mixed. The data favored posterior cortical involvement, which was more consistent with IIT's spatial prediction than with GNW's frontoparietal emphasis. Yet GNW's predicted "ignition at offset" was absent. The result was therefore neither theory fully confirmed and neither decisively refuted. Researchers have responded by planning a second large-scale experiment and by making the dataset publicly available, and one of the investigators notes that the field has tested only two of the many theories on offer.

The second development is the indicator approach. Its premise is that the method is to derive indicators from theories of consciousness, then assess whether AI systems are likely to be conscious by determining whether they possess these properties. The method is probabilistic by design. Satisfying the indicators would not mean that such an AI system would definitely be conscious. The indicator framework is therefore a way of updating credence under uncertainty, not a detector of experience. Whether it will remain useful depends on whether the indicators are sharp enough to discriminate between theories and whether they can be applied to systems whose internal workings are not fully known.

Ethical and social questions

If science became able to say reliably where experience is present or absent, several practical questions would follow. In medicine, uncertain diagnoses of awareness raise questions about consent, decision-making and the way families and clinicians communicate with patients who may be aware but cannot respond. In animal research and husbandry, reliable evidence of sentience would bear on the scope of protection. In artificial intelligence, the questions are sharper. Developers might create systems whose internal organization resembles the features that theories associate with experience, and users might attribute experiences to conversational systems that lack them. Legal and regulatory frameworks would need thresholds for when moral consideration is owed, and they would need to decide how to act when evidence is ambiguous. These questions are open, and they are as much about social institutions as about the science of consciousness itself.

References

1. Ned Blocken.wikipedia.org
2. Ned Block’s Phenomenal and Access Consciousness - by blairblairdongqi.substack.com
3. Grand Challenges in Consciousness Research Across Perception, Cognition, Self, and Emotionncbi.nlm.nih.gov
4. Phenomenal and Access Conciousnessprotoscience.fandom.com
5. (PDF) Against Access Consciousnessresearchgate.net
6. Demystifying Consciousness: an interview with Ned Block - The Philosophers' Magazine Archivearchive.philosophersmag.com
7. Mind in Indian Buddhist Philosophy (Stanford Encyclopedia of Philosophy)plato.stanford.edu
8. Abhidharma (Stanford Encyclopedia of Philosophy)plato.stanford.edu
9. Eight Consciousnessesen.wikipedia.org
10. Yogacara school and the theory of consciousnessfiveable.me
11. Alaya-Vijñanatheosophy.world
12. Mencius (Stanford Encyclopedia of Philosophy/Fall 2010 Edition)plato.stanford.edu
13. Mind (Heart-Mind) in Chinese Philosophy (Stanford Encyclopedia of Philosophy)plato.stanford.edu
14. The Emotions in Early Chinese Philosophy (Stanford Encyclopedia of Philosophy)plato.stanford.edu
15. Anatt%C4%81en.wikipedia.org
16. The No-Report Paradigm: A Revolution in Consciousness Research? - PMCpmc.ncbi.nlm.nih.gov
17. Seeing the invisible: The scope and limits of unconscious processing in binocular rivalry - Yonsei Universityyonsei.elsevierpure.com
18. A methodological dilemma for investigating consciousness empirically - ScienceDirectsciencedirect.com
19. Unconscious processing of invisible visual stimulinature.com
20. (PDF) Blindsight and Unconscious Vision: What They Teach Us about the Human Visual Systemresearchgate.net
21. Introspectionen.wikipedia.org
22. History of psychologyen.wikipedia.org
23. History of Psychologylouis.pressbooks.pub
24. Structuralism and functionalismebsco.com
