perception, philosophical perspectives subjectivity simply is what ‘intrinsicness’ amounts to. Chalmers, D. (1996). The Conscious Mind. Another consideration is that, given that the brain is CliVord, W. (1874/1886). ‘Body and mind’. In Fortnightly Review, 1886 the physical ground of consciousness, we might regard December. (Page references are to the reprint.) 1890 1950 The Principles of Psychology 1 experience as a kind of window into the intrinsic nature James, W. ( / ). , Vol. . Levine, J. (2001). Purple Haze. of at least this one, very complex, physical entity which 1991 Mind, Brain and the Quantum: The W Lockwood, M. ( ). we thus nd to be mentalistic in nature (see Lockwood Compound ‘I’. 1991 ). It is then perhaps a natural inference to the McLaughlin, B.(1992). ‘The rise and fall of British emergentism’. ubiquity of the mental, although the sort of experience In Beckermann, A. et al. (eds) Emergence or Reduction. that characterizes the physically fundamental entities in Nagel, T. (1979). ‘Panpsychism’. In Mortal Questions. the world is presumably altogether diVerent and radic- Rosenberg, G. (2004). A Place for Consciousness: Probing the Deep ally simpler than our own complex mental lives, reXect- Structure of the Natural World. 1999 Theories of Consciousness ing the diVerences in complexity between brains and Seager, W. ( ). . 2005 Panpsychism in the West electrons (see Rosenberg 2004, Strawson 2006). Skrbina, D. ( ). . Strawson, G. (2006). ‘Realistic monism: why physicalism entails Against these points stands the Xat intuitive implaus- panpsychism’. Journal of Consciousness Studies, 13. ibility of the hypothesis that everything has a mental Whitehead, A. N. (1929) Process and Reality: An Essay in Cos- dimension. But in addition there are some counter- mology. arguments. One is that the physical nature of reality seems to be entirely suYcient to account for everything that happens (this is often called the ‘causal closure’ perception, philosophical perspectives Percep- of the physical world). If the mental is not itself a tion is a way of acquiring information, beliefs, or physical feature at bottom (via some sort of reductive knowledge about the world by means of the senses. relation or some other acceptable dependence upon the In philosophy, ‘perceive’ and its derivatives ‘see’, ‘hear’, physical) then it threatens to become *epiphenomenal. and the like, are usually taken to be success verbs. A panpsychist might reply that at the fundamental level, Thus, when Macbeth claimed to see a dagger before him the mental features are essential to the causal powers of when there was no such dagger, he was mistaken; things (see Rosenberg 2004). he merely seemed to see a dagger. In fact, he *hallucinated Another objection holds that panpsychism has its a dagger. A major goal of philosophical theories of per- own emergence problem (see James 1890/1950:Ch. 6, ception is to provide an account of perception that differ- Seager 1999:Ch. 9). Somehow there is a transition from entiates it from hallucination and from other mental the ‘elemental’ mental features of the physical constitu- occurrences. Section 2 below looks at approaches to the ents of things to the complex minds possessed by com- latter, while sections 3 and 4 outline approaches to posite entities such as ourselves. If the panpsychist is the former. Another goal is to address the question willing to admit this kind of emergence why not simply of how perception can yield and justify belief, thus opt for an emergentist solution to the whole problem of making it a source of knowledge. In answering this second mind, and avoid the basic implausibility of panpsy- question philosophy of perception overlaps with episte- chism? In reply, it might be noted that emergence mology. of complexes of a given set of features is generally Some terminological preliminaries need to be much easier to understand than the apparently radical noted. First, when we consciously perceive the world emergent transformation of matter into consciousness. we have ‘perceptual experiences’. It is usually taken to Finally, there is a methodological objection to panpsy- be the case that we can also have perceptual experiences chism. One of our chief metaphysical goals is to under- when we are hallucinating (and hence not perceiving). stand everything in terms of the best accounts of the Whether the perceptual experiences involved in percep- world which we currently possess. The physical sciences tion and hallucination are the very same kinds of states provide these accounts and they do not avail themselves is an important question in recent philosophy of percep- of the panpsychist option. One might argue that it is thus tion and supporters of common-kind and disjunctive incumbent upon metaphysicians to exert every eVort to theories of perception, discussed in sections 3 and 4 understand mind from within the structure provided by below, endorse very diVerent answers. the physical sciences. There is a kind of cogency to such Secondly, most philosophers hold that perceptual an objection, if it is taken to encourage eVorts at a experiences are by deWnition conscious states. A great naturalistic metaphysics. But such a metaphysics might deal of philosophy of perception is concerned with the fail, and it is good to explore options which might come nature of these states. The sense-datum theory holds that in handy, even if they stretch imagination beyond its such states involve perceiving non-physical mind-depen- usual bounds. dent objects. Adverbialism conceives of such states WILLIAM SEAGER as states in which one is sensing in a certainly way. 502 perception, philosophical perspectives

