Consciousness is the subjective experience of being aware, of having an inner life that feels like something. Despite being the most familiar thing in your existence, it remains one of the hardest phenomena for science to pin down. Researchers can now measure proxies for it, track its disappearance under anesthesia, and debate whether birds or even octopuses share some version of it. But no single theory yet explains why certain patterns of brain activity give rise to experience at all. What science has managed, especially in the past two decades, is to get remarkably specific about what consciousness looks like from the outside, even while the inside remains mysterious.
Why Defining Consciousness Is So Difficult
Part of the problem is that consciousness is not one thing. Researchers typically split it into at least two dimensions. The first is the state of consciousness: whether you are awake, asleep, sedated, or in a coma. The second is the contents of consciousness: what you are actually experiencing at any given moment, such as the color of a sunset or the taste of coffee. These two dimensions can come apart. You can be in a wakeful state yet have your conscious contents sharply narrowed (think of highway hypnosis), or you can be in a sleeping state yet have vivid conscious experience (dreaming). The thalamus, a relay hub deep in the brain, appears to underlie both of these dimensions through two distinct populations of neurons: one type that helps maintain wakefulness and set perceptual thresholds, and another that sustains specific perceptual content and stability.
1PubMed Central. Thalamic contributions to the state and contents of consciousnessA class of deep cortical neurons links these thalamic circuits to the cortex, effectively coupling the state and the contents of consciousness together into a unified broadcasting system. When that coupling breaks down, as it does under anesthesia or after severe brain injury, experience either narrows dramatically or disappears.
2PubMed Central. Coupling the State and Contents of ConsciousnessThe Leading Theories
Several competing theories try to explain how brain activity becomes conscious experience. None has achieved consensus, but three get the most attention in the research community.
Global Neuronal Workspace (GNW) theory holds that most of what your brain processes stays unconscious, handled by specialized local circuits. A piece of information becomes conscious when it triggers a sudden, nonlinear “ignition” across a network of widely distributed neurons, making it available to many brain areas at once. Think of it like a performer stepping onto a brightly lit stage: the information goes from private backstage processing to being broadcast throughout the theater.
3PubMed Central. Conscious Processing and the Global Neuronal Workspace HypothesisIntegrated Information Theory (IIT) takes a fundamentally different approach. Rather than asking what the brain does when you are conscious, it asks what kind of system can be conscious at all. Its answer: consciousness corresponds to a system’s capacity to integrate information. A system that generates many different possible states (differentiation) and ties them together into a unified whole (integration) has consciousness to a degree measured by a quantity called Phi.
4PubMed Central. An information integration theory of consciousnessIIT is elegant in theory, but calculating Phi for real biological systems turns out to be a serious problem. A recent analysis found that a single neural subsystem could produce a wide spectrum of Phi values, ranging from zero to 0.83 bits, depending on how the calculation was set up. The single published value for that system was just one point in that range. If the true value turned out to be zero, the system would not count as integrated at all, and the entire biological interpretation would collapse. That kind of ambiguity makes IIT hard to test in practice.
5Neuroscience of Consciousness. On the non-uniqueness problem in integrated information theoryHigher-Order Theories propose that you are conscious of something only when your brain generates a representation of that mental state. In other words, seeing a red apple involves one neural process, but being aware that you are seeing a red apple requires a second, higher-order process that monitors the first. Recent cognitive neuroscience work has implicated the prefrontal cortex in these higher-order representations, consistent with the idea that the prefrontal region plays a key role in generating the cognitive states underlying awareness.
6Trends in Cognitive Sciences. Higher-Order Theories of Consciousness: Answering Common Objections and Clarifying Recent ProgressFront of the Brain or Back of the Brain
One of the liveliest debates in consciousness research is where in the cortex conscious experience actually lives. GNW and Higher-Order theories both point toward the prefrontal cortex, the front of the brain, as essential for awareness. Other researchers argue that the posterior cortex, especially parietal and occipital regions sometimes called the “posterior hot zone,” is where specific conscious contents are generated, and that prefrontal activity just reflects downstream processing like attention and decision-making rather than consciousness itself.
