Few phrases in philosophy have carried as much weight as the one Thomas Nagel placed at the centre of his 1974 paper. An organism has conscious mental states, he wrote, if and only if there is something it is like to be that organism, something it is like for the organism. Half a century later, the formulation remains the standard gloss on phenomenal consciousness. It appears in neuroscience reviews and metaphysics seminars alike, usually without comment.

Yet its ubiquity conceals a peculiar fact. The phrase is not a definition in any ordinary sense. It specifies no necessary and sufficient conditions that could be checked independently and offers no operational criterion. It also resists paraphrase into terms that do not already presuppose what is being defined. Critics have called it empty. Defenders insist that its apparent emptiness is precisely the point, since it gestures at something that third-person vocabulary cannot capture.

So what does subjective character actually mean? Is 'what it is like' a technical term with determinate content, a rhetorical pointer, or a grammatical confusion masquerading as insight? The answer matters beyond exegesis. How we construe the phrase shapes what counts as a theory of consciousness and what the hard problem demands of us. It also bears on whether the explanatory gap is a feature of reality or an artefact of how we have framed the question.

Nagel did not coin the phrase. B. A. Farrell used it in 1950, wondering what it would be like to be a bat, and Timothy Sprigge deployed a similar locution in 1971. What Nagel did was turn it into an argument. In What Is It Like to Be a Bat? the phrase becomes the fulcrum of a challenge to reductive physicalism. If experience has a subjective character accessible only from a particular point of view, then any objective, viewpoint-neutral description will necessarily leave it out.

The precise grammar matters. Nagel is emphatic that 'like' here does not signal comparison or resemblance. To say there is something it is like to be a bat is not to say that being a bat resembles anything else. Nor does the phrase mean 'how it feels to the bat' in a merely dispositional or behavioural sense. Instead, 'like' indexes a way of being for a subject. It marks a mode of presentation that exists only insofar as there is someone to whom it is presented.

This is why the qualifier 'for the organism' carries so much of the load. Nagel's point is not simply that bats have sensory states with particular qualities. It is that those qualities are given to a perspective. Echolocation is not merely a process occurring inside a bat; it is, presumably, something the bat undergoes. The subjective character of experience is the fact of its being had, not just the content of what is had.

The formulation thereby fuses two ideas that later philosophers would prise apart. One is a qualitative dimension: the specific feel of sonar perception. The other is a subjective dimension: that feel's being for someone. Nagel treats them as inseparable, and his anti-reductionist conclusion depends on that fusion. Facts about experience are essentially perspectival because experience is constitutively for a subject. Separate the two, and the argument's structure changes considerably.

Crucially, Nagel did not claim that physicalism was false. He claimed that we currently lack any conception of how it could be true. The 'what it is like' locution was meant to mark the precise location of that conceptual deficit, not to settle the metaphysics in advance. Read this way, the phrase is less a definition than a diagnosis. It names the place where our explanatory concepts run out, and it invites us to ask why they run out there.

Takeaway

Nagel's phrase is not about resemblance but about givenness. An experience has subjective character because it is for someone, not merely because it has distinctive qualities.

The most aggressive critique came from P. M. S. Hacker, who argued in 2002 that the philosophical use of the phrase is grammatically defective. 'What is it like to be X?' ordinarily invites a comparative or evaluative answer, such as 'exhausting' or 'wonderful'. Those answers concern the experience of occupying a role, not some ineffable inner quality. On Hacker's Wittgensteinian reading, philosophers detach the phrase from the practices that give it sense. What remains is a form of words without content.

Daniel Stoljar's semantic analysis offers a more constructive challenge. He argues that 'there is something it is like for x to be F' has a relational structure, roughly: there is a way being F affects x. If that is right, the locution is perfectly meaningful but far less metaphysically loaded than often assumed. It applies to plenty of non-phenomenal cases, since there is something it is like for a team to lose a final. So it cannot pick out phenomenal consciousness without additional stipulation.

