Luminosity and the KK Principle

Core Idea

The luminosity thesis holds that whenever one has a mental state, one is in a position to know that one has that state. Closely related is the KK principle: if you know something, you know that you know it. Williamson argues both principles face insurmountable obstacles given the margin-for-error account of knowledge and the vagueness of mental state boundaries.

How It's Best Learned

Test luminosity with subtle mental state differences: can you always detect when you're certain versus almost certain? When you believe something, are you always positioned to know you believe it? Examine Williamson's margins-for-error argument against these principles.

Common Misconceptions

• Rejecting luminosity doesn't mean introspection is entirely unreliable. - The KK principle is stronger than simple reflection; it requires knowledge of knowledge, not just awareness. - Luminosity concerns only whether one is 'in a position' to know; it doesn't claim such knowledge is easily obtained.

Explainer

The luminosity thesis captures an intuitive picture of the mind as self-transparent: mental states are "lit from within," fully available to the subject who has them. Whenever you are in pain, you know you are in pain. Whenever you believe something, you know you believe it. Whenever you feel certain, you know you feel certain. This seems to capture something important about the first-person perspective — unlike the external world, where evidence can be misleading or incomplete, your own mental states seem directly accessible in a way nothing else is. The closely related KK principle (if you know P, you know that you know P) extends this to knowledge itself: your epistemic states are also transparent to you.

Your background in justified true belief gives you the resources to see why luminosity is philosophically significant. If knowledge requires justification, and you can always know whether you have a mental state, then your beliefs about your own mental states are always justified from the inside. This would mean introspection is a privileged epistemic method — not infallible, perhaps, but systematically more reliable than perception of the external world. Many traditional epistemological frameworks assume something like this: Descartes' cogito and Locke's inner sense both presuppose that the mind has special access to itself. Luminosity is the contemporary formulation of this assumption.

Timothy Williamson's argument against luminosity in *Knowledge and Its Limits* (2000) is the most influential challenge. He uses a margins-for-error argument built on the gradual change of mental states. Consider warmth: imagine you cool down very slowly over many hours, one imperceptible degree at a time. At the start you feel warm; at the end you feel cold. At every intermediate point, your feeling is so close to the adjacent states that you cannot reliably distinguish "I feel warm" from "I feel slightly-less-warm-than-warm." Williamson argues that knowledge requires a safety margin — you can only know P if, in nearby possible situations, you also believe P and it is true. But given the gradual change, at every point where you believe you feel warm, nearby situations include ones where you no longer feel warm but still believe you do. So you never *know* that you feel warm; you only believe it. Since this argument applies to any gradually-changing mental state, luminosity fails generally.

The implications are striking: your higher-order evidence about your own mental states is not automatically authoritative. You might believe you are certain of something when you are actually only fairly confident; you might believe you are experiencing pain when you are experiencing something that would not quite count. This does not mean introspection is worthless — it means introspection is more like perception than traditional epistemology assumed, subject to the same kinds of margins for error and reliability constraints. The KK principle falls for the same reason: knowing P does not guarantee knowing you know P, because the standards for second-order knowledge may not be met even when first-order knowledge is. The luminosity debate thus reopens questions about what is distinctive about self-knowledge and whether the first-person perspective really does carry any special epistemic privilege.