Perceptual Aquaintance and Informational Content 89 Perceptual Aquaintance and Informational Content Donovan Wishon 1 Introduction Consider the case in which you are looking at a ripe red apple in standard viewing conditions.What is it that you are aware of? Intuitively, one is aware of the apple and its quality of redness. One is also in some sense aware of the quality of one’s experience of the ripe red apple; one is aware of “what it’s like” to perceive it visually in standard viewing conditions. What is the relation between these two things one is aware of, that is, between the qualitative character of one’s experience of seeing the apple and the redness of the apple itself? On the naïve realist view of perception, they are the same: the qualitative character of the experience of seeing red just is the redness that inheres in the surface of the apple. In the case of veridical perception, one is aware of the redness in the surface of the apple, and nothing else. On the indirect realist view of perception, on the other hand, what one sees directly is the qualitative character of one’s own mind-dependent sensation as of red, and one explicitly or implicitly infers the redness of the apple itself. Many currently working on a Russellian notion of perceptual acquaintance and its role in perceptual experience tend to treat naïve realism and indirect realism as an exhaustive disjunction of possible views.1 In this paper, I will articulate a third, intermediate view. This view of perceptual acquaintance is a form of direct realism, according to which one directly and literally sees the apple and its redness without seeing something mind-dependent and without making any intermediate inference. Nevertheless, it also maintains that the qualitative character of perceptual experience is a mind-dependent feature of our internal states of sentient awareness, and so is to be distinguished from the redness of the apple. Indeed, I believe that only this combination of direct realism and qualia internalism can provide an adequate characterization of our perceptual experience of things.2 1 2 I have in mind here Campbell 2002a, 2002b, and 2009 and Tye 2009. I will say more about how exactly we should understand the technical notion of ‘perceptual acquaintance’ in § 2 of this paper. Qualia internalism is the view that any duplicate of a subject will have the same qualitative experience regardless of the environment that he or she is in. Qualia externalism, on the Brought to you by | University of Mississippi Authenticated | 10.248.254.158 Download Date | 9/9/14 4:47 PM 90 Donovan Wishon To foreshadow things a bit, on my picture of perceptual experience we are aware of the qualitative character of our sensations simply in virtue of having them, and we are non-inferentially aware of the apple and its redness by being attuned to what the occurrence of our sensations tells us about the rest of the world. Consequently, we are presented with, and thus acquainted with, both the apple (and its redness) and the qualitative character of our sensation of the apple, albeit in very diferent ways. Indeed, perceptual experience has what I call a “two-faced presentational character”: we are presented all at once with external objects (and their perceptible features) and with the qualitative character of our perceptual experience itself. My view here particularly contrasts with recent work by John Campbell and Michael Tye on Russell’s notion of acquaintance and its role in perceptual experience, which has tended to emphasize the naïve realism of early Russell (and G. E. Moore).3 For example, Campbell has recently claimed: On a Relational View, the qualitative character of the experience is constituted by the qualitative character of the scene perceived…only this view, on which experience of an object is a simple relation holding between perceiver and object, can characterize the kind of acquaintance with objects that provides knowledge of reference (Campbell 2002a: p. 115). Similarly, Tye has argued that: Phenomenal character is manifest to us in our being aware of…external qualities. We cannot focus on it in any way that separates it from our focus on external things and qualities… On this view, the phenomenal character of the experience of red in a case of veridical perception is a feature of the surface the perceiver sees. The surface has the phenomenal character (Tye 2009: p. 120). On such views, when we have perceptual acquaintance with things, what we directly and literally perceive are external objects and their perceptible features, where the qualitative characters of our perceptual experiences inhere in, or are constituted by, those external objects and features. There are, however, a number of well-known objections to naïve realism. First of all, our current best scientiic theories of perception subscribe to the view that the right sort of neurobiological states of the brain are (at least) causally suicient for the occurrence of a perceptual experience as of external objects and features, even when there are no external objects 3 other hand, is the view that the qualitative characters of experiences are constituted by features of the external environment. This recent work includes Campbell 2002a, 2002b, and 2009 and Tye 2009. Brought to you by | University of Mississippi Authenticated | 10.248.254.158 Download Date | 9/9/14 4:47 PM Perceptual Aquaintance and Informational Content 91 present.4 And secondly, in cases of illusion or hallucination, subjects have the falsidical experience as of an external object having such-and-such features when either the object does not have those features, or there is no object whatsoever. Together, these objections suggest that the qualitative phenomenal features of which we are aware in perceptual experience are minddependent features of the experiential states themselves rather than mindindependent features of external objects. Campbell and Tye are unconvinced by these arguments, however. They urge that we should embrace some variety of disjunctivism, the view that there is no ‘common factor’ of phenomenal character or of content between veridical perceptions and the cases of illusion or hallucination.5 On such a disjunctivist view, in cases of veridical perception the perceived object or features are constituents of the perceptual experience itself and the qualitative characters of which we are aware inhere in them rather than in our internal states of sentient awareness. However, in cases of non-veridical perception or hallucination, the subject has an entirely distinct kind of perception-like experience, one with a diferent phenomenal