Can You Conceive of a Philosophical Zombie… or a Million-Sided Object?

i) Introduction ii) Conceivability iii) Philosophical Zombies iv) Four Conceivings: 1) A Round Square 2) Descartes’ Chiliagon 3) Goff’s Megagon 4) Water Being XYZ v) Saul Kripke on Misconceivings vi) Conclusion: Philip Goff on Misconceivings

Conceivability has played a vital part in much contemporary (mainly analytic) philosophy. This is the case specifically in metaphysics. It’s also very important to the work of the Australian philosopher David Chalmers and his championship of what came to be called “philosophical zombies”.

Much of this interest in conceivability begin with the work of Saul Kripke in the 1960s and 1970s. Yet Kripke himself acknowledged the possibility of people misconceiving things which they falsely believe they’ve actually conceived. And David Chalmers too acknowledges this possibility of misconceiving (metaphysical) possibilities.

But the first thing which needs to be done is to distinguish conceiving from imagining. This is an important distinction which many philosophers — since Descartes — have often stressed. Here’s one account of that distinction:

“[Descartes] first distinguishes between imagination and pure understanding. In the case of a triangle, he can perceive that a triangle is three-sided and derive all sorts of other properties using the understanding alone. He can also perceive these properties with the imagination, by picturing the triangle in his mind’s eye. However, the weaknesses of the imagination become clear when he considers a thousand-sided figure. It is very difficult to picture it in his mind’s eye, and more difficult still to differentiate that mental image from the mental image of a 999-sided figure. The pure understanding, however, dealing only in mathematical relations, can perceive all the properties of a thousand-sided figure just as easily as it can a triangle.”

As can be seen, the words above are all indebted to Descartes’ distinctions between imagining and conceiving (or “pure understanding”); as are the distinctions of David Chalmers and Philip Goff which follow.


To put it basically, some philosophers argue that (philosophical) conceivings don’t depend on the formation of any mental images whatsoever. Therefore such conceivings can be seen as being a more sophisticated form of …. well, for want of a better word, imagining. That is, conceiving x is (at least in part) imagining x without employing any mental imagery. (If that isn’t in itself an oxymoron.)

The general philosophical argument is that if we can conceive of any given x, then that x is possible. (Distinctions between logical possibility and metaphysical possibility will be made later.)That is, the fact that we can conceive of x, means that x is possible.

But why on earth should that be the case?

(Note: Most of the following quotes from Philip Goff are taken from his paper ‘The Phenomenal Bonding Solution to the Combination Problem’.)

In more technical terms, the English philosopher Philip Goff (more of whom later) expresses the relation between conceivability and possibility in this way:

“If P is conceivably true, then P is possibly true.”

This is also expressed in possible-worlds jargon as follows:

“If P is conceivably true (upon ideal reflection), then there is a possible world W, such that P is true at W considered as actual.”

Or, less technically, Goff also that

“Chalmers holds that every conceivably true proposition corresponds in this way to some genuine possibility”.

Goff adds that

“conceivability entails possibility when you completely understand what you’re conceiving of”.

And then Goff goes much further than all the above.

Not only is his argument that the conceiving of x is a reason for believing that x is metaphysically possible, Goff also argues that it may be the case (or is the case) that

“metaphysical possibility is just a special kind of conceivability”.

Note the use of the (possible) “‘is’ of identity” in the sentence above. We’re told that metaphysical possibility literally is (“a special kind of”) conceivability. Thus it’s not just that our conceiving of x may — or does — give us a reason to believe that x is possible. It’s also the case that this very conceiving of x seems to (as it were) bring about the metaphysical possibility of x.

Despite all that, Goff seems to assume that there’s a determinate and precise meaning of the words “conceivably” and “conceivably true”. Yet conceivability-to-possibility arguments may not get off the ground (at least in some cases) in the first place. And that’s because nothing at all is really conceived. Alternatively, that which is conceived of is literally unbelievable (as Goff himself acknowledges later).

So it can be argued that the least a philosopher can argue is that conceivability is a (rough) guide to possibility. This lesser claims would an alternative to outrightly stating that because subject S has conceived of any given x, then that means that x is possible. Of course we’ll now need to know what the words “guide to possibility” mean and how, exactly, conceiving of x is a guide to its possibility.

So let’s be more concrete about conceivability by bringing in philosophical zombies.

Philosophical Zombies

What is a philosophical zombie?

This is one account:

“A philosophical zombie or p-zombie is a hypothetical being that is physically identical to and indistinguishable from a normal person but does not have conscious experience, qualia, or sentience… Proponents of philosophical zombie arguments, such as the philosopher David Chalmers argue that since a philosophical zombie is by definition physically identical to a conscious person, even its logical possibility would refute physicalism, because it would establish the existence of conscious experience as a further fact.”