*Representationalism argues that such states represent phenomenal character. Moreover, as stated above, per- the world to be a certain way. A great deal of modern ceptual experiences have their phenomenal character philosophy of perception has been concerned with essentially, but beliefs, qua beliefs, do not. Which par- the precise nature of perceptual representation. Sense- ticular belief a belief is, say the belief that crabs pinch datum theory, adverbialism and representationalism are hard, is a matter solely of which content it has, and any discussed in sections 6–8 below. phenomenal character that a particular instance of Finally, perceptual experiences are said to have *phe- a belief may have is irrelevant to its being that nomenal character. This means that there is ‘something belief. (Note that many people think beliefs themselves that it is like’ to undergo those experiences. Philosophy have no phenomenal character—they are simply usually of perception comes closest to philosophy of mind when accompanied by states that do.) it discusses the nature of perceptual phenomenal char- However, the traditional view that perceptual experi- acter and consciousness and whether a physical account ences are diVerent from sensations and beliefs has of it can be given. Recently, many philosophers have been challenged. On the one hand, arguing that sensa- sought an answer by enquiring about the precise rela- tions and perceptual experiences are not dissimilar, tionship between phenomenal character and represen- some representational views take sensations to be per- tation and whether the former can be explained solely in ceptions of one’s own body. For example, pains might terms of the latter. Section 8 elaborates on this issue. be thought of as states that represent damage or dis- 1. Perception, sensation, and belief order at a location in one’s body. On the other hand, 2. Unconscious perception arguing that perceptual experiences are more akin to 3. The common-kind view and the causal theory of beliefs than the traditional view, some doxastic views of perception experience hold that to have a perceptual experience 4. Disjunctivism is simply to believe, or to be inclined to acquire a belief, 5 that we are immediately perceiving some physical ob- . Empirically informed direct realist views V 6. Sense-datum theory ject or state of a airs by means of the senses. However, 7. Adverbialism this view is not widely endorsed for, plausibly, unlike 8. Representational theories belief-like states, experiences are necessarily conscious and occurrent and relatively unaVected by one’s other 1. Perception, sensation, and belief beliefs. Furthermore, the content of perceptual experi- Traditionally, philosophers have contrasted perception ence is sometimes held to be diVerent from that of belief with sensation. Perception was taken to be a process in various ways, notably by being non-conceptual. that involved states that represented—or that were about—something. For example, typical visual experi- 2. Unconscious perception ences had at a beach might represent sand, crabs, or the Can perception occur without a perceptual experience blueness of the sea. These experiences might accurately and without any conscious state? Recent empirical represent the beach or misrepresent it, if undergoing an Wndings have led some to answer positively (see uncon- illusion or hallucination. Sensations like pains, itches, scious perception). Consider the phenomenon of and tickles were not taken to be representational. For *blindsight, in which people claim to be blind. None- example, the feeling of pain was not taken to be theless, when asked to guess what is in front of them in ‘about’ anything—it was a mere feeling. At the same a forced-choice paradigm, they select the right answer time, philosophers have traditionally recognized that more frequently than chance. Is this evidence of uncon- sensations and perceptual experiences are alike in scious perception? It depends, Wrst, on whether the some respects. Both types of state have phenomenal subject really lacks a perceptual experience. This issue character, and which phenomenal character they have in turn depends on a commonly encountered question determines or partly determines, which particular kind in consciousness studies: to what extent can belief or of sensation or experience they are. *introspection about experience be inaccurate? Sec- Perceptual experiences have also usually been con- ondly, it depends on whether the accurate guessing trasted with beliefs. Although, like perceptual experi- behaviour of the subject warrants our claiming that ences, beliefs have been thought of as representing perception is occurring. This question arises because the world to be a certain way (in the case of belief, the not any state of a subject that reliably indicates a stimu- way the subject of the belief takes the world to be) they lus is a perceptual state. For example, to have a verruca are dissimilar in other respects. Beliefs need have no is to be in state that reliably indicates the presence of the phenomenal character (for example, they can be uncon- human papilloma virus, but it seems incorrect to think scious) whereas perceptual experiences necessarily have that having a verruca amounts to perceiving the virus.