character and diferent content, yet one that the subject cannot introspectively discern to be diferent from the veridical case, despite the fact that it is. Elsewhere, I have argued that the move to disjunctivism is unpromising.6 In particular, it is diicult to see how disjunctivism can provide an epistemologically and metaphysically plausible account of non-veridical perceptual experiences. But in any case, I intend here to take a very diferent route in developing an account of perceptual acquaintance. I’ll argue that once we recognize the “two-faced presentational character” of perceptual experience, we can safely avoid the prima facie dilemma between naïve realism and indirect realism.7 Indeed, once we fully recognize the informational character of perceptual experience, we can see how our perceptual acquaintance puts 4 5 6 7 I say “(at least) causally suicient” because the relation between the relevant neurobiological states of the brain and the conscious perceptual experience will be even more intimate on any physicalist ontology. In point of fact, Tye thinks that there is a common factor of phenomenal character and of content schema, but no common factor of singular content. However, as I argue elsewhere, it is hard to square this view with his professed qualia externalism. See chapter 3 of my dissertation Russellian Acquaintance and Phenomenal Concepts. See chapter 3 of my dissertation Russellian Acquaintance and Phenomenal Concepts. It is worth noting that Russell’s own theorizing about the nature of perceptual acquaintance was plagued by this false dilemma. One goal of this paper is to show that we needn’t follow the early Russell in embracing naïve realism and its problematic qualia externalism, nor the later Russell in combining qualia internalism with an undesirable indirect theory of perceptual experience. Brought to you by | University of Mississippi Authenticated | 10.248.254.158 Download Date | 9/9/14 4:47 PM 92 Donovan Wishon us in direct experiential contact with external objects, features, and states-ofafairs even though the qualitative characters of our experiences are minddependent features of our own states of sentient awareness. Thus, we can have the best of both worlds; we can be direct realists about the content of perceptual experience and internalists about its qualitative character. I will proceed as follows: In §2, I will explain how Russell conceived of acquaintance and its role in cognition, reference, and perception. In §3, I will present John Perry’s theory of information and informational content, which I think is a useful framework for thinking about naturalized content in general. In §4, I will explain how this theory of information and informational content helps us see how we can be direct realists about the content of perceptual acquaintance. Finally, in §5, I will draw on recent work on perceptual intentionality by Searle to develop my own account of the “two-faced presentational character” of perceptual experience, which helps explain how we can be internalists about its phenomenal character while being externalists about its content. 2 Knowledge of Things and Knowledge of Thruths Recent interest in a Russellian notion of perceptual acquaintance has been fueled by the increased appreciation of Russell’s more general distinction between two kinds of knowledge: knowledge of things and knowledge of truths. Russell’s distinction here roughly corresponds with an ordinary distinction drawn in many natural languages between two uses of the term ‘know’. First of all, we ordinarily talk about knowing that such-and-such is the case.This sort of knowledge paradigmatically involves conceptually articulated beliefs and is propositional in character; it involves cognitive attitudes that can be evaluated in terms of their truth or falsity. However, there is also another way in which we use the term ‘know’ in ordinary language: we talk of knowing individuals in the sense that we have directly encountered them and therefore have some familiarity with them. This is roughly what Russell has in mind with his technical notion of knowledge of things, though it is not entirely equivalent to the ordinary one. According to Russell, our knowledge of things is a distinctive epistemic relation that plays a fundamental role in reference and cognition. Indeed, Russell introduces the notion of knowledge of things, or acquaintance, in order to place a substantive cognitive constraint on what it takes for an individual to have genuinely singular thoughts about, or to make genuinely singular Brought to you by | University of Mississippi Authenticated | 10.248.254.158 Download Date | 9/9/14 4:47 PM Perceptual Aquaintance and Informational Content 93 reference to, individuals. In other words, our knowledge of things explains what it takes to have thoughts or utterances whose contents are constituted by the very individuals and features, if any, that the thoughts or utterances are about rather than by some way of getting at or identifying them.8 In fact, he argues that all cognition and linguistic designation ultimately rests on our fundamental epistemic capacity to be acquainted with, or consciously aware of, individuals and their features. Russell’s basic picture of cognition and reference is this: There is some special class of individuals (concrete and abstract alike) and their features with which we have experiential contact. On the basis of this experiential contact, we are in a position to direct conscious attention to those individuals and features. Our conscious attention to these experienced individuals and features in turn grounds our ability to make demonstrative, singular reference to them (to designate them with a “logically proper name”) and puts us in a position to acquire knowledge of such demonstrative, singular reference. Furthermore, our conscious attention to experienced individuals and features also puts us in a position to pick up information about, and thereby form some conception of, these individuals and features. Finally, on the basis of our conceptions of these individuals and features and our more general conception of how the world works, we can designate objects and features with which we lack experiential contact. We do so by employing representations that encode identifying conditions that an object or feature must uniquely satisfy in order to be the designated individual or feature of the representation, or by employing representations that are purely quantiicational in character. But what exactly is acquaintance? Basically, Russell conceives of acquaintance as a fundamental experiential relation between a conscious