How would a philosopher (in a strict sense) know that he’d conceived of a philosophical zombie? What if you or I can’t conceive of a phil-zombie even if other people can do so? And even if you and I did (seemingly) conceive of a p-zombie, then how could we then know that we’d done so? Indeed wouldn’t the best philosophical policy be to accept (or believe) that we hadn’t actually conceived of a p-zombie?

Isn’t the very notion of conceiving of any given x far too psychologistic (or mentalistic) an idea to be sustaining metaphysical theses? Moreover, if other minds are a philosophical “problem”, then other minds conceiving of p-zombies (or anything else) is a problem within an already-existing problem.

More concretely, if we actually came across an entity which acted and looked — in every single detail — like an ordinary human being (which is the whole point of the p-zombies thought experiment), then how could we extract from all this that it may, in fact, be a p-zombie? If we ever did come across a p-zombie which acted and looked just like us, then the obvious conclusion would be that it had consciousness (or experiences) just like us. In other words, we’d assume that this “zombie” felt pain, sometimes gets angry, could smell garlic, etc.

Moreover, when Mary supposedly conceives of a p-zombie, how does it differ from her conceiving of her mother, father or best friend? After all, when she conceives of the latter, she can’t literally conceive of his/her consciousness or of them actually being conscious. Instead, all she’s got to go on is their physical and verbal behaviour. And that’s also true of p-zombies.

Perhaps that’s Chalmers’ point!

If it is Chalmers’ point, then let’s spell it out again.

Conceiving of a p-zombie is no different from conceiving of your mother, father or best friend.

So conceiving of a p-zombies is no different from conceiving of your mother, father or best friend and perhaps that’s partly why Chalmers believes conceiving of a p-zombie is possible. And if conceiving of p-zombie is possible, then its actual existence is (metaphysically) possible.

But does all that actually work?

When we conceive of a p-zombie (just as when we conceive of our mother, father or best friend), we don’t conceive of an entity (or being) actually lacking consciousness because such a thing can’t be conceived. We’re supposed to be capable of conceiving of a being (or entity) behaving like our mother, father or best friend — though not instantiating consciousness. However, how is that different to conceiving of one’s mother, father or best friend? And it’s no different because we can neither conceive of the consciousness of a fellow human being nor of a p-zombie. In other words, consciousness is “known” because it’s behaviourally or physically expressed. Unexpressed consciousness, on the other hand, is not conceivable. So it’s just as hard to conceive of another being’s consciousness as it is to conceive of its lack of consciousness.

Of course Chalmers — for one — never really discusses coming across real (or actual) p-zombies because the whole point of his thesis is the importance of conceiving of them. He’s also keen to show us how that very conceiving establishes certain other(!) philosophical positions. In this case, that other philosophical position is that consciousness (or experience) is over and above everything physical and functional. In other words, even though this conceived of (or real) p-zombie behaves and looks just like a normal human being, it still may not (or does not) instantiate consciousness at all. This means that, for all intents and purposes, a p-zombie is a (as it were) robot or machine in human clothing.

Despite all the above, can the contrary ever be categorically stated?

Can it be stated that no one could ever conceive of a p-zombie (or a “p-zombie world”)? In other words, if Chalmers, etc. can’t categorically claim that they have conceived of a p-zombie, then perhaps we can’t categorically claim that they haven’t done so. How could you, I or anyone else know what Chalmers has (or hasn’t) conceived of? Consequently, this may mean that all the sceptical questions aimed at such conceivings can now be aimed at the sceptics about such conceivings.

Yet that last conclusion surely demonstrates the central point of this issue.

That point being that since we’re talking about mentalistic conceivings, whether subject S has (or hasn’t) truly conceived of a p-zombie can never be established. And if it can never be established, then how can such conceivings provide the basis of philosophical theses?

This is a good place to look at some other (philosophical) conceivings.

Four Conceivings

A Round Square

Here’s an extreme example of the conceivability-to-possibility argument:

i) If we can conceive of a round square, ii) then it is possible that there is a round square.

This is a rejoinder to that argument:

i) We cannot conceive of a round square. ii) Therefore round squares are impossible.

But what about this? -

i) If we can conceive of p-zombies, ii) then p-zombies are possible.

Here’s the counter-argument:

i) We can’t conceive of p-zombies. ii) Therefore p-zombies are impossible.

Of course the examples of a round square and a p-zombie are very different (as are a megagon and p-zombie). However, this round square example was chosen simply to get the point across.