503 perception, philosophical perspectives

3. The common-kind view and the causal theory of perience on the scene before the eyes. Thus, Lewis perception would claim that because Macbeth would have con- It is widely agreed that, when you perceive, things seem tinued to have the visual experience of a dagger even a certain way to you—but that things could seem just if it had been a claymore that had activated the machine; that way when you were hallucinating. In other words, his experience fails to be suitably dependent on the for each case of perception there is a possible case presence of a dagger, and this explains why he does of subjectively indistinguishable hallucination. The not see it. The merits of this and other responses have ‘common-kind’ view claims that perceptual experiences been widely debated. and their indistinguishable hallucinatory counterparts are fundamentally the same kind of state. The experi- 4. Disjunctivism ences will both have the same phenomenal character, Common-kind theories of perception stand in contrast will represent the same thing, and be similar in to metaphysical disjunctivism. The main claim of meta- their intrinsic mental properties. The experiences will physical disjunctivism has sometimes been articulated as only diVer in factors extrinsic to the experience such as being that there is nothing in common between the their causal origin, and whether they represent accur- experience involved in perception and that involved in ately. This reXects a common thought: the major pur- an indistinguishable hallucination. That claim is too pose of ascribing experiences to subjects is to try to hasty when one considers that both states will seem to capture how things seem to them. a subject to be indistinguishable from the experience The paradigmatic form of the common-kind view is involved in perception and both will be mental states. the causal theory of perception. On this theory, A more accurate articulation is that metaphysical dis- one perceives an object or property if and only if one junctivism claims that a state indistinguishable from has an experience that to some degree represents that perception is either a state that constitutes perception object or property, and that experience is caused in an (in the ‘good case’) or a state involved in hallucination appropriate way by that object or property. How accur- (in the ‘bad case’), and that these states exhibit further ate the representation needs to be is a tricky issue. We diVerences which amount to the states being diVerent want to allow that perceptual *illusions (where one ‘fundamental kinds’ (to use M. G. F. Martin’s termin- perceives, but inaccurately) are possible. For example, ology). What are these further diVerences? in the Mu¨ller–Lyer illusion two lines that are the same Some metaphysical disjunctivists claim that, in the length look unequal in length but we do not want to say good case, the objects and properties that one perceives that this prevents us from perceiving the lines. partly constitute one’s experience. This is not true in the The causal condition is required in order to account bad case where nothing is perceived. Such disjunctivists for veridical hallucinations—hallucinations that nonethe- not only deny the common-kind view but also deny less accurately represent the world. For example, im- that in perception the relevant experience is caused by agine that an evil scientist gives you a hallucinogenic the objects and properties perceived. This is because drug that makes you have a visual experience of a causes and eVects must be distinct states and this theory starWsh. At the same time, by chance, there just happens denies that the experience and what is perceived are to be a starWsh in front of you. The causal theory distinct. holds that such cases are not perception, as an appro- Some metaphysical disjunctivists think that experi- priate causal connection between what is perceived ences involved in perception represent, but others and the perceptual experience is missing. Spelling out claim that such experiences consist of a direct relation the nature of the connection is diYcult because or openness to the objects and properties perceived and of ‘deviant causal chains’. The diYculty is that, intui- therefore that representation is not required. This latter tively, not every causal connection suYces for percep- view is called a no-content view. One can see that tion even when the experience accurately represents some rationale for it would come from holding that that which causes the experience. Imagine that Mac- the perceived objects and properties partially constitute beth’s brain is connected to a machine, which has a the experiences and thus that representation is otiose. touch-sensitive pad that turns it on. The machine pro- One motivation for metaphysical disjunctivism is the duces in Macbeth an experience as of a dagger when it desire to maintain a naive or direct realism about percep- is activated. If such a dagger came to rest on the tion. This is the view that we are directly or immediately sensitive pad and activated the machine, Macbeth aware of objects in perception. Another motivation is an would not see it, despite it causing his experience. unwillingness to