The Dog and the Zombie
Susanna Siegel *
According to Mike’s disjunctivism, when someone hallucinates a sausage, their experience consists merely in the property of being indiscriminable from a veridical perception of a sausage. This is (III) on Mike’s handout. As Mike understands indiscriminability, this property is a negative and epistemic one: the subject could not know by reflection that she is not veridically perceiving a sausage.
Mike thinks that this view of hallucinations strikes many people as unbelievable. The supposedly unbelievable claim is a conditional:
I à E,
..where this means, if an event has the property of being indiscriminable from a veridical perception, then it is an experience. It’s built in to an event’s being an experience that it has some phenomenal character. (Strictly speaking IàE is a schema. You fill in a more specific indiscriminability property, such as the property of being indiscriminable from a VP of a sausage, and you get a correspondingly specific experience, such as the experience as of a sausage. So Mike’s view is about the properties that constitute hallucinations – a different I property for each specific hallucination).
Mike’s paper is an attempt to articulate why someone might have a hard time believing IàE. I’m going to focus on two objections to it that Mike discusses, the zombie objection and the dog objection, and on what he says in reply to them.
For the purposes of the argument, a zombie is anyone who lacks the sort of phenomenal consciousness that typically accompanies perceptual experience. The other part of the usual definition of zombiehood (that it’s a physical or merely functional duplicate of a normal perceiver) is not relevant. Nor is the stipulation that zombiehood involves complete absence of consciousness. So a zombie in the sense here could equally well be a blindsighter. The argument is then this:
Z1. A zombie could have I.
So
Z2. I doesn’t suffice for E.
So
Z3. (III) is false.
(If the ‘zombie’ were a blindsighter, it would presumably be one who mistook her non-phenomenal state for a phenomenal one). Mike states the objection when he says (just before section 2, p. 8):
“The objection is surely that it is incredible to suppose that being phenomenally conscious like this, indicating the way one is now (indifferently between perception and hallucination) could just be a matter of the obtaining of a certain [negative] epistemological condition.”
In order to succeed, Mike’s reply to this objection has to give us a way of understanding the indiscriminability property whereby it suffices for phenomenal consciousness. His strategy here is indirect. He assumes - and he assumes that the objector will agree – that having the indiscriminability property will suffice for having some sort of self-awareness. Let’s call this conditional that Mike thinks should be common ground
Ià S.
He then identifies as the point of substantive disagreement the claim that the self-awareness in question in turn suffices for phenomenal consciousness.
Sà E.
Here is a passage from the paper in which Mike seems to be describing SàE as the point of disagreement.
“To reject the objection the disjunctivist must insist that the obtaining of the negative epistemological condition in itself guarantees the presence of subjectivity…One way of putting the disagreement here…is to think of it in terms of contrasting conceptions of phenomenal consciousness, and how such consciousness relates to self-awareness and self-conscious knowledge of experience…
Underlying the objection…is a picture of phenomenal consciousness…as independent of one’s recognition that one is having experience…It is the presence of this independent condition which the disjunctivist seems to be denying is present in the case of…hallucination.
[In reply,] [t]he disjunctivist may insist that we cannot conceive of phenomenal consciousness independent of the higher-order perspective that can be taken on it. And hence to conceive of someone’s position as being indiscriminable for them from veridical perception is thereby to imagine them as properly having a point of view and hence subjectivity.”
By employing this strategy and identifying SàE as the fundamental disagreement, Mike is assuming that the other conditional, IàS, is unproblematic. If he thought that it too was contentious, then there would be two stages of the reply to the objection, instead of just one.
What kind of self-awareness does Mike think having the indiscriminability suffices for? In the passage above he characterizes this alternately as ‘recognition that one is having experience’ and a ‘higher-order perspective that can be taken on phenomenal consciousness’. No matter how this is further spelled out, IàS will need some more defense. Here are two prima facie challenges to it.
First, the indiscriminability property as Mike defines it is a negative epistemic property: not being able to know by reflection that one is not veridically perceiving. How could an inability to know that I’m not veridically perceiving suffice for positive self-awareness of any kind, such as knowing that I am having an experience, or that I have the negative epistemic property?