Uriah Kriegel takes the opposite approach, accepting the phrase's richness and dissecting it. He distinguishes qualitative character, the reddish or painful aspect of an experience, from subjective character, its for-me-ness. On his self-representational theory, subjective character consists in an experience representing itself in a peripheral, non-reflective way. This gives Nagel's vague 'for the organism' clause a determinate theoretical shape. The cost is commitment to a contested account of how that shape is realised.

Eric Schwitzgebel proposes retreating from analysis altogether. Phenomenal consciousness, he suggests, is best defined by example. It is the property most obviously shared by positive cases, such as sensory experience, imagery, inner speech and felt emotion. It is absent from negative cases, such as immune responses or implicit knowledge. On this view, the 'what it is like' phrase becomes one pointer among several. It is valuable for gesturing at the phenomenon rather than specifying its nature.

Each strategy buys clarity at a price. Hacker's dissolution leaves a residue that grammatical therapy struggles to explain away. Stoljar's semantics strips the phrase of its distinctive force, and Kriegel's analysis builds theory into what was meant to be a neutral starting point. Schwitzgebel's ostension depends on a shared intuitive grasp that illusionists such as Keith Frankish dispute. No single refinement has yet achieved both precision and neutrality, and it is an open question whether any could.

Takeaway

Every attempt to sharpen 'what it is like' trades neutrality for precision. When someone claims to have clarified the phrase, ask which of the two they paid with.

Given these difficulties, how should researchers actually use the notion? The first principle is to treat 'what it is like' as fixing a referent rather than revealing an essence. Like 'water' before chemistry, the term can pick out a phenomenon without telling us what that phenomenon fundamentally is. This separates the definitional task from the explanatory one. It also blocks a common slide: from 'consciousness is defined subjectively' to 'consciousness cannot be explained objectively'.

Second, disaggregate. Following Kriegel and Ned Block, distinguish qualitative character, subjective character, access consciousness, and reflective self-awareness. Much confusion in the literature arises when an argument trades on one sense and draws conclusions about another. Evidence that a patient lacks reportable access does not, by itself, establish an absence of phenomenality. Equally, an account of for-me-ness is not automatically an account of qualitative feel.

Third, remain theory-neutral at the point of definition. A definition that already entails dualism, representationalism or a higher-order theory cannot adjudicate between them. This matters acutely for empirical work. Integrated information theory, global workspace theory and recurrent processing accounts all claim to explain phenomenal consciousness. If each defines its target in its own terms, adversarial collaborations risk measuring different things under a single name and calling the results a contest.

Fourth, anticipate the illusionist challenge. Suppose 'what it is like' denotes something genuinely ineffable, intrinsic and private; illusionists deny that anything answers to it. Suppose instead it denotes whatever introspection represents as present; illusionists can happily accept it. Stating which construal is in play prevents disputes that are merely verbal. Frankish's distinction between phenomenal and quasi-phenomenal properties is a useful tool here, even for those who reject his conclusion.

Finally, accept that some residual indeterminacy may be ineliminable, and even informative. That the phrase resists paraphrase is itself data. Perhaps subjectivity is a primitive that cannot be analysed in other terms. Perhaps instead our concept of it is structured in ways that systematically mislead us. Both possibilities are substantive hypotheses, and a well-used definition keeps both open rather than foreclosing one by stipulation.

Takeaway

Let 'what it is like' point at the phenomenon without deciding its nature. A definition that settles the metaphysics in advance can no longer serve as common ground for testing rival theories.

Nagel's phrase has survived half a century of scrutiny not because it is precise, but because it locates something no precise alternative has managed to capture. Its persistence across rival camps suggests a shared referent, even amid deep disagreement about what that referent is.

The clarification literature shows that the phrase bundles distinct notions: qualitative feel, perspectival givenness, and a relation to a subject. These behave differently under analysis. Disentangling them does not dissolve the hard problem, but it does sharpen it, showing which component resists explanation and which may yield to functional or representational accounts.

Perhaps the deepest lesson is methodological. Subjective character may be the one phenomenon where our very concept is part of what needs explaining. To ask what 'what it is like' really means is already to ask what kind of beings we are, such that this phrase feels both unavoidable and impossible to cash out.