Transparency and our introspective grasp on phenomenal properties

Thalabard 2014 – Transparency & intrinsic properties

Transparency and our introspective grasp on intrinsic properties. (22/11/2014, Université Paris-Sorbonne, Workshop on Derk Pereboom) Conscious mental states are those which, according to Nagel's classic quasi-definition are those for which there is a sense to ask 'how they are like' or 'what it is like to undergo them' for the subject enjoying them. In other words, concious mental states have a distinct 'qualitative feel' or 'phenomenal character'. Qualia, or phenomenal properties, are usually introduced as the properties responsible for the state having a particular phenomenal character – depending on one's favourite account of conscious experience, these may be properties of experience itself, or properties of the subject of experience. Phenomenal character is a troublesome feature of conscious experience, wheneve one tries to explain it in natural terms, be them physical or functional : straightforward characterizations of experience in natural terms seem to leave out 'what it is like' to have a conscious experience. Thus the 'hard problem' of consciousness is to explain why certain of our mental states have such a phenomenal feel. That is, explain why certain states are phenomenally conscious at the exclusion of others – which properties of these states correspond to their being conscious ; and why they come to have the distinctive feel they have. Phenomenal properties are traditionally construed as introspectible and intrinsic properties of experience. By 'intrinsic', all that is meant is that those properties are not determined by a subject's environment, or by the relations that the states possessing them have with other mental states : qualia are sui generis properties, that elude any kind of naturalistic explanation. Derk Pereboom diagnoses some of our anti-naturalistic intuitions, making them derive from introspective misrepresentations of the qualities of our experience. Pereboom's discussion of the knowledge argument and his qualitative inaccuracy hypothesis are an intriguing piece of philosophy. All I have to offer here is a brief set of remarks. I will however suggest that the open possibility of,error regarding the nature of phenomenal properties which Pereboom appeals to might be obtained through the transparency thesis, without making introspection as such a source of error concerning the nature of phenomenal properties. Ultimately, my account is simpler than Pereboom's, and cheaper, as it is not committed to any substantial claim regarding our folk concept of introspection regarding phenomenal properties. My talk will be divided in two parts : in the first section, I will present and discuss 1

Thalabard 2014 – Transparency & intrinsic properties

Pereboom's Qualitative Inacuracy Hypothesis (hereafter, QIH) argument against antiphysicalism, and try to evaluate how to understand the idea of introspective inaccuracy, in order for the argument to bear on the question of the physical nature of qualia. In the second part, I will briefly present and discuss the transparency thesis, thereby drawing some conclusions on the limitations of introspective data. I wish to contrast what introspective evidence the transparency thesis affords with the picture which Pereboom presses on us, and will advocate a modest use of introspection in the investigation of phenomenal properties.

1) The Qualitative inaccuracy hypothesis. In his 2011 Consciousness and the Prospects of Physicalism, Derk Pereboom devises an introspective argument, designed to counter the Knowledge Argument. This argument is an expansion of an earlier version (1994). QIH : It is an open possibility that we systematically misrepresent the features of experience which we are aware of through introspection. Here are 2 formulations of QIH : « Thus, even if phenomenal modes of presentation, such as what it is like to sense red, represent sensation a having a character that is not intersubjectively accessible, a modified reply to the knowledge argument, along Lycan-Loar lines is effective. For since mental states might well not be as they appear in introspection, sensations might yet be intersubjectively accessible and wholly physically realized. (…) Perhaps it is plausible to say that the content of an introspective representation might fail to ressemble the mental state it represents. » (1994, p323) « Introspective representations mediate the subject's awareness of these mental states, rendering this awareness in a sense indirect. As a consequence, the subject may represent a mental state as being a certain way, even though it is s not really that way, or at least not as it is in itself. » (2011, p9)