Regarding this challenge, two distinctions are useful to keep in view. The first is between trivial and non-trivial indiscriminability properties. Suppose you’re unable to be aware anything at all – say because you’re a rock. Then you have all the trivial indiscriminability properties. In contrast, you don’t get to have non-trivial indiscriminability properties just because you have these inabilities. Something different needs to be true of you in order to have them. (Non-trivial indiscriminability properties need more explanation that just the gloss on I given.)
The second distinction that is useful to keep in mind is the distinction between having the capacity to be aware of things, on the one hand, being in a state of being aware of them, on the other.
You might think that you can have non-trivial indiscriminability properties only if you have the capacity to be aware of things. But that alone won’t get you IàS. S requires that you be in a state of being aware of something. Just having the capacity for that isn’t enough.
Now for the second challenge to IàS. This challenge arises in light of other commitments of disjunctivism. By Mike’s lights, the nature of the hallucination differs from the nature of a veridical experience. (They belong to different ‘fundamental kinds’). This at least suggests that if the self-awareness for which I supposedly suffices is a contentful state – say, a state of knowing by introspection that one is having an experience – then the content of introspection-based knowledge would differ. But that’s incompatible with the indiscriminability property sufficing for knowing that one has an experience. Perhaps Mike will deny the first step, and say that despite the differences in kind between veridical and hallucinatory experience, the content of self-knowledge in each case is the same. (Perhaps it is knowledge that one has the indiscriminability property). If that’s the reply it would be useful to hear (a) what that single content is supposed to be exactly, and (b) what reasons there are to think that we have states with that content in the case of veridical perception.
I suspect that Mike will appeal to the ‘insubstantial’ nature of knowledge that one is having an experience (or to knowledge that one has the negative epistemic property). The insubstantial nature is illustrated by this contrast: the cup can be on the table without my knowing that it is, but (Mike thinks) I can’t have an experience (or the negative epistemic property) without knowing that I do. The cup-knowledge is substantial, the experience-knowledge is insubstantial. So the insubstantial status of the knowledge comes to either EàS (if the thing I know is that I have an experience), or IàS (if the thing I know is that I have the negative epistemic property). Question: How does this help answer the zombie objection? EàS seems irrelevant to the dialectic; IàS is relevant, but what’s the argument for it?
I now turn to the dog objection and the notion of impersonal indiscriminability that Mike invokes to answer it.
The dog objection is that by disjunctivist lights there seems to be no difference for the dog between any two hallucinations. Since the dog can’t reflect on his experience, he can’t know anything by reflection, and a fortiori he can’t know that he’s not veridically perceiving a carrot, or a sausage. So the dog seems to have every indiscriminability property all the time! This result hits hardest in the case of hallucination, where the disjunctivist says the indiscriminability property is all there is to the dog’s phenomenal character. (In the veridical cases, at least by disjunctivist’s lights, there are the world-involving properties to differentiate the dog’s experiences).
Mike’s reply to the dog objection is that it fixes on the wrong kind of indiscriminability property. In terminology from before, it fixes on trivial indiscriminability properties. The relevant privation of ability, Mike says, is not the dog’s privation, but rather some impersonal privation.
“…the relevant claim of indiscriminability to fix the kinds of experience which John [or the dog] are having are the impersonal or objective such claims: that is we are interested in the claim that [they] are in a situation for which it is impossible simpliciter and not just impossible for John [or the dog] to tell apart through introspective reflection from a veridical perception [of a sausage].” (p. 10).
“We are not to suppose that its being impossible to know through reflection that x is not one of the F’s is true in relation to … the dog’s experience, because of certain further psychological facts…(p. 11)…”
These passages state that there is such a thing as impersonal, objective indiscriminability properties. But they do not tell us what these are. Mike’s suggestion seems to be that we should understand this by comparison with invisibility. We can make sense of the idea that a bit of mending on a jacket objectively invisible, by considering whether someone ideally placed to see things – an ideal visual perceiver in ideal circumstances - could see the mending or not. If not, then the mending is invisible. By analogy, he seems to suggest, we can make sense of the idea that an experience is objectively indiscriminable from a veridical perception of a sausage while being objectively discriminable from a veridical perception of carrots, by considering whether someone ideally placed to reflect on experience – an ideal introspector – could know that it’s not a veridical perception of a sausage. If they could not know this, and if they could know that it’s not a veridical perception of a carrot, then the experience is (non-trivially) indiscriminable from a veridical perception of a sausage.