25. What Is Structuralism In Psychology?simplypsychology.org
26. (PDF) The decline and fall of introspectionacademia.edu
27. Neural correlates of consciousness: progress and problemsnature.com
28. Christof Koch and the Neural Correlates of Consciousnessneuroanthropology.net
29. Will we ever have Conscious Machines?arxiv.org
30. Neural correlates of consciousness in humansnature.com
31. What is a Neural Correlate of Consciousness?consc.net
32. CONSCIOUSNESS AND THE PREFRONTAL CORTEXphilarchive.org
33. An integrative view of the role of prefrontal cortex in consciousness - ScienceDirectsciencedirect.com
34. Decoding internally generated transitions of conscious contents in the prefrontal cortex without subjective reportsncbi.nlm.nih.gov
35. Are the neural correlates of consciousness in the front or in the back of the cerebral cortex? Clinical and neuroimaging evidencebiorxiv.org
36. Broadband slow-wave modulation in posterior and anterior cortex tracks distinct states of propofol-induced unconsciousness - PMCpmc.ncbi.nlm.nih.gov
37. Neuroanatomical correlates of brainstem comaresearchgate.net
38. A human brain network derived from coma-causing brainstem lesionsresearchgate.net
39. Rewriting the brain pathway for consciousnesssciencedaily.com
40. Neuroanatomical Basis of Coma in Acute Ischemic Strokebiorxiv.org
41. Posterior and anterior cortex — where is the difference that makes the difference?nature.com
42. Does the Prefrontal Cortex Play an Essential Role in Consciousness? Insights from Intracranial Electrical Stimulation of the Human Brainjneurosci.org
43. Can Pure Thalamic Strokes Lead to Severe Impairment of Arousal? - Jaakkola - 2025 - European Journal of Neurology - Wiley Online Libraryonlinelibrary.wiley.com
44. Thalamic, Hypothalamic, and Brainstem Microbleeds Found in Patients with Severe Traumatic Brain Injury and Comaneurologytoday.aan.com
45. Advances in understanding and treating disorders of consciousness caused by brainstem injurymednexus.org
46. Persistent Vegetative State and Minimally Conscious State: A systematic review and meta-analysis of diagnostic procedures - PMCpmc.ncbi.nlm.nih.gov
47. The Covert Consciousness Dilemmamagazine.hms.harvard.edu
48. Disorders of consciousnessmedlink.com
49. “DoC-feeling”: a new behavioural tool to help diagnose the Minimally Conscious Statebiorxiv.org
50. Diagnostic Developments in Differentiating Unresponsive Wakefulness Syndrome and the Minimally Conscious Statencbi.nlm.nih.gov
51. Functional MRI Detects Residual Awareness in Vegetative Patientsmedscape.com
52. The Clinical Utility of fMRI for Identifying Covert Awareness in the Vegetative State: A Comparison of Sensitivity between 3T and 1.5Tjournals.plos.org
53. Detecting Awareness in the Vegetative State: Electroencephalographic Evidence for Attempted Movements to Commandjournals.plos.org
54. Meta-analysis of the diagnostic value of functional magnetic resonance imaging for distinguishing unresponsive wakefulness syndrome/vegetative state and minimally conscious statencbi.nlm.nih.gov
55. Measuring the Complexity of Consciousnessarxiv.org
56. Perturbational Complexity Indexen.wikipedia.org
57. Assessing the Intra- and Inter-Subject Reliability of the Perturbational Complexity Index (PCI) of Consciousness for Three Brain Regions Using TMS-EEGbiorxiv.org
58. Doidoi.org
59. Detecting the Potential for Consciousness in Unresponsive Patients Using the Perturbational Complexity Indexncbi.nlm.nih.gov
60. The perturbational complexity index detects capacity for consciousness earlier than the recovery of behavioral responsiveness in subacute brain-injured patients - Brain Stimulation: Basic, Translational, and Clinical Research in Neuromodulationbrainstimjrnl.com
61. Diagnostic Developments in Differentiating Unresponsive Wakefulness Syndrome and the Minimally Conscious State - PMCpmc.ncbi.nlm.nih.gov
62. Covert Consciousness: Epistemic Limits of Its Definition and Detection - PubMedpubmed.ncbi.nlm.nih.gov