In other cases, the move form conceivings to metaphysical possibilities may well be legitimate. Alternatively, every move from conceivability to metaphysical possibility may be somewhat suspect!

Descartes’ Chiliagon

The definitional quote in the introduction above mentions a “999-sided figure”; whereas René Descartes himself referred to a chiliagon - which is a 1000-sided polygon. (I presume the odd and arbitrary number of 999 sides — rather than the 1000 sides of a chiliagon — was chosen to emphasise our imaginative limitations.)

Even if a chiliagon were massive in size, it would still be visually indistinguishable from a circle. It can also be argued that a chiliagon can’t be conceived of either — even if we do have a concept of it. Though that, of course, will depend on what’s meant by the words “conceived of”. It can be argued that mathematicians (or geometricians) can conceive of a chiliagon (to lesser or greater degrees); though a layperson can’t.

Moreover, a chiliagon is a mathematical/geometrical abstract object; not a concrete object. In other words, it couldn’t be found or even made. Indeed both the chiliagon and megagon are not a constructible polygons. Nonetheless, that doesn’t stop them from being well-defined concepts.

Philip Goff’s Million-Sided Object

Philip Goff himself cites the example of a “million-sided object” to stress the distinction between conceiving and imagining any given x.

The megagon (a million-sided polygon) is obviously similar to Descartes’ example of the chiliagon in that it’s deemed to be a “well-defined concept” which, nevertheless, cannot be visualised.

In one sense it can be said that we can indeed conceive of a megagon. Or, more helpfully, if I were to ask someone this question:

What do you conceive of when you say that you’ve “conceived of a million-sided object”?

That person may answer:

I conceived of an object which has a million sides.

But what does that mean? What is he conceiving of? Is he simply assuming the following? -

i) A million-sided object has a million sides. ii) Therefore I’ve conceived of a million-sided object.

Is there any more to it than that? If he simply knows that if an object has a million sides, then he’s conceived of an object which has a million sides?

In any case, no one can picture or imagine a million-sided object. So that’s ruled out for a start. What’s left? Again, the words “conceiving a million-sided object” may seem vacuous. Yet despite all that Goff says that “the concept million-sided object is transparent”. That is,

“it is a priori (for someone possessing the concept, and in virtue of possessing the concept) what it is for something to have a million sides”.


“when one conceives of a million-sided object one completely understands, or is in principle able to reason one’s way to a complete understanding of, the situation being conceived of”.

Goff goes further when he says that

“it is a priori for the conceiver what it is for the state of affairs they are conceiving of [i.e., a million-sided object] to obtain”.

Thus we reach the important conclusion which Goff has been leading up to all along. Namely,

“that we can move from the conceivability (upon ideal reflection) of the states of affairs so conceived, to its genuine possibility”.

Goff’s quote above is simply a rerun of what’s already been said. That is:

i) What is it to conceive of something which has a million sides? ii) It is to conceive of a million-sided object.

Nonetheless, perhaps my position is too psychologistic in nature — i.e., too dependent on our contingent mental states and their content (as Descartes warned us against). Perhaps Goff’s position is purely logical. Alternatively, perhaps Goff’s position is purely mathematical/geometrical (therefore abstract) in nature. That would mean that it may be an entirely logical and/or mathematical point to say that “the concept million-sided object is transparent”.

Then again, what does that claim amount to? Indeed how different is conceiving of a million-sided object to conceiving of a round square? That said, it’s clearly the case that a round square isn’t in the same logical space as a million-sided object. A round square is a logically impossible object. (Or is it conceptually impossible?) A megagon, on the other hand, is not a logically impossible object — even though it isn’t a constructible polygon.

Moreover, even if the words “having a well-defined concept” and “conceiving of” are taken as (virtual) synonyms, it’s still the case that both the layperson and the mathematician would need to conceive of (or have a well-defined concept of) a megagon.

So can Goff — or any other philosopher who isn’t a mathematician — really conceive of a megagon without also (to use Goff’s own words) “completely understanding” all the detail in the following copy-and-paste? Here goes:

The question now is:

Are p-zombies in the least bit analogous to a million-sided polygon (i.e., a megagon) when it comes to conceivability leading to possibility?

Water is XYZ

David Chalmers also offers us an example of conceiving of water being XYZ rather than H₂O.

Again, what does it mean to conceive that water is XYZ?

Surely it’s no use Chalmers going into to further detail if this isn’t established in the first place.

Is water being XYZ conceivable simply because we’re simply imagining (to use (Chalmers’ own words) “watery stuff” (i.e., Chalmers’ “primary intension”) which we also conceive to be XYZ? But are we conceiving of water actually being XYZ (i.e., rather than simply conceiving of water stuff)? Isn’t that something completely different? So here goes:

1) The first act of conceiving has to take on board what XYZ actually is. (Is XYZ meant to be some kind of fictional — though possible — molecule?) 2) And then one needs to conceive of XYZ actually being water.