2

Thalabard 2014 – Transparency & intrinsic properties

QIH is directed against one of two current intuitions about introspection and qualia : Intuition 1 : physical and introspective modes of presentation represent a phenomenal property as having a specific qualitative nature ; what qualitative nature the introspective mode of presentation attributes to a phenomenal property, is not included in the physical mode of presentation of that property. Intuition 2 : introspection accurately represents the qualitative nature of phenomenal properties. QIH targets the second intuition. That is to say : according to Pereboom, it is an open possibility that introspection misrepresents the qualitative nature of phenomenal properties. Pereboom's view of introspection is roughly the following : introspection amounts to representing our experience - we access the properties of our experience through the mediation of a representation, which is caused by the experience we are enjoying. This causal relation between experience and our introspective representation opens the possibility of misrepresenting the introspected mental state. The open possibility that we introspectively misrepresent phenomenal properties might provide an answer to the knowledge argument : since this misrepresentation would in fact allow us to restore a derivability relation between our physical knowledge and our phenomenal knowledge. If we misrepresent phenomenal properties as non physical, then it is still an open possibility that these properties are in fact physical, and thus amenable to a naturalist framework. What is the misrepresentation at play here ? There are two senses in which one might misrepresent the properties of one's phenomenal states. Here is one of Pereboom's examples : while at the dentist, I need an anesthetic injection. The dentist hides the needle from me, knowing how scared I am, and tells me he is just going to drop cold water in my mouth. I do experience something unpleasant we he actually injects me, but don't feel the usual sting, only experiencing something that I interpret as cold dripping water in my mouth. (Pereboom, 2014, p23) According to Pereboom, this example (along with others) should show that we may sometimes become aware of a discrepancy between the real nature of some of our phenomenal states, and what features we may introspectively grasp. Much like when we realize that we do not perceive the real nature of colors after seeing a given shade under different lightings In this example, we fail to introspect the painfulness of the anesthetic shot. Phenomenal properties are accessed under a perspective, just like colors in the real world are seen from a perspective, and may 3

Thalabard 2014 – Transparency & intrinsic properties

have different appearances under different lighting conditions. It seems to me that Pereboom's examples show at best that we may fail to access some of the properties of our phenomenal states, or that there may be a discrepency between the properties we would attribute to our experience on a third person basis, and the properties which we are actually aware of. This could be explained, along an overflow hypothesis à la Block (2007), by the limitation of conscious access, as opposed to the richness of phenomenology : by diverting my attention from the needle, the dentist just guides me away from noticing any painful quality my experience has. I strongly doubt that this would count as misrepresenting the nature of the phenomenal properties of my experience : after all, the possibility of phenomenal overflow is compatible with my representing correctly the properties of experience I am introspectively aware of. But even this is contentious : it might well be that I do not undergo any experience of pain in this case – an access theory of phenomenal consciousness would deny that I felt any pain when I received the shot. Under such an understanding of consciousness, limiting conscious experience to what I was actually aware of, the intuition of a discrepancy between what I did feel (and can introspectively report) and the actual features of my experience is a metacognitive illusion : I would misrepresent my own experience precisely in endowing it with phenomenal properties (or a phenomenal content) I could not report. Setting this possibility aside, we may grant that we sometimes misrepresent which phenomenal properties our experience has : for instance, I might mistakenly represent my perception as being a perception as of green, whereas my state is a perception as of orange. That is to say, my access to phenomenal properties is not incorrigible, and I may be wrong about which phenomenal properties are actually instantiated in a given conscious perceptual state. This possibility of misrepresentation is of no use when the question of the physical nature of phenomenal properties is at stake : for it does not concern the nature of the instantiated phenomenal properties. For QIH to counter the knowledge argument, introspection has to (mis)represent phenomenal properties as not having a physical nature. That we may be mistaken about which properties our experience has is indeed compatible with misrepresenting the nature of these properties, but it does not logically entail that we are mistaken about the nature of phenomenal properties. In fact, it is not evident that introspection does deliver any representation about the nature of phenomenal properties. Claiming that it does is a contentious claim, whatever our cognitive relation with phenomenal properties really is.

4

Thalabard 2014 – Transparency & intrinsic properties

Now, of course, the examples provided by Pereboom are not meant as knockdown arguments in favor of qualitative inaccuracy. They are hints, suggesting that introspection is not incorrigible – and indeed, should we introspectively construe qualia as non-physical properties of experience, we might well be mistaken given the fact that we may misrepresent the very properties experience has.