63. Application of Fast Perturbational Complexity Index to the Diagnosis and Prognosis for Disorders of Consciousness - PubMedpubmed.ncbi.nlm.nih.gov
64. Conscious Processing and the Global Neuronal Workspace Hypothesis - ScienceDirectsciencedirect.com
65. (PDF) The Global Neuronal Workspace Model of Conscious Access: From Neuronal Architectures to Clinical Applicationsresearchgate.net
66. Conscious Processing and the Global Neuronal Workspace Hypothesisresearchgate.net
67. On the utility of toy models for theories of consciousnessarxiv.org
68. evolutionary origins of the Global Neuronal Workspace in vertebratesacademic.oup.com
69. Deep Learning and the Global Workspace Theoryarxiv.org
70. Conscious Access as Continuous-to-Discrete Translationarxiv.org
71. Adversarial testing of global neuronal workspace and integrated information theories of consciousness - Tel Aviv Universitycris.tau.ac.il
72. How Does Consciousness Work? Dueling Scientists Tested Two Big Theories but Found No Winnersingularityhub.com
73. COGITATE and Adversarial Collaborations - The Standard Model of Consciousnessfmt.matthiasgruber.com
74. Integrated Information Theory of Consciousnessiep.utm.edu
75. Integrated information theory - Scholarpediascholarpedia.org
76. A Consciousness-Based Quantum Objective Collapse Modelarxiv.org
77. Tononi’s Integrated Information Theoryloc.closertotruth.com
78. The Integrated Information Theory of Consciousnessphilosophynow.org
79. (PDF) Integrated Information Theory: A Consciousness-First Approach to What Existsresearchgate.net
80. Does integrated information theory make testable predictions about the role of silent neurons in consciousness?academic.oup.com
81. Measuring Integrated Information from the Decoding Perspectivencbi.nlm.nih.gov
82. Integrated information theory: the good, the bad and the misunderstoodarxiv.org
83. An Intriguing and Controversial Theory of Consciousness: IITpsychologytoday.com
84. The Problem with Phi: A Critique of Integrated Information Theoryncbi.nlm.nih.gov
85. The integrated information Φ of an integrate and fire network - PMCpmc.ncbi.nlm.nih.gov
86. Higher-Order Theories of Consciousness — OECSoecs.mit.edu
87. Higher-order theories of consciousness - Scholarpediascholarpedia.org
88. Rosenthal’s Higher-Order Thoughtloc.closertotruth.com
89. How neuroscience will change our view on consciousnessresearchgate.net
90. The Predictive Brain and the 'Hard Problem' of Consciousnesspsychologytoday.com
91. Seth Hallucinations and Recursive Stoppingmedium.com
92. Predictive Coding and the Predictive Brain: A Deep Dive - PeopleShiftpeople-shift.com
93. Anil Seth: "Reality is a controlled hallucination"lab.cccb.org
94. Anil Seth on Consciousness as a Controlled Hallucination - Biology Insightsbiologyinsights.com
95. Understanding Higher-Order Theories of Consciousnesspsychologytoday.com
96. Lamme’s Recurrent Processing Theoryloc.closertotruth.com
97. Seth's “Beast Machine” Theory - Landscape of Consciousnessloc.closertotruth.com
98. Adversarial testing of global neuronal workspace and integrated information theories of consciousnessnature.com
99. (PDF) Adversarial testing of global neuronal workspace and integrated information theories of consciousnessresearchgate.net
100. COGITATE: Testing Two Contrasting Theories of Consciousness (GNWT) & (IIT)templetonworldcharity.org
101. Landmark study puts leading theories of consciousness to the test— neither comes out unscathed - University of Birminghambirmingham.ac.uk
102. IIT Wiki - Supplement of Adversarial Collaborationiit.wiki
103. Neither Theory Survived What the Cogitate Consortium's Adversarial Test Foundtheconsciousness.ai
104. David Chalmersen.wikipedia.org
105. Hard problem of consciousness - Scholarpediascholarpedia.org
106. Hard Problem of Consciousnessiep.utm.edu
107. Knowledge Argument Against Physicalismiep.utm.edu
108. Knowledge argumenten.wikipedia.org
109. Qualia: The Knowledge Argument (Stanford Encyclopedia of Philosophy)plato.stanford.edu