Colorless Green Ideas Sleep Furiously

Now take Chomsky’s famous surreal sentence:

“Colorless green ideas sleep furiously.”

All the words and their concepts in the sentence above are (to use Goff’s word) “transparent” when taken individually. We can also say that the sentence itself is grammatically and logically correct. We can understand the words within that statement. However, can we also conceive of that sentence being true? Or, more accurately, can we conceive of a (to use Goff’s words) “state of affairs” in which colorless green ideas sleep furiously?

Here it seems that grammatical (or even logical) correctness runs free of conceivability. In other words, perhaps we don’t — and can’t — actually conceive of colorless green ideas sleeping furiously.

But is the same conclusion true of conceiving of a p-zombie?

Saul Kripke on Misconceivings

The American philosopher and logician Saul Kripke cited various examples of people believing that they’ve conceived of something — which they deemed to be possible - and their not actually doing so.

Kripke offered us the example of conceiving that water is not H₂O. He also cited the case of a person seemingly conceiving of a true mathematical theorem to be false or false theorem to be true. (In the former case, Kripke gave the example of Goldbach’s conjecture.)

Kripke — like Chalmers and Goff — claimed that he was working with his own “Cartesian intuitions” when he — specifically — tackled the mind-body problem. And many of those intuitions were also about what is and what isn’t metaphysically possible.

In the first instance, Kripke cited an act of imagination (i.e., he didn’t use the word “conceiving”) which may mislead us. He wrote:

“[W]e thought erroneously that we could imagine a situation in which heat was not the motion of molecules. Because although we can say that we pick out heat contingently by the contingent property that it affects us in such and such way.”

In this case “we” conceived of the effects of the motion of molecules on bodies and the environment. However, we didn’t — or couldn’t — conceive of heat actually being something other than the motion of molecules.

So let’s say that heat is XYZ and that it has nothing to do with molecular motion. That raises the following question:

What is it to conceive that heat is XYZ rather than molecular motion?

Kripke was claiming that those who imagine heat being something that’s not molecular motion aren’t really imagining heat at all. They just believe that they are imagining heat because they’re actually imagining the “contingent properties” of heat — i.e., its effects on persons and the environment.

To repeat: we can’t conceive of such a thing as heat being XYZ. In other words:

1) The argument is not only that there is a radical separation between our conceiving of XYZ being heat (or conceiving p-zombies) and its actual possibility. 2) The argument is also that we don’t actually conceive of such a thing in the first place.

Yet Kripke also argued that our imagination can tell us something both true and important about the world. In Kripke’s own words:

“[J]ust as it seems that the brain state could have existed without any pain, so it seems that the pain could have existed without the corresponding brain state.”

Kripke stressed our ability to imagine a pain state without its correlated brain state (formerly characterised as the “firing of C-fibres”). Thus Kripke (to paraphrase) concluded:

If we can imagine a mental state without its correlated brain state, then such a states of affairs is possible.

Alternatively, Kripke was also arguing that there’s no necessary identity between mental states and brain states.

Conclusion: Philip Goff on Misconceivings

Philip Goff also expresses the possibility that conceivability may not always give us metaphysical (i., not logical) possibility. That is, even if we do allow various moves from conceivability to metaphysical possibility, sometimes what we argue is metaphysically possible still remains unbelievable. Or as Goff himself puts it:

“When metaphysical possibility is so radically divorced from conceptual coherence [] I start to lose my grip on what metaphysical possibility is supposed to be.”

Goff is arguing that — in some cases at least — metaphysical possibility can move too far beyond “conceptual coherence”. That must also mean that — at least some — moves from conceivability to metaphysical possibility are illegitimate. That is, a specific conceiving may not warrant the metaphysical possibility which is supposedly derived from it. To stress that point, Goff also says that

“a radical separation between what is conceivable and what is possible has the potential to make our knowledge of possibility problematic”.

Yet, despite all the above, both Goff and Chalmers seem to provide what they take to be a tight link between conceivability and metaphysical possibility. And, if that’s the case, then how can there ever be a (to use Goff’s words) “radical separation” between the two?

That said, the central position taken in this piece is that if that link between conceivability and metaphysical possibility is — at least in some cases — broken, then that will be (at least partly) due to the fact that such conceivings aren’t really genuine conceivings at all. More clearly, some moves from conceiving to (metaphysical) possibility are completely bogus precisely because the given x wasn’t conceived of in the first place.