subject and individuals or features in the world, where the ‘knowledge of things’ that it constitutes is logically independent of, but serves as the ultimate enabling condition for, our knowledge of truths about, and conceptions of, the things or features with which we are acquainted. As Russell puts it: Knowledge of things, when it is of the kind we call knowledge by acquaintance, is essentially simpler than any knowledge of truths, and logically independent of knowledge of truths, though it would be rash to assume that human beings ever, in fact, have acquaintance with things without at the same time knowing some truth about them (Russell 1912/1997: p. 46). For Russell, acquaintance is relational in the straightforward sense that it is a relation that obtains between a conscious subject and the individuals or 8 I am being shamelessly terminologically anachronistic here, but I think doing so is helpful. Brought to you by | University of Mississippi Authenticated | 10.248.254.158 Download Date | 9/9/14 4:47 PM 94 Donovan Wishon features with which the subject is acquainted; both the subject and the individuals or features are constituents of the acquaintance relation.9 He holds that acquaintance is fundamental in that: [Acquaintance is] the simplest and most pervading aspect of experience… All cognitive relations—attention, sensation, memory, imagination, believing, disbelieving, etc.—presuppose acquaintance (Russell 1992: p. 5). Indeed, he holds that our acquaintance with things constitutes our most basic form of intentionality, or object-directedness, and grounds our ability to make genuine singular reference to the individuals and features with which we are acquainted. He contends that: The faculty of being acquainted with things other than itself is the main characteristic of a mind. Acquaintance with objects essentially consists in a relation between the mind and something other than the mind; it is this that constitutes the mind’s power of knowing things (Russell 1912/1997: p. 42). Russell is quite clear in holding that acquaintance is essentially an experiential relation in which the conscious subject is presented with the objects of his or her awareness. In his 1913 Theory of Knowledge, he remarks: Now, since we have decided that experience is constituted by a relation, … we shall employ synonymously the two words ‘acquaintance’ and ‘awareness’, generally the former.Thus when A experiences an object O, we shall say that A is acquainted with O (Russell 1992: p. 35). Elsewhere, Russell asserts that: The distinction between acquaintance and knowledge about is the distinction between the things we have presentations of, and the things we only reach by denoting phrases (Russell 1905/2000: p. 212). For Russell, an individual or feature is presented to a subject just in case it makes the right sort of direct diference to that subject’s conscious and cognitive life such that the subject is aware of it. Finally, Russell holds that the knowledge of things of which acquaintance consists is unrelective, non-conceptual, and non-propositional. Knowledge of things is ‘non-conceptual’ in the sense that it does not require a subject’s having antecedent possession of any concept or conception of the thing or feature with which he or she is acquainted. Indeed, acquaintance is precisely the sort of unrelective and ‘objectual’ conscious awareness of which many non-linguistic sentient animals and small children are capable. 9 More precisely, both the subject and the individuals or features are constituents of the relational fact. Brought to you by | University of Mississippi Authenticated | 10.248.254.158 Download Date | 9/9/14 4:47 PM Perceptual Aquaintance and Informational Content 95 Knowledge of things is ‘non-propositional’ in the straightforward sense that it lacks genuine veridicality conditions. Indeed, on Russell’s view, either a subject experiences such-and-such individuals or features, or the subject does not. It is in this sense (alone) that we know a thing “perfectly and completely” when we are acquainted with it.10 Of course, it goes without saying that the epistemic mechanisms and faculties through which we become acquainted with things are themselves fallible in that they can misrepresent the individuals and features we are experiencing. But acquaintance itself does not admit of error; it is all or nothing.11 3 Information and Intentionality In the previous section, I provided a basic sketch of Russell’s notion of acquaintance and its role in reference and cognition. In this section, I want to provide a framework that I believe is crucial for thinking about the nature of perceptual acquaintance. And while I ultimately aim to defend an internalist account of the qualitative characters of our perceptual experiences, I will here provide an account of the nature of perception that is, in its broadest features, compatible with both naïve realism and its competitors. Indeed, I think that any adequate account of perception must begin with a more primitive theory about information, informational content, and intentionality. In developing my theory of perception, I will rely on the theory of information and informational content advanced by John Perry in his “Information, Action, and Persons.”12 To begin with, the basic unit of information on Perry’s account is a signal.A signal is an object having a property, where both the object and the property may be quite complex. Perry calls this object the carrier of information and its relevant property the indicating property.13 Given this technical notion of a signal, Perry characterizes information as: What one part or aspect of the universe (the signal) shows about some other part or aspect [of the universe] (the subject matter) (Perry 2002: p. 174). 