7PubMed Central. Are the Neural Correlates of Consciousness in the Front or in the Back of the Cerebral Cortex? Clinical and Neuroimaging EvidenceDreaming has become a useful testing ground for this debate, since dreamers are clearly conscious yet their prefrontal cortex is often much less active than during waking life. Some researchers have argued that this proves consciousness does not require the frontal regions. However, the evidence is murkier than it first appears. Neither the classical claim about prefrontal inactivity during REM sleep nor the more recent emphasis on posterior hot zone activity during dreaming holds up cleanly under scrutiny, leaving the front-versus-back debate genuinely unresolved.
8PubMed Central. Dream experiences and the neural correlates of perceptual consciousness and cognitive accessMeasuring Consciousness at the Bedside
Theoretical debates aside, clinicians face a concrete version of the consciousness problem every day: is this unresponsive patient aware? A person who cannot move or speak may still be conscious, and misdiagnosing them as unaware has devastating consequences for their care and their family’s decisions.
The perturbational complexity index, or PCI, has emerged as one of the most promising tools for this. It works by sending a magnetic pulse into the cortex using transcranial magnetic stimulation and then measuring how complex the resulting brain response is. A brain that is truly unconscious produces a simple, stereotyped reaction. A conscious brain produces a response that is both widespread and complex, reflecting the kind of integrated information processing the theories predict. PCI reliably distinguished conscious from unconscious states across healthy volunteers during wakefulness, dreaming, and various depths of anesthesia, as well as in patients who had emerged from coma.
9PubMed. A theoretically based index of consciousness independent of sensory processing and behaviorWhat makes PCI valuable is that it does not depend on the patient being able to see, hear, or move. It detects consciousness even when sensory and motor pathways are damaged, which means it can catch cases where a patient is aware but physically unable to demonstrate it.
10Brain Stimulation. Assessing the Capacity for Consciousness in Prolonged Disorders of Consciousness Using TMS-EEGPCI is also being explored as a way to predict which patients might respond to brain stimulation treatment. In a controlled trial of patients with prolonged disorders of consciousness, those who responded to repetitive transcranial magnetic stimulation showed significant increases in PCI, while non-responders did not. Patients who started with higher baseline PCI values were more likely to be responders in the first place, suggesting the measure could help clinicians target treatment to those most likely to benefit.
11PubMed Central. Perturbational complexity index in assessing responsiveness to rTMS treatment in patients with disorders of consciousness: a cross-over randomized controlled trial studyHidden Awareness in Unresponsive Patients
Brain imaging has revealed that some patients diagnosed as being in a vegetative state, now more commonly called unresponsive wakefulness syndrome, show brain activity patterns that look remarkably like those of minimally conscious patients. In one study, brain scans during acoustic stimulation showed that ten patients clinically classified as vegetative had activation patterns very similar to those of minimally conscious patients. Six months later, those ten patients had improved enough to be reclassified as minimally conscious, while the other vegetative-state patients remained stable.
12PubMed Central. Functional Evaluation of Awareness in Vegetative and Minimally Conscious StateA meta-analysis of the diagnostic performance of functional MRI in distinguishing these two states concluded that the technique has good accuracy, and can serve as a tool both for evaluating prognosis and for guiding rehabilitation.
13PubMed Central. Meta-analysis of the diagnostic value of functional magnetic resonance imaging for distinguishing unresponsive wakefulness syndrome/vegetative state and minimally conscious stateThe practical takeaway here matters enormously. Behavioral assessment alone misses a meaningful fraction of patients who have some form of inner life. Brain-based tools like PCI and functional MRI are slowly shifting the standard of care, but they are not yet routine in most hospitals, which means misdiagnosis remains a real and ongoing problem.
What Anesthesia Reveals
Anesthesia offers a controlled way to watch consciousness disappear and return. Under ketamine sedation, researchers have tracked how specific brain networks break down in sequence. The default mode network, a set of brain regions active during mind-wandering and self-referential thought, loses internal connectivity as sedation deepens. Connectivity between the medial prefrontal cortex and the rest of the default mode network dropped dramatically at deeper sedation levels. The normally antagonistic relationship between the default mode network and other brain regions also collapsed. Meanwhile, sensory and motor networks were relatively unaffected.
