CAT 2025 Slot 3 VARC: Competing Theories of Consciousness

Consciousness presents a peculiar problem for science: it is the one phenomenon we know with total certainty exists, because each of us experiences it directly, yet it is also the one phenomenon that resists the standard scientific method of explaining a system by describing the mechanical interactions of its parts. You can explain digestion by describing enzymes breaking molecular bonds, and the explanation feels complete — nothing further needs accounting for. But even a perfect, atom-by-atom description of every neuron firing in a brain seems to leave something unexplained: why should any of that electrochemical activity be accompanied by subjective experience at all, rather than occurring "in the dark," with no one home to notice it? Philosophers call this the "hard problem" of consciousness, to distinguish it from the merely difficult problems of explaining how the brain processes information, focuses attention, or reports on its own states — problems that are hard technically but not, in principle, mysterious.

Several major theories now compete to explain consciousness, each drawing the line between "conscious" and "unconscious" processing in a different place. Global Workspace Theory, associated with Bernard Baars and later developed by Stanislas Dehaene, holds that consciousness arises when information is broadcast widely across the brain, becoming available to many specialised subsystems at once — memory, language, planning — rather than remaining confined to the single module that first processed it. On this view, a stimulus becomes conscious the moment it "wins" a competition for access to this shared workspace and gets distributed everywhere.

Integrated Information Theory, developed principally by Giulio Tononi, takes a strikingly different approach: it starts not from a functional description of what conscious systems do, but from a mathematical measure — denoted Φ (phi) — of how much information a system generates as an integrated whole, over and above what its parts generate independently. A system with high Φ is, on this account, conscious to a correspondingly high degree, whatever it is made of; a system with low Φ is not, no matter how sophisticated its behaviour looks from outside. This theory produces a genuinely unsettling implication: it does not obviously rule out the possibility that certain non-biological systems, if wired appropriately, could possess some degree of consciousness, while also implying that some biological arrangements most people would expect to be conscious might not be, if their internal information happens not to integrate the right way.

A third contender, Higher-Order Theory, locates consciousness not in the breadth of information's spread nor in a measure of integration, but in a specific kind of self-representation: a mental state becomes conscious, on this view, only when the brain forms a further, higher-order representation about that state — a thought about the thought, so to speak. This has the advantage of explaining, intuitively, why we feel conscious states as somehow "ours" and reportable, but critics note that it risks an infinite regress: if a state needs a higher-order thought to become conscious, does that higher-order thought itself need a still-higher representation?

None of these theories has been decisively confirmed or refuted, partly because they make few enough distinguishing empirical predictions and partly because we lack any agreed, independent test for whether a given system is conscious — the very thing the theories are meant to explain. Some researchers now argue that progress will come less from a single winning theory than from carefully designed experiments — adversarial collaborations, in which rival theorists agree in advance on what result would count against their own position — that at least narrow the field, even if a full explanation of subjective experience itself remains out of reach.