10 11 12 13 Russell 1912/1997: pp. 46–7. Note that Russell’s view here should not be understood as an endorsement of the so-called “thesis of revelation”. Perry 2002. Perry’s earlier work with David Israel and Syun Tutiya is also excellent for thinking about the nature of information, but I will focus only on his later work on the topic. Perry 2002: p. 174. Brought to you by | University of Mississippi Authenticated | 10.248.254.158 Download Date | 9/9/14 4:47 PM 96 Donovan Wishon However, as Perry rightly notes, the mere occurrence of some state or event in the world says nothing about how things are with the rest of reality unless its occurrence is against the background of the way the world works, or at least some conception thereof. Indeed, it is a common fault of informational-semantic theories that they fail to emphasize that signals only carry information about other parts or features of the universe given a background of constraints, where constraints are understood as states-of-afairs, some of which are true. Perry nicely puts this point as follows: [That a signal carries information about some other part or feature of the universe] is possible only because events are constrained by laws of nature, or as I prefer, because of its more liberal, common-sense, loose, and nonreductive connotations, by the way that things happen. The information carried by a signal is what else things have to be like, for the signal to have occurred, given the way things happen (Perry 2002: p. 175). Perry’s idea is that a signal S carries the information that P if there are principles of how things actually happen such that given those principles, the signal would not have occurred unless P were the case. One crucial thing to notice about his notion of information here is that it is factive. In other words, given the way the actual world works, it is not possible for a signal to carry the information that P and for it not to be the case that P. This is one reason that information is not a viable candidate to which we can reduce intentionality, for intentional states of systems can be directed at, can have as conditions of satisfaction, states-of-afairs that are not the case. Another reason that we cannot reduce intentionality to Perry’s notion of information is that any occurrence in the world can constitute a signal that carries information about any other state of the world relative to the right constraints about how the world actually works. But one of the essential features of genuine original intentionality is that it is directed at, or places conditions on, particular objects or states-ofafairs that, again, might or might not in fact obtain. Moreover, it is generally directed in some particular way at them. Thus, it is not enough for the signal simply to causally-covary with some particular indicated object or state-of-afairs. Perry himself, of course, is quite aware of the fact that we cannot plausibly identify intentionality with, or reduce it to, mere information. Instead, he maintains that naturalized intentionality is to be identiied with the right sort of informational content. To see how informational content difers from mere information, it is helpful to have the following rough general schema: Brought to you by | University of Mississippi Authenticated | 10.248.254.158 Download Date | 9/9/14 4:47 PM Perceptual Aquaintance and Informational Content 97 A signal S has the informational content that P if and only if, given constraints C, S occurs if and only if P. In the case of mere information, the constraints against which a signal acquires its informational content are true constraints governing the way the actual world works. However, what is crucial for Perry’s notion of informational content is that the constraints we rely on in determining the content of a particular signal need not be true, a fact which allows us to consider counterfactual circumstances and which, as we’ll see, explains many cases of informational error. Furthermore, Perry maintains that for any signal there are numerous layers of informational content they have depending on which constraints and circumstances are assumed in the background. Some of this information content is what Perry calls relexive information, or information about the signal itself.14 Other levels of the signal’s informational content are about states or features of the rest of the world; Perry calls information of this sort incremental information.15 Perry thinks we get from the relexive information about the signal itself to the subject-matter, incremental information it has by adding to the background constraints particular details about the circumstances of the signal’s occurrence.16 In other words, a signal will give us information about what some other part or aspect of the world must be like given the signal’s occurrence, the circumstances in which it occurred, and the way the world works. Borrowing one example from Perry, an x-ray has the informational content that so-and-so has a cavity in such-and-such tooth given the way that x-rays and tooth decay work and given the circumstances that the x-ray was exposed to so-and-so’s tooth.17 Perry thinks that many systems and devices depend on information and informational content having these features. Indeed, he thinks that many systems and devices are constructed in such a way that they harness the informational content of a signal in order to satisfy some goal. They do this by having some sort of architectural design such that the occurrence of a particular signal is also the cause an action that will be successful in just those circumstances. To borrow one of Perry’s examples, the springing of a mousetrap carries the information that there is a mouse in the trap relative 14 15 16 17 Perry 2002: p. 175. Perry 2002: p. 176. Perry 2002: p. 179. Perry 2002: pp. 175–6. Brought to you by | University of Mississippi Authenticated | 10.248.254.158 Download Date | 9/9/14 4:47 PM 98 Donovan Wishon to the constraints that only mice will spring it, and it is designed to succeed in trapping mice in precisely those conditions in which a mouse is in it.18 If, however, the constraints are not satisied, then the occurrence of the signal will fail in bringing about the designed efect. Of course, in the case of the mousetrap, its informational content is ‘derived’ rather than ‘original’. That is, the occurrence of the trap’s signal has the informational content that there is a mouse in the trap given the relevant constraints and circumstances only because we have given it the job or goal of trapping mice. For this reason, Perry acknowledges that it is not enough for a signal simply to have informational content for it to have genuine intentionality. But what more is needed on Perry’s account to get to genuine original intentionality? His answer is that what we need is a system whose indicating states have natural jobs or goals to indicate some state-of-afairs in the world and to guide actions that make sense given their occurrence. However, unlike so-called ‘teleosemantic’ approaches to naturalized intentional content, Perry does not limit the sources of natural functions of states to those supplied by natural selection, learning, and social accretion. Indeed, one striking feature of Perry’s account is that the qualitative characters of many of our states of sentient awareness are a vital source of naturally supplied goals. In particular, he maintains that the pleasant or unpleasant dimensions of the qualitative characters