14Anesthesiology. Resting-state Network-specific Breakdown of Functional Connectivity during Ketamine Alteration of Consciousness in VolunteersThis selective breakdown is revealing. Consciousness does not fade because the whole brain shuts down. Instead, specific network relationships dissolve, especially those involving self-referential processing and the dynamic push-pull between internally and externally focused brain regions. The fact that sensory networks keep humming along while consciousness fades suggests that raw sensory processing is not enough for experience. You also need the higher-level integration those network relationships provide.
Altered States and the Boundaries of Normal Consciousness
Psychedelics push consciousness in the opposite direction from anesthesia: instead of simplifying brain activity, they make it more chaotic. Under substances like psilocybin and LSD, the default mode network loses its normal internal coherence and its usual oscillatory rhythms desynchronize. The brain becomes more disordered, which paradoxically seems to produce a state that feels richer and less constrained than ordinary waking life. This aligns with what researchers call the entropic brain hypothesis: psychedelics increase neural entropy, resulting in a mind that is more flexible and less bound by the usual patterns of self-referential thought.
15PubMed Central. Default Mode Network Modulation by Psychedelics: A Systematic ReviewLucid dreaming offers yet another variant. During a lucid dream, you know you are dreaming and can sometimes control the dream’s content. Brain imaging of lucid dreamers shows strong activation in the precuneus, parietal lobules, and prefrontal and occipito-temporal regions compared with ordinary REM sleep. These are areas normally quiet during dreaming, and their reactivation corresponds to the recovery of reflective cognitive abilities that define lucidity.
16Sleep. Neural Correlates of Dream Lucidity Obtained from Contrasting Lucid versus Non-Lucid REM Sleep: A Combined EEG/fMRI Case StudySome researchers have even proposed that lucid dreaming and psychosis sit at opposite ends of a spectrum. Lucid dreaming involves a physiological awakening of frontal activity during sleep, while psychosis involves a pathological intrusion of dreamlike features during waking life due to reduced frontal function.
17Medical Hypotheses. Neurobiology and clinical implications of lucid dreamingPerception as Controlled Hallucination
Even ordinary waking consciousness is stranger than it feels. Under predictive processing accounts of the brain, what you perceive is not a direct readout of sensory input. Instead, your brain constantly generates predictions about what is out there and checks those predictions against incoming data. What you experience is the brain’s best guess, updated by sensory error signals. This has led some researchers to describe perception as “controlled hallucination,” where the hallucination is your brain’s model and the control comes from sensory feedback keeping it in check.
18Analytic Philosophy. Perception as controlled hallucinationYour sense of having a body and a self follows similar principles. The anterior insular cortex plays a critical role in interoceptive attention, the ability to notice signals from inside your own body like your heartbeat or breathing. Brain imaging has confirmed the insular cortex’s involvement in this process, and patients with focal damage to this region show reduced ability to attend to internal bodily signals.
19PubMed Central. Anterior insular cortex plays a critical role in interoceptive attentionThis means your conscious sense of being a self in a body is not just passively given to you. It is actively constructed by brain circuits monitoring internal states. Disrupt those circuits and the sense of self can fragment, as happens in some neurological and psychiatric conditions.
Split Brains and the Unity of Experience
One of the most striking findings about consciousness comes from patients whose corpus callosum, the massive bundle of fibers connecting the brain’s two hemispheres, has been surgically severed. Classic split-brain research suggested these patients effectively had two separate conscious agents in one skull, each hemisphere unaware of what the other was experiencing. More recent and carefully controlled work challenges that picture. Testing across a wide variety of tasks showed that split-brain patients could respond accurately to stimuli appearing anywhere in the visual field, regardless of whether they responded verbally or with either hand. They could not compare stimuli across visual half-fields, and each hemisphere showed separate performance levels for certain tasks. But unity in consciousness and behavioral responding was largely preserved.
20Brain. Split brain: divided perception but undivided consciousnessThis raises an intriguing possibility: the cortical connections between hemispheres may be less critical for unified consciousness than previously assumed. Subcortical structures, including the thalamus and brainstem, may be doing more of the binding work than the classic model gave them credit for.