2) The transparency thesis and introspective opacity By means of contrast, I wish to examine a different view concerning introspection. My discussion will draw certain conclusions about the strength of introspective data concerning the metaphysics of conscious experience. The view I have in mind is the transparency thesis about introspection. Several formulations of transparency may be found in the litterature. Here are two typical statements : « Look at a tree and try to turn your attention to the intrinsic features of your visual experience. I predict you will find that the only features there to turn your attention to will be features of the presented tree, including relational features of the tree 'from here'. » (Harman, 1990, p478) « In general what makes a sensation of blue a mental one seems to escape from us ; it seems, if I may use a metaphor, to be transparent. We look through it and see nothing but the blue ; we may be convinced that there is something but what it is, no philosopher, I think, has clearly recognized. » (Moore, 1922, p20) The transparency thesis about conscious experience therefore holds the following : whenever we try to focus on our phenomenal states, we find nothing else to direct our attention to than the features of the represented objects of our environment. This seems to capture the innocent and oft remarked idea that perception puts us in contact with worldly features and objects (or at least seems to do so, in the case of illusion or hallucination). Transparency formulations may be found independently of any concern about qualia, and even be embraced by qualia realists (Kind, 5

Thalabard 2014 – Transparency & intrinsic properties

2003). Transparency usually plays a role in the defence of representationalism or disjunctivism. Although what I have to say here does not depend on the acceptance of either of these positions, I will briefly state the argument from transparency in favor of representationalism. This argument, as articulated by Tye, concludes, from the introspective invisibility of qualia, construed as intrinsic properties of experience, to the representational nature of phenomenal character. Since qualia are supposed to be introspectible features of experience, our incapacity to focus our attention on them seems to make representational features the only candidate properties to explain phenomenology. As it is used in the argument, the transparency thesis is a negative claim concerning the features we can introspect ; under the strong Tye-Dretske reading, it amounts to denying that we can become aware, directly, of any feature of experience. This straightforward argument in favor of representationalism has of course been criticized : the relationship between transparency and representationalism is highly intuitive, but logically loose. To pause. The straightforward link between representationalim and transparency seems to depend upon the acceptance of intuition 2 – precisely the intuition targeted by the QIH argument. Making transparency directly support representationalism requires that the nature of our phenomenal properties is revealed to us in introspection. Now, a QIH-like argument could certainly be devised against representationalism : perhaps the introspected representational nature of phenomenal character is as much a misrepresentation as taking phenomenal properties to be nonphysical. I do not wish to pursue this avenue here. Rather, the point I want to make concerns the scope of the transparency thesis. There are two main strategies for undermining the transparency claim, and thus, the representationalists' reliance on introspection. The first option, consists in plainly denying transparency. This would be Ned Block's (1996) view : « According to me, in normal perception, one can be aware of the mental paint – the sensory quality that does the representing. » (1996, p35). This option amounts to begging the question, no matter how much it concedes to the defender of transparency – by granting, for instance, that introspecting qualia is difficult, along a Moorean line of thought, or that we have only an oblique access to phenomenal properties (Loar, 1990). Simply denying transparency is not fruitful, and appeals to introspection as a warrant to ascertain metaphysical claims about the properties of experience is « no way to do philosophy » (Block, 1990, 6

Thalabard 2014 – Transparency & intrinsic properties

p73) – this methodological point would certainly be endorsed by Pereboom. A more interesting strategy would be the 'indirect' approach. It is possible to grant transparency, while showing that it is compatible with qualia realism. Qualia are troublesome properties. « Don't we find qualia staring at us in the face, when we attend introspectively to our experiences ? Well, not in a way that provides a quick knockdown refutation of the qualia quiners. Indeed, introspection can seem to provide confirmation rather than refutation of their view. The only thing that seems to correspond to 'attending introspectively to one's visual experience' is atending to how things appear to one visually, and offhand this seems to tell one what the representational contentnof one's experience is without telling anything about what the non intentional features of one's experience are that encode this content. » (Shoemaker, 1990, p100) Transparency appears to block any appeal to direct introspection in defence of qualia. Should such properties exist, we might not be directly aware of them, or might not be aware of them qua intrinsic properties of experience. In the light of these considerations on transparency, it seems safer to take transparency as an epistemic claim about our introspective grasp on the properties of experience : there is something it is like to undergo a phenomenal episode, and we can become aware of its character by reflecting on our current experience, but going further, and trying to directly introspect the phenomenal features responsible for this character will prove disappointing. This is certainly compatible with the rejection of intuition 2, according to which the nature of phenomenal properties is revealed through introspection. Where does this short discussion leave us ? The transparency thesis, taken as a claim about which properties we are aware of upon introspecting our phenomenal episodes, may be considered as uncommital as to the nature of phenomenal properties : introspection does not reveal anything about the real nature of the properties of experience. If this is the case, then, any belief concerning the nature of these properties, stemming from introspective data alone, is bound to be unwaranted – and hence possibly mistaken. Now, we do make reference to how our experience feels, and make reference to it through its phenomenal mode of presentation : but, as far as this way of picking our experience by its phenomenal character relies on introspection, it is silent on the metaphysical nature of the properties accounting for this mode of presentation, or phenomenal character. That is to say, singling out our experience by its phenomenal subjective feel does not commit one to any 7