of many of our experiential states provide us with a crucial source for deriving natural goals, a fact that natural selection and social accretion make good use of in generating more sophisticated natural goals for us.19 Hence, Perry’s own view bears a strong resemblance to those according to which we cannot explain intentionality naturalistically, but only by appealing to consciousness itself. However, where he parts company with such opponents of naturalized intentionality is in holding that consciousness itself, and its qualitative characters, can in principle be naturalized. Of course, it is absolutely crucial to point out that Perry holds that the physical domain must be such that consciousness, in all of its subjective and qualitative richness, can arise within it. So he is not the kind of traditional naturalist who would seek to make consciousness out to be less than it in fact is simply with the goal of making it it easier into an overly impoverished conception of physical reality. 18 19 Perry 2002: pp. 178–9. Of course, this can’t be the full story since many experiences, such as our experiences of color, do not have a recognizable dimension of pleasantness or unpleasantness. Brought to you by | University of Mississippi Authenticated | 10.248.254.158 Download Date | 9/9/14 4:47 PM Perceptual Aquaintance and Informational Content 99 4 Informational Content and Perceptual Content In the previous section, I presented Perry’s account of informational content as what one aspect of the world tells us about another aspect of the world given relevant constraints and circumstances. But what exactly does this have to do with perception? I think that what is central to perception is not its causal character, as many believe, but rather its informational character. It is that the states of our perceptual systems have informational content about states of the world outside of our perceptual system in a way that we can harness this information to guide our thought and action. In fact, our perceptual systems have been designed through processes of natural selection, social accretion, and learning to provide us with information about our environment in order to help us be more successful in navigating it and thereby satisfying our various practical goals. Perception accomplishes this, of course, because we and our perceptual systems are, as Perry puts it, ‘attuned’ to more-or-less accurate regularities that hold between distal objects in our environment and the internal states of our perceptual systems. The crucial notion of attunement here is of an unrelective sensitivity to, or even diferential responsiveness to, some moreor-less accurate constraints on the way that the world works. This sensitivity to constraints allows us to track and harness the information carried by states of our perceptual system about particular events in our environment. We might even construe attunement as a set of capacities, abilities, or knowhow of the subject, or of his or her perceptual faculties.20 For example, many birds are attuned to the constraint that the path to any clearly visible object is unobstructed; indeed, their attunement to this constraint guides their (often highly complex) behavior even though they lack an explicit and relective appreciation of the constraint and even though the constraint is false (given the advent of transparent windows).21 One beneit of thinking of perception in terms of informational content is that it helps explain how direct realism might be true. For, as Searle remarks in a similar context, it is important to note that an account of perception does not get to declare itself a version of direct realism for free. Rather, direct realism should be a consequence of an independently motivated account of the intentionality of perception. And indeed, one crucial beneit of think20 I think that Perry’s notion of attunement to constraints bears important similarities to Searle’s notion of the ‘background’ against which our intentional states have their particular conditions of satisfaction. See Searle 1983. 21 I borrow this example from Perry 2002: p. 184. Brought to you by | University of Mississippi Authenticated | 10.248.254.158 Download Date | 9/9/14 4:47 PM 100 Donovan Wishon ing of perception in terms of informational content is that we evidently can explain how direct realism about perception can be true. This is because the informational content of a signal is literally about the very external objects, feature-instances, or states-of-afairs, if any, for which, in the particular circumstances of its occurrence and relative to appropriate constraints, it has the job of indicating. And in the case of perceptual experience, the constraints governing our perceptual capacities are such that our states of perceptual awareness are directly about perceptible external objects, feature-instances, or states-of-afairs in our environment. What’s more, this fact about the nature of informational content explains how it is that perception can have several other important intentional features noted by Searle and others.22 First, it explains why perception has an essentially indexical element; indeed, our perceptions are always about particular objects, feature-instances, or states-of-afairs that bear some indexical relation to us and our perceptual faculties. Secondly, it explains why perception is, as Searle puts it, ‘causally self-referential’. The basic idea here is that in cases of veridical perception, we are presented with the very objects, featureinstances, or states-of-afairs that cause our perceptual experience as of them. And indeed, it is very natural to see the causal self-referential character of all perception as a relexive informational content that is a constitutive feature of perceptual signals. But we can also maintain the intuitive picture that the subject-matter or incremental informational content of perceptual experiences is exhausted by the objects, feature-instances, or states-of-afairs, if any, that constitute their conditions of satisfaction. In other words, we can straightforwardly maintain that a perceptual experience is veridical if and only if the particular things experienced are the way the perceptual experience represents them as being. Most importantly, thinking of perception in terms of its informational content also helps explain how it is that we can have non-veridical perceptual experiences or hallucinations, ones that have a qualitative ‘common factor’ with (or subjectively indistinguishable diference from) veridical cases. This is because a perceptual signal has its particular informational content only relative