What Are Higher-Order Theories of Consciousness?
Higher-Order Theories (HOTs) of consciousness share a deceptively simple core idea: a mental state is conscious when, and only when, it is the object of an appropriate higher-order mental representation. You see red — that is a first-order visual state. You become aware that you are seeing red — that is a higher-order state. According to HOT theories, the second step is what makes the experience conscious. Without it, the first-order state still occurs and influences behavior, but there is nobody home to experience it.
This framework offers a clean explanation for why so much of our mental life is unconscious. Subliminal perception, blindsight, automatic motor routines, and implicit learning all involve first-order processing that lacks higher-order accompaniment. Consciousness, on this view, is not a property of certain brain regions or types of information — it is a relational property that emerges when the mind represents its own states.
The Core Framework
David Rosenthal's Higher-Order Thought theory, the most developed version, holds that a mental state M is conscious when the subject has a concurrent thought to the effect that they are in state M. This higher-order thought must be non-inferential (not arrived at by reasoning), concurrent with the first-order state, and assertoric (representing the state as actually occurring, not merely possible).
Crucially, the higher-order thought itself is typically not conscious — it does not require a yet higher-order thought about it. This avoids the infinite regress that plagued earlier introspectionist theories. The higher-order thought operates in the background, conferring consciousness on its target without itself needing to be experienced.
Who Proposed It
David Rosenthal at the City University of New York Graduate Center developed HOT theory across several decades of publications beginning in the 1980s, drawing on earlier ideas from philosophers like Franz Brentano and David Armstrong. Hakwan Lau, now at the RIKEN Center for Brain Science (previously UCLA), has been the leading neuroscientific proponent, developing Perceptual Reality Monitoring (PRM) theory — a neural implementation of higher-order principles that identifies prefrontal cortical circuits as the substrate of higher-order representations. Richard Brown at LaGuardia Community College has contributed important theoretical refinements.
Key Evidence
The strongest evidence for higher-order theories comes from dissociations between perception and awareness. Blindsight patients, whose primary visual cortex is damaged, can accurately guess visual properties of stimuli they report not seeing. They have first-order visual representations without higher-order awareness — exactly what HOT predicts should produce unconscious vision.
Hakwan Lau's neuroimaging studies have shown that prefrontal cortical activity predicts subjective visual awareness even when controlling for perceptual accuracy. In signal detection terms, prefrontal activity correlates with metacognitive sensitivity (knowing when you perceived correctly) rather than perceptual sensitivity (perceiving correctly). This supports the idea that prefrontal regions implement the higher-order representations that generate conscious awareness.
Additional evidence comes from studies of metacognition in neurological patients. Damage to prefrontal cortex can produce anosognosia — unawareness of one's own deficits — suggesting that higher-order monitoring mechanisms are neurally dissociable from the first-order processes they monitor.
Key Objections
Higher-order theories face several challenges. The "targetless representation" problem asks: what happens when a higher-order thought misrepresents? If you have a higher-order thought that you are in pain when you are not, does HOT predict you are conscious of pain? Rosenthal bites this bullet, arguing that such cases produce genuine conscious experience of pain — a counterintuitive result.
Ned Block has argued forcefully that higher-order theories conflate access consciousness (information available for report and reasoning) with phenomenal consciousness (the subjective quality of experience). Block contends there can be phenomenal consciousness without higher-order access, as in cases of "overflow" where visual experience is richer than what attention and report can capture.
There is also debate about whether prefrontal cortex is truly necessary for consciousness. Studies of conscious experience in animals with limited prefrontal cortex, and in human patients with prefrontal damage who still report conscious experiences, challenge the neural predictions of HOT-based theories.
Why It Matters
Higher-order theories matter because they offer one of the clearest answers to a question other theories struggle with: what is the difference between conscious and unconscious mental states? By locating consciousness in the relationship between mental states rather than in their intrinsic properties, HOT theories provide a principled criterion for when consciousness is present. This has direct implications for assessing consciousness in non-communicative patients, animals, and AI systems. If higher-order representation is the key, then the question becomes empirically tractable: does the system represent its own internal states? The answer determines whether the lights are on.