110. Hard problem of consciousness — The Map of Consciousnessmapadelaconsciencia.es
111. Binding problemen.wikipedia.org
112. Synchrony, Binding, and Consciousness - Cees van Leeuwen, 2007journals.sagepub.com
113. (PDF) Why neural synchrony fails to explain the unity of visual consciousnessacademia.edu
114. Consciousness and the Binding Problemblankhorizons.com
115. The Unity of Consciousness and The Binding Problemsuzitravis.substack.com
116. Gazzaniga on Whole Braingurney.itweb.ucf.edu
117. Interaction in isolation: 50 years of insights from split-brain researchacademic.oup.com
118. Split-Brain - an overviewsciencedirect.com
119. Split-brainen.wikipedia.org
120. Who's in Charge of Our Minds? The Interpreterfs.blog
121. Forty-five years of split-brain research and still going strongresearchgate.net
122. Minimal self-models and the free energy principlencbi.nlm.nih.gov
123. From self-organizing systems to subjective temporal extensionarxiv.org
124. Self-modelen.wikipedia.org
125. (PDF) Losing Your Self Being No-One: the Self-Model Theory of Subjectivityresearchgate.net
126. Do we really exist? Eastern inspirations in Thomas Metzinger’s self-model theory of subjectivityejournals.eu
127. The power of direct observation: discovery of REM sleepnature.com
128. Aserinsky Discovers REM Sleepebsco.com
129. Who discovered REM sleep? - PMC - NIHpmc.ncbi.nlm.nih.gov
130. Nathaniel Kleitmanen.wikipedia.org
131. The Biology of REM Sleep - ScienceDirectsciencedirect.com
132. Review Anesthesia and the neurobiology of consciousnesssciencedirect.com
133. Mechanisms of unconsciousness during general anaesthesia - ScienceDirectsciencedirect.com
134. Anesthesia-induced loss of consciousness disrupts auditory responses beyond primary cortexpnas.org
135. General Anesthetics and Molecular Mechanisms of Unconsciousness - PMCpmc.ncbi.nlm.nih.gov
136. Feature Review Towards a Comprehensive Understanding ofmed.stanford.edu
137. Shannon entropy of brain functional complex networks under the influence of the psychedelic Ayahuascaarxiv.org
138. The entropic brain: a theory of conscious states informed by neuroimaging research with psychedelic drugsncbi.nlm.nih.gov
139. The entropic brain - revisited - ScienceDirectsciencedirect.com
140. A critical review of brain entropy as a biomarker of the psychedelic state - PubMedpubmed.ncbi.nlm.nih.gov
141. Psychedelics, entropic brain theory, and the taxonomy of conscious states: a summary of debates and perspectivesacademic.oup.com
142. Consciousness makes sense in the light of evolution - ScienceDirectsciencedirect.com
143. Frontiersfrontiersin.org
144. The evolutionary functions of consciousnessroyalsocietypublishing.org
145. Nicholas Humphrey Explainedeverything.explained.today
146. Category: Book reviews - Digital Minds - WordPress.comdigitalminds2016.wordpress.com
147. The social origins of consciousnessroyalsocietypublishing.org
148. Evolution of Consciousnessunfinishablemap.org
149. Nicholas Humphreyen.wikipedia.org
150. Evolution of Consciousnessncbi.nlm.nih.gov
151. Can biological quantum networks solve NP-hard problems?arxiv.org
152. Orchestrated Objective Reduction of Quantum Coherence ...esalq.usp.br
153. REVIEW articlefrontiersin.org
154. The Importance of Quantum Decoherence in Brain Processesresearchgate.net
155. Quantum Computation in Brain Microtubules?Decoherence and Biological Feasibilityarxiv.org
156. Understanding Physical Processes in Describing a State of Consciousness: A Reviewarxiv.org
157. How biological microtubules may avoid decoherence (Conference)osti.gov
158. Frontiersfrontiersin.org
159. The revised Penrose-Hameroff orchestrated objective-reduction proposal for human consciousness is not scientifically justified Comment on "Consciousness in the universe: A review of the 'Orch OR' theory" by Hameroff and Penroseresearchgate.net
160. New research on anesthesia unlocks important clues about the nature of consciousnesssciencedaily.com