Here are three counterfactuals that could make explicit what the dog’s having the impersonal indiscriminability property consists in:
1.If the dog could ideally reflect on his experience, he would not be able to know by reflection that he was not veridically perceiving a sausage.
2. If I could ideally reflect on the dog’s experience, I would not be able to know
by reflection that he was not veridically perceiving a sausage.
3. If an ideal introspector were in the dog’s experiential situation, she (the introspector) would not be able to know by reflection that she was not veridically perceiving a sausage.
(Note that if the bracketed part is supposed to be there, then any of these explications of the indiscriminabilty property would also purport to explicate how it is that positive knowledge can come of privation of knowledge.)
If one of these explications is a way to understand what impersonal indiscriminability properties are, then given Mike’s claim that IàE, it must support this conclusion:
C. The dog has an experience as of a sausage.
None of these counterfactuals, however, seems strong enough to support the conclusion. Consider the counterfactual: If a rock could see as well as I can (with my glasses), it couldn’t tell by looking that there’s invisible mending on my shirt, but could tell that it is white. That counterfactual seems true. But we wouldn’t conclude that (say) the color of my shirt is visible to the rock in the impersonal sense, while the mending is invisible to the rock in the impersonal sense. We certainly wouldn’t conclude that the rock has a point of view on the world or any subjectivity. If the holding of 1, 2 or 3 is all it takes to have such an indiscriminabilty property, then it is doubtful that having I suffices to give the dog ‘subjectivity’ or ‘a point of view on the world’. This leaves us with a fresh doubt that IàE. (And, for that matter, a fresh doubt that Ià S). The fresh doubt arises from the way that Mike seems to want to explicate indiscriminability.
If Mike denies that the holding of 1, 2 or 3 suffices for the dog to have an experience as of a sausage, while holding on to Ià E, then there must be something else that impersonal indiscriminability properties are, if they are anything at all. Mike has left it a bit obscure what these properties are. He denies that sausage-hallucinations involve underlying robust properties (akin to the look shared by a lemon and a lemony soap) in virtue of which they are indiscriminable from sausage-perceptions, since if they had such robust properties, that would violate claim (III), that the hallucinator’s experience consists merely in a negative epistemic property. That suggests that they should be understood as some sort of counterfactual about reflection. But Mike also denies that having an indiscriminability property involves the actual employment of a mechanism of reflection. (He says, in speaking of impersonal indiscriminability properties, that “nothing requires that the restriction ‘through reflection’ actually indicate a given mechanism being employed.” (p.14).) That suggests that the antecedent of the conditional can’t invoke any mechanism or process. Perhaps, then, 3 is closest to capturing what the property is supposed to be.
At this point a question arises about counterfactual 3, which talks about ‘being in the dog’s situation’. What is it to ‘be in the [hallucinating] dog’s situation’? To be in this situation can’t be to have a robust phenomenal character, by Mike’s lights. And it can’t to be simply to have the property of being impersonally indiscriminable from a VP of a sausage -- or else 3 would mean the same as the trivially true 3*:
3*. If an ideal introspector couldn’t know by reflection that she wasn’t veridically perceiving a sausage, then she couldn’t know by reflection that she wasn’t veridicially perceiving a sausage.
If we’re driven to 3 (or similar counterfactuals with the same antecedent), then apparently there is no non-trivial account of what impersonal indiscriminability properties are, or else Mike is illicitly relying on a notion of robust phenomenal character after all. To avoid 3’s antecedent, apparently we need to talk about actual employment of introspective mechanisms after all.
Overall the prospects for making the claim that there are impersonal indiscriminability properties intelligible seem dim.