Consciousness Without a Cortex
Mammals are not the only animals with plausible claims to consciousness. Crows lack a layered cerebral cortex entirely, yet single-neuron recordings from their pallial endbrain show activity patterns that correlate with the birds’ subjective perception of whether a visual stimulus was present or absent. When a crow reported seeing something, specific neurons fired in a pattern consistent with perceptual awareness, and these responses could not be explained by the stimulus alone since they correlated with the bird’s report rather than the physical stimulus properties.
21PubMed. A neural correlate of sensory consciousness in a corvid birdThese results suggest that consciousness does not require the specific layered architecture of the mammalian cortex. The bird pallium evolved independently, uses different cell types, and organizes itself differently, yet still appears to support something that looks like sensory awareness. This implies that the broader anatomical and physiological principles of the telencephalic pallium offer a substrate for consciousness to evolve independently across vertebrate lineages.
22PubMed. Consciousness without cortexCephalopods push the boundaries even further. Octopuses are invertebrates separated from vertebrates by over 500 million years of evolution, yet they display sophisticated learning, flexible behavior, and nervous systems complex enough that researchers have proposed they provide a unique test case for examining basal consciousness among invertebrates.
23Frontiers in Systems Neuroscience. Cephalopod Behavior: From Neural Plasticity to ConsciousnessOctopuses may demonstrate perceptual richness, neural unity, temporal awareness, and affective evaluation, all properties considered markers of consciousness.
24PubMed Central. The Case for Octopus Consciousness: ValenceOne especially provocative idea is that octopus arms, which contain the majority of the animal’s neurons and can sense and act even when amputated, might support their own local form of primary consciousness: a direct awareness of the world that operates semi-independently from the central brain.
25Frontiers in Systems Neuroscience. Where Is It Like to Be an Octopus?How Far Back Does Consciousness Go
If crows and octopuses both have something like consciousness, the question becomes whether awareness evolved once, deep in evolutionary history, or arose multiple times independently. One proposal uses the presence of multiple levels of somatotopic neural maps, where the brain contains organized spatial representations of the body and the sensory world, as an objective marker for sensory consciousness. All living vertebrates have these maps, which suggests that consciousness in some form extends back to the origin of vertebrates in the Cambrian period, over 520 million years ago. The first conscious sense may have been vision, which then combined with additional sensory systems to transform simple reflexive brains into organs that could actually perceive the world.
26PubMed Central. The evolutionary and genetic origins of consciousness in the Cambrian Period over 500 million years agoThe crow findings support this timeline. Since the avian and mammalian lineages diverged hundreds of millions of years ago, consciousness-supporting brain structures either evolved before that split or arose independently at least twice. Either scenario means consciousness is not some lucky accident of the mammalian cortex. It is a feature that nervous systems arrive at repeatedly, possibly because it offers something so useful that evolution keeps reinventing it.
27PubMed. A neural correlate of sensory consciousness in a corvid birdCan Machines Be Conscious
The rapid development of AI has made the question of machine consciousness move from philosophy seminar rooms into mainstream discussion. A large-scale survey of scientific theories of consciousness, including GNW, IIT, higher-order theories, predictive processing, and attention schema theory, derived a set of “indicator properties” of consciousness stated in computational terms that could, in principle, be evaluated in AI systems.
28arXiv. Consciousness in Artificial Intelligence: Insights from the Science of ConsciousnessCurrent AI systems do not meet most of these indicators. They process information and generate sophisticated outputs, but they generally lack the recurrent, self-monitoring, globally integrated architecture that the leading theories associate with consciousness.
An interesting philosophical argument holds that the tests we use to assess consciousness in animals could be adapted for machines. If an AI system displayed clusters of cognitive capacities normally linked to consciousness in humans, and those capacities could be selectively disrupted in ways that mirror masking effects in human perception, and the system reproduced human-like behavior in carefully designed tasks, then the argument goes that we should treat it as conscious.
29Erkenntnis. Tests of Animal Consciousness are Tests of Machine ConsciousnessWhether this standard will ever be met, and whether meeting it would truly mean the machine is conscious rather than merely behaving as if it were, remains one of the deepest open questions at the intersection of neuroscience, computer science, and philosophy. The hard problem of consciousness, the question of why any physical process gives rise to subjective experience at all, does not get easier just because the substrate changes from carbon to silicon.