assert a certain form of infallibility con- David Lewis imaginatively addresses this problem, cerning our own minds: this view denies that if two claiming that what is required for perception is a suit- mental states seem the same to a subject then they able pattern of counterfactual dependence of visual ex- must be the same. A third motivation is a desire to 504 perception, philosophical perspectives explain how perception can ground our knowledge of states. This type of view often draws heavily on empir- the world. If hallucination and perception involve the ical work in psychology and neuroscience. An early same fundamental kind of state, but hallucination does version of the view is J. J. Gibson’s ecological approach. not yield knowledge, then how can perception do so? If According to Gibson, there is enough information in the experiences involved in perception and hallucination the ‘ambient optic array’—the pattern of light in space are importantly diVerent, as the metaphysical disjuncti- and time that directly stimulates an observer—such that vist would have it, then this problem might be overcome. the visual system of an observer need not process the (Epistemological disjunctivism is the view that the ex- direct stimulus to produce representations of the world. periences in the good and bad cases have a diVerent Perception consists, not in the forming of mental repre- epistemic status. Although metaphysical and epistemo- sentations, but in a direct response to invariances in the logical disjunctivism have a clear aYnity, the positions optic array. These invariances include surfaces and are distinct and neither entails the other.) edges in the environment and ‘aVordances’. AVordances One concern for metaphysical disjunctivism is how to are what the environment provides or invites, such account for illusion (where one sees, but inaccurately). as somewhere to shelter or something to eat. Gibson If illusory experiences are treated like veridical percep- also stresses that how you can act—what movements tual experiences then the illusory aspect is unaccounted you can make—will aVect what aVordances there are for. If they are treated like hallucinatory experiences for you and therefore what you can perceive. This the illusory aspect is explained but at the expense of type of view has recently been elaborated upon with the perceptual aspect. Disjunctivists themselves disagree the development of dynamic or sensorimotor theories about how to treat illusion. of perception. (Some *sensorimotor theories invoke Another concern is that metaphysical disjunctivism mental representations but insist that having such rep- typically has little to say about the mental states in- resentations requires an ability to interact knowingly volved in hallucination apart from the fact that with one’s environment.) they are indistinguishable from experiences involved in The assumptions that underlie these theories are perception (and indeed some varieties hold that there is highly controversial. Both thought experiments and em- nothing further that can be said). This lacuna provides pirical work question whether action is necessary for future work for disjunctivists. perception and whether Gibson’s assumptions about the A third concern is whether metaphysical disjuncti- ambient optic array and the working of the brain vism can provide a better epistemology of perception are correct. The extent to which mental representations than common-kind views, as some of its proponents are required in perception, and how to account for have claimed. The view certainly does not rule out illusion and hallucination if they are not, is the subject the possibility that we are hallucinating all the of much contemporary debate. time. Does it show how we could come to know 6. Sense-datum theory about the world if we do perceive? Opponents Returning to common-kind theories now, and to the complain that it does not as, if we are perceiving, we question of how such theories characterize perceptual do not know that we are (due to the indistinguishability experience, we Wnd that there are three main views: from hallucination). The disjunctivist is liable to claim sense-datum, adverbial, and representational theories. that one can know something without knowing that These, in turn, form the subject matter of the next one knows it. In the end, the debate seems to come three sections. down to whether one thinks that in the case of percep- Sense-datum theory was popular in the Wrst half tion there is something that is available to the subject that of the 20th century and is often attributed to earlier grounds knowledge that is not present in the hallucin- empiricist thinkers, such as Locke. This view endorses atory case. The disjunctivist will aYrm this, as they the following: (1) in hallucination there are no worldly will say that the world is directly available to the perceiv- objects that answer to what one seems to be aware of; ing subject. The opponent will deny it because the sub- (2) if it appears to one as if one is perceptually aware of ject cannot tell by reXection alone whether the world an object with a certain property then there must be is so available. something that one is aware of that has that property; 5. Empirically informed direct realist views and (3) the common-kind view. Point (2) is supported There is a view in the philosophy of perception that by the phenomenal character of experience: it at least shares certain features with some kinds of disjunctivism: seems as if we are aware of something when we are the endorsement of direct realism and the eschewing, at experiencing. Sense-datum theorists conclude that, in least to some degree, of the role of representational cases of hallucination, we are aware of mental or