to constraints and circumstances to which we are attuned, which suggests a very natural explanation of why, in cases of non-veridical or perceptual experience, the states of our perceptual systems do not carry the information we take them to. Indeed, the basic idea here is that in such non-veridical or hallucinatory cases, we (or our perceptual systems) are attuned to false con22 Searle, in draft: pp. 19–25. Brought to you by | University of Mississippi Authenticated | 10.248.254.158 Download Date | 9/9/14 4:47 PM Perceptual Aquaintance and Informational Content 101 straints on how the world works and we ind ourselves in circumstances that bring out this fact. In cases of perceptual error and illusion, the conditions under which we perceive some object, feature-instance, or state-of-afairs are such that we have a perceptual experience whose qualitative character and informational content do not accord with the actual information that the occurrence of the perceptual state carries. Hence, we take the perceived object, feature-instance, or state-of-afairs to be other than it in fact is. Moreover, in cases of hallucination, we ind ourselves in such abnormal circumstances that we have the relevant perceptual experience without there being any sort of corresponding object, feature-instance, or state-of-afairs required for the satisfaction of its veridicality conditions.23 In fact, once we recognize that the singular content of perceptual experience is incremental informational content, we have a natural explanation for why our perceptual experiences are error-prone in the irst place. For like many information-harnessing devices, we needn’t be attuned to particularly accurate constraints and circumstances for our perceptual faculties to perform well enough at helping us pick up information about the environment suitable for guiding successful actions. All that is required is that our perceptual faculties work well enough in the circumstances in which we most often ind ourselves, and there is suicient reason to suppose that they do. In any event, my account of the informational content of perception provides an intuitive explanation about why there at least seems to be a common factor between the veridical and non veridical cases of perceptual experience. The reason is that many of the possible errors can occur even if we have exactly the same signal. For, as I’ve just noted, the veridicality of a particular informational signal depends on factors external to it, such as background constraints and circumstances. Hence, if the occurrence of the signal itself suices for the instantiation of the qualitative character of our perceptual experience, as I believe is the case, then we have a straightforward explanation for why there is a qualitative ‘common factor’. Indeed, the common factor in such cases just is the occurrence of one and the same informational content-bearing signal. 23 Obviously, this story is extremely over-simpliied. For instance, we are also attuned to the fact that our perceptual faculties are fallible, which explains why we do not take visual blackouts, double-vision, blurry vision, etc., to be qualitative features of the world. Indeed, we generally take such perceptual experiences to provide information about the improper functioning of our own perceptual faculties rather than information or misinformation about the world. Brought to you by | University of Mississippi Authenticated | 10.248.254.158 Download Date | 9/9/14 4:47 PM 102 Donovan Wishon 5 Two Kinds of Awareness Now that we have at least a sketch of a framework of the informational character of perception, I want to consider how we might be internalists about the qualitative character of perceptual experience without being indirect realists about perception. My central claim in this section will be that given a proper informational framework for thinking about perception, we can accomplish this, provided that we are careful to distinguish between intentional and phenomenal awareness. I will also return to the question of how I see the resulting theory of perception as itting into a broader Russell-inspired account of acquaintance. What I hope to show, at least in outline, is that we can indeed have an adequate model of perceptual acquaintance without adopting naïve realism or disjunctivism. In fact, I think we took the irst steps towards developing an adequate model of perceptual acquaintance in the previous section when we made the distinction between signals and their informational contents. Indeed, what we saw was that we can usefully think of perceptual experiences as signals whose incremental informational content is directly about the external objects, if any, for which they have the job of standing, relative to appropriate constraints and circumstances. Thus, we can respect the naïve and direct realist insight that our perceptions are directly of, or about, external objects, feature-instances, and states-of-afairs. On the other hand, we can still suppose that the qualitative characters of our perceptual experiences are identical with, or supervene on, intrinsic features of the signal itself. In fact, I think that one of the crucial insights of Perry’s theory of information and informational content is that it explains why information is useful in the irst place. For recall that, fundamentally, it is what one state or aspect of the world says about another state or aspect of the world, relative to constraints and circumstances, in virtue of the properties it (i.e. the signal) has. Indeed, the structure of information and informational content is important because it allows us to get at the properties of an accessible signal in order to acquire information about distal state-of-afairs to which we do not have as ready access. Of course, we have to be extremely careful here. Unlike in many cases in which information is useful, we don’t literally look at, or perceive, the signals in the case of our own perceptual experiences. Indeed, supposing that we do so is exactly the mistake that the indirect realist about perception makes. Instead, we simply have the perceptual experiences, and we are, in a relevant sense, aware of (or at least sensitive to) the qualitative features they possess Brought to you by | University of Mississippi Authenticated | 10.248.254.158 Download Date | 9/9/14 4:47 PM Perceptual Aquaintance and Informational Content 103 just by having them.24 Moreover, it is in virtue of having, and thereby being aware of, such perceptual experiences that, in the right conditions, we directly and literally perceive external objects, feature-instances, or states-of-afairs, without