Thalabard 2014 – Transparency & intrinsic properties

claim concerning the real nature - representational, physical, intrinsically mental or what have you - of its phenomenal properties. I want to suggest that phenomenal properties are introspectively opaque, as far as their nature is concerned : this opacity does not commit one to misrepresenting phenomenal properties – misrepresentation arises only if one wants to go beyond the deliverance of introspection. The purpose of this short discusson of the transparency thesis was to provide an example of an introspective thesis which denies that we are aware in any way of the features of experience as irreducible intrinsic properties. I am not quite sure why Pereboom wants us to adhere to the intuition that the nature of phenomenal properties is given in introspection. It is an open question whether our natural attitude takes introspection to reveal the nature of phenomenal properties. The upshot is that we might achieve the same result as QIH through the weaker transparency claim about introspection : given introspection, it is an open possibility that we (systematically) misrepresent the nature of phenomenal properties – all that the possibility of qualitative misrepresentation requires is that our concept of phenomenal property go beyond the deliverance of introspection. But it is a contentious point that it does.

Conclusion To sum up. I have discussed the Qualitative inaccuracy hypothesis, and what the misrepresentation at stake amounts to. If all it amounts to is our possibility to miss out on some phenomenal properties of our experience, or to mistake some properties for others, it seems that this can be accomodated without appealing to a misrepresentation of the nature of these properties. What should be counted as a decisive point in favor of QIH is that we indeed represent phenomenal properties of experience as non-physical properties : but should phenomenal properties indeed be non-physical, it might also well be the case that we misrepresent them, and it is a contentious point, as the transparency thesis should teach, that we represent phenomenal properties, let alone any property of experience, as having any kind of nature. Now, it certainly is an open possibility that we misrepresent phenomenal properties, but I am not certain, without certain further hypotheses, that this is enough to counter the knowledge argument. I do no want to be unfair to the QIH. Pereboom diagnoses the false beliefs based on introspection as 'errors of ignorance', following Hill's terminology : « The open possibility I am envisioning would have us make errors of ignorance in our 8

Thalabard 2014 – Transparency & intrinsic properties

introspection-based beliefs about phenomenal properties, since such beliefs wuld be based on appearances that fail to do justice to the real qualitative nature of those properties. » (2011, p23, n33) It seems to me that the same error of ignorance about phenomenal properties may be obtained through neutral introspective representations of phenomenal properties.

References Block Ned (2007) Consciousness, Access, and the Mesh between Psychology and Neuroscience (Behavioral and Brain Sciences, vol.30) Harman Gilbert (1990) The Intrinsic Quality of Experience (Philosophical Perspectives, vol., p452) Loar Brian (1990) Phenomenal States (Philosophical Perspectives, vol.4, p81-108) Martin Michael (2002) The Transparency of Experience (Mind and Language, vol.17, p376-425) Martin Michael (2004) The Limits of Self-Awareness (Philosophical Studies, vol.120, p37-89) Moore George Edward (1922) Philosophical Studies (K. Paul, Trench, Trubner & Co, London) Pereboom Derk (1994) Bats, Brain Scientists, and the Limitation of Introspection (Phil.& Phen. Research, vol.54(2), p315-329) Pereboom Derk (2011) Consciousness and the Prospects of Physicalism (Oxford University Press) Shoemaker Sydney (1994) Phenomenal Character (Noûs, vol.28(1), p21-38) Tye Michael (1995) Ten Problems of Consciousness (MIT Press)

9