I submitted my PhD thesis titled Relationalism in the Face of Hallucinations in October 2016, at the University of Warwick and at Paris-1, in the context of a co-tutelle program. Discussed in December 2016, no correction.
My supervisors were Prof. Matthew Soteriou (Warwick) and Prof. Maximilian Kistler (Paris-1).
In my thesis I defend relationalism in the face of the argument from illusion and, most importantly, hallucination. For relationalism the phenomenal character of perception should be accounted for in terms of a non-representational relation of acquaintance with the objects one seems to perceive.
It is often believed that illusions and hallucinations pose a serious threat to this view, which is deemed unable to account for these phenomena and for the fact that they are or can be introspectively indistinguishable from genuine perception.
Although this line of argument is pervasive in the history of philosophy and in the current debate, it is not clear exactly how this argument is supposed to threat relationalism. Various versions have been proposed, but so far a valid and compelling argument has yet to be offered.
With my thesis I aim to fill this gap: I formulate a novel challenge that seriously threatens relationalism and that relationalists ought to take seiously. Once clarified what’s really the source of the troubles for relationalists, I explain how they should respond to this challenge.
Very roughly, this challenge goes as follows:
The relationalists need to explain how the phenomenal character of a perception and an introspectively indistinguishable hallucination can be different without committing to the unacceptable claim that phenomenal characters have some sort of depth, that they are not superficial.
We can frame the source of the challenge in the form of a dilemma:
(1) On the one hand, we have the intuition that phenomenal characters are superficial.
This means (according to the understanding of superficiality I advance in my thesis) that their nature must be accessible through introspection.
(2) On the other hand, relationalism is committed to claimnig that, even when hallucinations are indistinguishable from perception, their phenomenal character is not the same as that of perception. But this seems to entail that the real phenomenal character of the indistinguishable hallucination is hidden to introspection- thus not superficial.
So it seems that either one denies the superficiality of phenomenal character or one forsakes relationalism. To avoid this dilemma, the relationalist needs to explain how these two apparently incompatible claims can be reconciled.
A quite straightforward answer to this challenge is available when we turn to ordinary hallucinations, those documented by psychopathology. The relationalist can claim that the inability to distinguish one’s current hallucination form the class of perception is due to the occurrence of conditions which interfere with the subject’s normal abilities to introspect (such as altered states of mind, cognitive impairment, neurological malfunctioning).
So ordinary hallucinations really do not pose problems for the relationalist. What causes a problem are merely hypothetical cases of hallucinations: cases where one has a hallucination that is brought about through neural replication of the brain state one is in when genuinely perceiving something. In these cases, ex hypothesis all is fine with the subject’s brain and cognition, so nothing interferes with the subject’s introspective capacities. We should say that these hallucinations are impersonally indistinguishable through introspection from the target perception. This makes the explanation given for ordinary hallucinations unavailable in this case.
So relationalists need to explain how they can maintain that a perception and a hallucination that is brought about through neural replication have different phenomenal characters. But it is not obvious how they can do so without denying the superficiality of phenomenal characters. In the absence of a response to this challenge, relationalism is undermined.
Ultimately I argue that a response to this challenge can be derived from the way Hinton formulates disjunctivism in his seminal work, and from Martin's further proposal of a negative account of hallucination. The phenomenal character of an impersonally indistinguishable hallucination through introspection simply consists in the fact that it seems to one as if one were perceiving a certain object (but one doesn’t know than one is not perceiving). This does not violate the superficiality constraint because the fact that it seems to one as if one were perceiving is accessible through introspection.
But what I take to be the most relevant aspect of my discussion is that it reconfigures the dialectic of the debate around relationalism and the argument from hallucination. It shows that both sides in the debate have so far overlooked what’s really at stake in the dialectic, and what’s the real source of the problem for relationalism. My reassessment of what goes on–or should be going on–in the debate undercuts a series of arguments against relationalism that are often considered very convincing. One example concerns the notion of impersonal indiscriminability. Many have objected to a version of disjunctivism—which I argue a relationalist should adopt—because it employs a notion of impersonal indistinguishability, and this notion is incoherent. I argued that this notion is coherent, but most importantly, I have shown that the proponent of the argument from hallucination needs to help herself to this notion, or else the argument from hallucination doesn’t even get off the ground. So if it turns out that the notion is indeed incoherent, that would be a problem for the proponent of the argument from hallucination and relationalism would be safe.