161. pmc.ncbi.nlm.nih.govpmc.ncbi.nlm.nih.gov
162. hameroff.arizona.eduhameroff.arizona.edu
163. Finite-Memory Extension of Tegmark's Decoherence Bound in Biological Mediaarxiv.org
164. Penrose-Hameroff’s Orchestrated Objective Reductionloc.closertotruth.com
165. Roger Penrose & Stuart Hameroff, Orchestrated objective reduction of quantum coherence in brain microtubules: The "orch OR" model for consciousness - PhilPapersphilpapers.org
166. Quantum computation in brain microtubules: Decoherence and biological feasibilitylink.aps.org
167. Microtubule-Stabilizer Epothilone B Delays Anesthetic-Induced Unconsciousness in Ratseneuro.org
168. Anesthesia study unlocks clues about the nature of consciousnessmedicalxpress.com
169. Identifying indicators of consciousness in AI systemsresearchgate.net
170. Insights from the Science of Consciousnessarxiv.org
171. Turing, A.M. (1950). Computing machinery and intelligence. Mind, 59, 433-460.cs.ox.ac.uk
172. Summary of 'Computing Machinery And Intelligence' (1950) by Alan Turingjackhoy.com
173. The Chinese Room Argument (Stanford Encyclopedia of Philosophy)plato.stanford.edu
174. Chinese Room Argumentiep.utm.edu
175. Chinese Room Argument: Syntax, Semantics and Computation - Interactivembrenndoerfer.com
176. Chinese room argumentbritannica.com
177. Five years of the Cambridge Declaration on Consciousnessanimal-ethics.org
178. The Cambridge declaration on consciousness · Pablo Stafforinistafforini.com
179. Frankfurt Declaration on the Cambridge ...forum.effectivealtruism.org
180. The Cambridge Declaration on Consciousness - Animal Cognitionanimalcognition.org
181. Mirror testen.wikipedia.org
182. Graded Mirror Self-Recognition by Clark’s Nutcrackersncbi.nlm.nih.gov
183. Rethinking self-awareness: insights and biases from self-recognition research - ScienceDirectsciencedirect.com
184. The Mirror Test: Which Animals Pass, Which Fail, and What It Reveals About Animal Intelligence and Consciousnesstimeslife.com
185. A ‘Self-Aware’ Fish Raises Doubts About a Cognitive Testquantamagazine.org
186. www.frontiersin.orgfrontiersin.org
187. Cephalopods and decapod crustaceans Sentiencefactuallyvegan.com
188. Volume 31 Issue 1 - February 2022 Report - UFAWufaw.org.uk
189. Consciousness, AI, and the Limits of Scientific Explanationarxiv.org
190. Jonathan Birch (philosopher)en.wikipedia.org
191. Call for evidence: background - GOV.UKgov.uk
192. Animal Welfare (Sentience) Act formally recognises animals as sentient beingsworldanimalprotection.org.uk
193. Decapod Crustaceans under the Law - UK Human Rights Blogukhumanrightsblog.com
194. Call for evidence on the use of decapod crustaceans in research and education - British Pharmacological Societybps.ac.uk
195. Stream of Consciousness, 1890-1929: Causes and Legacyexplaininghistory.org
196. Stream of Consciousness_Baiduwikibaike.baidu.com
197. (PDF) Stream of Consciousness and Interior Monologue Techniques in Dokuzuncu Hariciye Koğuşu and Mrs Dallowayresearchgate.net
198. Stream of Consciousness: How to Unravel Our Messy Thoughtsthemindcollection.com
199. Free Indirect Discourse: Definition, Examples & How to Use It - Hearthhearth.sh
200. Mrs Dallowayen.wikipedia.org
201. Narrative Techniques of Virginia Woolf - The Classroomtheclassroom.com
202. Narrative Techniques in the Modernist English Novelseejph.com
203. Stream of Consciousness - Routledge Encyclopedia of Modernismrem.routledge.com
204. Stream Of Consciousnessencyclopedia.com
205. Stream of consciousness (psychology)en.wikipedia.org
206. Narrative Technique of Virginia Woolf in Mrs. Dalloway Suman Raniijcrt.org
207. Stream Of Consciousness technique used in Virginia ...researchguru.net
208. Landmark collaboration probes the foundations of consciousness - Health e-Newshealthenews.mcgill.ca
209. Identifying indicators of consciousness in AI systems: Trends in Cognitive Sciencescell.com
210. Rethinking Consciousness: When Science Puts Itself to the Test - Max Planck Neurosciencemaxplanckneuroscience.org