any intermediate inference. But how can we be aware of the qualitative features of our experiential states just by having them? Unfortunately, I cannot hope to answer such a deep question here. What I will say is that whatever metaphysical account of consciousness we ofer, it must ultimately account for this somewhat remarkable fact. Indeed, I take it as a datum to be explained that we have experiential states where the having is, in the relevant non-relective and non-propositional sense, the knowing.25 This brings us to John Searle’s crucial distinction between the two senses of ‘of ’ that he thinks are relevant to our perceptual awareness. According to Searle, there are two fundamentally diferent kinds of awareness relation, both of which are marked in English by the preposition ‘of ’.The irst is what he calls ‘the ‘of ’ of constitution’, and the second is what he calls ‘the ‘of ’ of intentionality’.26 To see what the distinction between these two relations are, Searle asks us to consider the following two cases: (1) When I feel my pain, I am aware of, or conscious of, the pain. (2) When I see something red, I am aware of, or conscious of, the instance of red. What Searle hopes we’ll recognize is that the awareness relations involved in (1) and (2) are quite diferent. Intuitively, in the irst case the awareness of the pain just is the pain, but in the second case, the awareness of the red is not itself the instance of red.27 Rather, in the case of (2), the instance of red is the 24 I want to note that the locution “aware of ” is somewhat problematic for characterizing the phenomenal awareness we enjoy simply in virtue of having a perceptual experience. Indeed, the locution “aware of ” suggests intentional awareness, which is to be distinguished from phenomenal awareness. Unfortunately, I lack a better locution. 25 Galen Strawson makes roughly the same point on page 286 of Strawson 2009. 26 Searle, in draft: 14. It is important to note that we do not have follow Searle’s somewhat dubious semantic thesis about the English preposition ‘of ’ to agree with him that there is a distinction between the two sorts of awareness involved in perceptual experience. 27 I think that there are perhaps two diferent ways of interpreting Searle’s important notion of ‘the ‘of ’ of constitution’. First, we can follow him in holding that the awareness of pain, for example, just is the pain. But second, we might instead hold that the awareness of pain is numerically identical with our having the pain. In my opinion, it is quite natural to say that when we have a perceptual experience, and are thereby aware of it, this just is the state of sentience, and so I prefer the second interpretation of Searle’s notion. However, I think Brought to you by | University of Mississippi Authenticated | 10.248.254.158 Download Date | 9/9/14 4:47 PM 104 Donovan Wishon intentional object of our (in this case) veridical awareness as of something red; it is what our perceptual experience as of red is directed at, or about.28 Searle quickly characterizes his distinction as follows: In the case of pain, the pain is identical with the awareness or the consciousness. There aren’t two things, the pain and the awareness of the pain.That is what I call the “of ” of constitution. In this case the awareness is constituted by the thing that I am aware of. But that “of ” is diferent from the “of ” of intentionality. When I see a red object, I am indeed aware of red, but the awareness in this case is of the red in the red object. The red is not identical with the awareness, the “of ” is the “of ” of intentionality (Searle, in draft: p. 13). Indeed, Searle thinks that it is absolutely crucial that we do not confuse the ‘of ’ of constitution with the intentional relation we bear to objects, featureinstances, or states-of-afairs in perception. If one does so, he argues, then it is far too easy to form the mistaken belief behind indirect realism that: Intentionality consists invariably of some sort of representation, and the subject who has the intentional state has some sort of [conscious] relation to these representations [rather than to their intentional objects] (Searle, in draft: p. 14.) What’s more, he continues: That is what forces the analogy between the intentional theory of perception and [the] idea that intentionality would be like reading a newspaper about the real world. I think, frankly, this is quite an absurd conception of intentionality of perception… [And] if you think that all intentionality is a matter of [conscious] relation to a representation, that the object of the intentionality is the representation or some element of it, and that on an intentionalistic account the awareness in the awareness of a hallucination must be the same kind of awareness as the awareness of an object in a veridical perception, then it will seem to you that an intentionalistic account of perception involves a denial of naive realism (Searle, in draft: p. 15).29 In fact, Searle thinks that it is roughly this line of reasoning that forces one into the false dilemma of choosing naïve realism or adopting some version of indirect realism. However, Searle thinks that we shouldn’t be moved by this either way of construing the ‘of ’ of constitution is compatible with our English usage of ‘aware of ’ and little turns on the issue. 28 I take it that Searle holds that in this second case, what we are aware of, in the sense of the ‘of ’ of constitution, is the qualitative character of the experience as of red. 29 Note: Searle himself uses the term ‘naïve realism’ in the same fashion as I use ‘direct realism’. Brought to you by | University of Mississippi Authenticated | 10.248.254.158 Download Date | 9/9/14 4:47 PM Perceptual Aquaintance and Informational Content 105 argument, and that we won’t be, provided we distinguish between the ‘of ’ of constitution and the ‘of ’ of intentionality. One lesson we should take from Searle’s distinction is that there is a fundamental diference between what we might call ‘phenomenal awareness’ and ‘intentional awareness’. In the case of phenomenal awareness, we are nonrelectively and non-propositionally aware of the qualitative characters of our experiential states simply in virtue of having them. In the case of intentional awareness, on the other hand, we are aware of some state of the world in virtue of being in some representational state that is directed at it. Like our phenomenal awareness, such intentional awareness of things needn’t be relective or propositional. Indeed, the subject need only be attuned to the informational content (and to relevant background constraints