Chapter 1
I introduce relationalism and its commitment to disjunctivism, the claim that perception and hallucination have different phenomenal characters, even when they are subjectively indistinguishable. I argue that relationalism accounts for how perception strikes us introspectively in a less revisionary way than its opponents. Qualia theorists and representationalists need to deny the manifest relationality of perceptual experience, while sense-datum theories need to commit to the claim that we are systematically wrong with respect to the nature of the object one perceives.
I also discuss the notion of phenomenal character and reject an understanding of it that would make it incompatible with a relationalist account.
Chapter 2
I outline the general structure of the argument from illusion and hallucination, composed of:
-
a base case, where it is claimed that relationalism doesn’t apply to certain cases of experiences (typically, cases of conflicting appearances, illusions, and hallucinations).
-
and a spreading step, which generalises the conclusion for the base case to all experiences, including genuine perception.
I argue that the base case fails for conflicting appearances and illusions, for which relationalism can indeed offer an account. But it can be established for hallucinations. However it is not clear what motivates and support the spreading step. In the traditional version of the argument the spreading step is motivated by the indistinguishability principle. This claims that when two experiences are introspectively indistinguishable, they have the same phenomenal character.
Chapter 3
I look at how this principle can be motivated. I reject two traditional strategies: one that relies on the idea that introspection is infallible and one relying on the idea that phenomenal characters do not allow for a seem/is distinction. Instead, I suggest that the indistinguishability principle relies on the intuition that phenomenal characters are superficial and that relationalists are committed to violating the intuition that phenomenal characters are superficial I spell out what is entailed by the superficiality of phenomenal characters and, with this understanding in place, I formulate the challenge for relationalism that I outlined at the beginning of this presentation.
In chapter 3 I also argue that the superficiality constraint may be used to support the indistinguishability principle only if the notion of indistinguishability that features in the principle is understood in an impersonal way. It remains to see whether there are indeed hallucinations that are introspectively indistinguishable from perception in an impersonal way.
Chapter 4
I argue that there is no available empirical evidence for the claim that ordinary hallucination documented by psychopathologists are indistinguishable in the relevant sense from perception. And actually, it is not clear what kind of evidence could support that claim.
This brings about two general lessons for the current debate. First, it recommends scepticism for the idea that one can provide a comprehensive account of the hallucinations, which are more varied in their phenomenology and aetiology than philosophers tend to recognise. Second, it undercuts a widespread way of presenting the argument from hallucination as a phenomenological argument, where the problem seems to be that relationalism cannot accommodate certain quite widespread and documented phenomena: hallucinations.
Chapter 5
I consider how the proponent of the argument can introduce the metaphysical possibility of hallucinations that are impersonally indistinguishable through introspection helping themselves to the idea of brain replication. I argue that two main arguments to support this hypothesis (the argument from local supervenience and the argument from the same proximate cause, same immediate effect) beg the question against relationalism. However, I conclude that it is not clear how the relationalist could deny this possibility and so we should concede it to the opponent.
Chapter 6
I present my own solution to the challenge I set out in the rest of the thesis. After considering and ruling out less promising options (including one that appeals to the multiple realisability of phenomenal characters), I argue that the relationalist can answer the challenge I set up by adopting a negative view of hallucination. I discuss how we should understand this proposal, by comparing it with that of eminent proponents of it, such as Martin and Fish, and I respond to some standard objections to this view, helping myself with the conceptual tools I put in place throughout the previous chapters. The response to these objections largely rests on understanding the relation between introspection and phenomenal character and between introspection and perception. The key ideas of this proposal are that (1) the phenomenal character of experience doesn’t exist prior to and independent from one’s introspective access to it and (2) the subject’s introspective perspective on her experience depends on her perceptual perspective on the world. This concludes my defence of relationalism.