Philosophical Zombies
A p-zombie is a hypothetical creature molecule-for-molecule identical to a conscious human, behaviorally indistinguishable in every respect, but with no inner experience (see the SEP entry). No “what it’s like” to be it. The lights are off. David Chalmers uses the conceivability of such a being as a modal argument against physicalism.
Why does this threaten physicalism?
If a physical duplicate of you could lack experience, then experience isn’t fixed by the physical facts, and physicalism owes us an account of why not.
The conceivability argument:
- Zombies are conceivable: nothing we can know a priori rules them out; the idea is coherent
- Whatever is (ideally) conceivable is metaphysically possible
- Therefore zombies are possible
- If zombies are possible, physicalism is false, because consciousness can’t be identical to or entailed by the physical facts (a world physically identical to ours could lack it)
The point: if consciousness is just the physical story playing out, the zombie is incoherent. It isn’t obviously incoherent. So consciousness isn’t just the physical story.
Chalmers’s hard problem is explaining why there is something it’s like to be a system that processes information at all. Easy problems (attention, discrimination, reportability) are about functions a brain performs. The hard problem is the extra fact of subjective experience riding along with those functions. P-zombies dramatize the hard problem: a being that does all the functional work, with no extra fact. If that’s coherent, the extra fact is extra, not determined by the functional/physical facts alone.
Main objections (per SEP):
- Type-B physicalism (Hill, Loar, McLaughlin): grant conceivability, deny it entails possibility. There are a posteriori necessities, truths that hold in all possible worlds but can’t be known from the armchair (e.g., “water is H₂O”). Mind-brain identity may be one of these: zombies seem conceivable but aren’t actually possible
- Phenomenal concepts strategy: the explanatory gap between physical description and experience reflects only a conceptual difference. We have two kinds of concepts (physical, phenomenal) that pick out the same underlying property. The gap is in our concepts, not in the world
- Dennett: p-zombies aren’t actually conceivable on close inspection; the thought experiment sneaks in hidden contradictions. If a zombie behaves identically to a conscious person (discusses its experiences, reports its pain, etc.) the hypothesis that it has no experience is doing no real work, or is incoherent
- Deny the inference from conceivability to possibility in general: plenty of coherent-seeming scenarios (Goldbach’s conjecture being false, say) might be genuinely impossible. Why trust conceivability here?
Three pokes at the same gap
Mary’s Room, Nagel’s bat, and zombies are pointing at the same pressure point from different angles:
- Mary (knowledge argument): has all physical facts about color, learns something new on seeing red. → physical facts don’t exhaust the facts
- Nagel’s bat (“What Is It Like to Be a Bat?“): physical description of echolocation doesn’t convey what bat experience is like. → physical description leaves something out
- Zombies: a physically complete duplicate could lack experience. → consciousness isn’t fixed by the physical
My take
I’m suspicious of the conceivability-to-possibility step, which is step 2, the load-bearing one. “I can imagine X” is a weaker claim than “X is metaphysically possible in principle,” and the history of science is full of things that seemed clearly conceivable and turned out not to be (instantaneous action at a distance, absolute simultaneity, the luminiferous ether). I don’t know why consciousness should be the domain where my imagination is a reliable guide to modal reality.
That said, Dennett’s reply feels like it’s dodging rather than answering. Saying “zombies aren’t really conceivable” requires you to have already decided that functional equivalence is consciousness, which is the thing under debate. He’s entitled to that position, but the zombie argument is aimed at it, not refuted by it.
Where I end up, tentatively: the zombie argument is the cleanest statement I know of why consciousness feels like an extra fact, and I think that feeling is tracking something. But I don’t think conceivability arguments can actually establish metaphysical conclusions about consciousness, and the Chinese Room is a better way to press the same intuition because it shows a specific mechanism (syntactic rule-following) failing to produce meaning, rather than stipulating the failure by thought experiment.
This pairs with my note on Mary’s Room: the knowledge-argument, the zombie argument, and the Chinese Room are three pokes at the same pressure point, that there seems to be something about first-person experience that doesn’t reduce. Whether that “seeming” is an ontological fact or a stubborn feature of our concepts is, as far as I can tell, genuinely unresolved.