and circumstances) indicated by the representational state. However, unlike in the case of phenomenal awareness, intentional awareness can be relective and propositional for sophisticated concept-wielding subjects. In fact, sophisticated enough conceptwielding subjects can even direct their relective intentional awareness to think about their own states of phenomenal awareness using higher-order so-called ‘phenomenal concepts’ of them. The crucial point, however, is that this intentional awareness of our conscious experiences is a further cognitive achievement from our more basic phenomenal awareness of them. Returning to the central theme of this paper, what I think these considerations show is that there are various kinds of acquaintance involved in our conscious perception of things. On one hand, in having a conscious perception, we are acquainted with the perceptual experience itself. Our states of sentient awareness are in this sense ‘self-presenting’; we are phenomenally aware of them when we have them in precisely the same way that non-linguistic sentient animals are aware of their own conscious experiences when they have them. And on the other, in having a conscious perception, we are acquainted with the external intentional objects, if any, of the perceptual experience. We are thereby presented with the objects, feature-instances, or states-of-afairs, if any, of which it is a perception. This fact follows from the very informational character of perception. Putting these two aspects together, we can say that conscious perception has what we might call a “two-faced presentational character”; we are presented in conscious perception both with the object of our perceptual experience, if any, and with the subjective, qualitative character of the experience itself. In saying this, we must be careful to note that for Russell, acquaintance does not carry with it any commitment to dubious epistemic theses such as the so-called ‘thesis of revelation’ or overly strong views about the ‘intimating’ Brought to you by | University of Mississippi Authenticated | 10.248.254.158 Download Date | 9/9/14 4:47 PM 106 Donovan Wishon character of the objects of our acquaintance.30 Though Russell holds that acquaintance presents us with its objects, he does not claim that we thereby have an exhaustive presentation of the essential nature of that object, as the thesis of revelation would maintain. Nor do we automatically acquire any knowledge of truths about a thing simply by being presented with it in experience. Indeed, I think we should take Russell’s contention that knowledge of things is logically independent of any knowledge of truths absolutely seriously. This goes for the alleged ‘self-intimating’ character of our states of sentient awareness. We simply do not know that we are having such-and-such experiences simply in virtue of having them, at least if what we have in mind is relective, semantically-articulated knowledge-that. Again, this is not to say that we sophisticated, concept-wielding adults don’t often think about them using phenomenal concepts, because we certainly do. The point is that this is a cognitive achievement above and beyond what acquaintance with our experiences provides all by itself.31 6 Conclusion In this paper, I have argued that once we recognize what I’ve called the “twofaced presentational character” of perceptual experience, we can be direct realists about the content of perceptual experience and internalists about its qualitative character. We can be direct realists because the informational character of perceptual acquaintance puts us in direct experiential contact with external objects, features, and states-of-afairs. And we can be qualia internalists because of the special self-presenting character of the perceptual signals that have such informational content about the world. What I hope I have shown, at least in rough outline, is that we can have an adequate account of Russellian perceptual acquaintance without embracing problematic theories of perception such as naïve realism, disjunctivism, or indirect realism. 30 I grant, however, that there are some unfortunate passages that suggest otherwise. See Wishon, In draft A. 31 Though it is beyond the scope of this paper, I want to at least note that one consequence of this view is that our phenomenal judgments are prone to error. Contrary to the opinions of many, I think this is a welcome consequence. For more on the relation between acquaintance and phenomenal concepts, see chapter 5 of my dissertation. Brought to you by | University of Mississippi Authenticated | 10.248.254.158 Download Date | 9/9/14 4:47 PM Perceptual Aquaintance and Informational Content 107 Acknowledgements I would like to thank John Perry, Krista Lawlor, Mark Crimmins, Daginn Føllesdal, David Beisecker, John Campbell, and John Searle for comments and suggestions regarding earlier drafts of this paper. I would also like to thank my wife, Christy, for her unwavering support. References Campbell, J. 2002a. Reference and Consciousness. Oxford: Oxford University Press. —. 2002b. Berkeley’s Puzzle. In Conceivability and Possibility, edited by T. Gendler and J. Hawthorne, pp. 127–43. Oxford: Oxford University Press. Campbell, J. 2009. Consciousness and Reference. In Oxford Handbook to Philosophy of Mind, edited by B. McLaughlin, A. Beckermann, and S.Walter, pp. 648–62. Oxford: Oxford University Press. Perry, J. 2002. Information, Actions, and Persons. In Identity, Personal Identity, and the Self, pp. 167–88. Indianapolis: Hackett. Russell, B. 1905/2000. On Denoting. In The Philosophy of Language, Fourth Edition, edited by A. Martinich, pp. 212–20. New York: Oxford University Press. —. 1912/1997. The Problems of Philosophy. New York: Oxford University Press. —. 1992. Theory of Knowledge:The 1913 Manuscript. New York: Routledge. Searle , J. 1983. Intentionality. New York: Cambridge University Press. —. In draft. Disjunctivism and the Intentionality of Perception. Strawson, G. 2009. Selves: An Essay in Revisionary Metaphysics. New York: Oxford University Press. Strawson, P. 1959. Individuals. New York: Routledge. Tye, M. 2009. Consciousness Revisited: Materialism without Phenomenal Concepts. Cambridge, MA: MIT Press. Wishon, D. In draft A. Russellian Acquaintance without Discriminating Knowledge. —. In draft B. Russellian Acquaintance and Phenomenal Concepts. Doctoral Dissertation. Brought to you by | University of Mississippi Authenticated | 10.248.254.158 Download Date | 9/9/14 4:47 PM 108 Donovan Wishon Brought to you by | University of Mississippi Authenticated | 10.248.254.158 Download Date | 9/9/14 4:47 PM