505 perception, philosophical perspectives mind-dependent objects and, due to (3), that this is true 7. Adverbialism when we perceive the world too. These special objects The sense-datum theory claims every experience in- are termed sense-data. They do not exist in public space volves a subject’s act of awareness of some object. and they possess any property that they appear to have. A number of philosophers, starting around the middle (Note that when the term ‘sense-data’ was Wrst intro- of the 20th century, wished to reject such an ‘act–object’ duced it was used to refer to the direct objects theory but, nonetheless, wished to remain common- of perception, whatever they were. At that time, the kind theorists. Rather than holding that a subject’s see- discussion in the literature about perception was ing or hallucinating redness was to be explained by a whether sense-data were mental objects or public ob- subject bearing some relation to a red sense-datum, the jects. Many people at that time concluded that sense- subject was held to be experiencing in a certain manner: data were mental objects, and so the term came to be in this example, redly. This view is adverbialism. associated with only such objects.) One challenge facing adverbialism is explaining com- Realist versions of sense-datum theory hold that plex experiences. Consider two experiences: (a) an ex- we indirectly perceive the mind-independent world in perience as of a red circle to the left of a blue triangle virtue of directly perceiving sense-data. We can do and (b) one as of a red triangle to the left of a blue this because sense-data resemble or represent the square. A description like ‘experiencing redly and circu- mind-independent world. Irrealist sense-datum theorists larly and bluely and triangularly and to the leftly’ does are either idealists, who hold that what we normally not pick out (a) rather than (b). Adverbialists have take to be physical objects are simply collections of oVered solutions to these problems but some have actual sense-data, or phenomenalists, who hold that argued that any plausible solution forces the adverbial- they are collections of actual and possible sense-data. ists to elaborate their theory in a way that attributes Sense-datum theories have been heavily criticized. experiences with representational content and thus this One criticism is that there is no good reason to believe shows that a plausible adverbialism is just a species of (2); hence the motivation for the view is undermined. the representational view. To see this, note that one The issue turns on whether the phenomenology of way the adverbialist could answer the challenge is to experience provides a good reason. Another criticism say that when one has experience (a) one experiences in is that sense-data are ontologically queer mind-depen- a certain manner, the manner is one in which one seems dent objects—for example, they do not exist in public to be presented with a red circle to the left of a blue space but seem to have some spatial characteristics— triangle. This seems to be equivalent to saying that one thus they should not be countenanced. A third is that is having an experience that represents this. (The alter- not all aspects of phenomenal character seem to be native would be to think of the manner of experiencing explained by positing objects and their properties. as not essentially representational, and thus as more like For example, what would be the nature of sense-data sensations as construed by the traditional view and corresponding to experiences as of impossible Wgures or outlined in section 1.) experiences that seem to represent something indeter- Adverbial views, like sense-datum views, face worries minate? A fourth is that realist versions of the theory about how experiences can give rise to knowledge of provide a circular account of perception: perception of the external world. The debate on this matter follows a the mind-independent world is explained in terms of pattern similar to that outlined above for sense-datum perception of sense-data, which is not itself explained. theories, with the exception that adverbialists some- A Wfth is that realist sense-datum theories make a plaus- times claim that they are not committed to the prob- ible account of our knowledge of the mind-independent lematic indirect view of perception that the sense-datum world impossible. It is said that such views impose a theorist is, as they are not committed to mental inter- Lockean ‘veil of perception’ between us and the mind- mediaries. Whether this is any advantage is disputable. independent world. A sense-datum theorist might at- tempt to answer this charge by saying that a mind- 8. Representational theories independent world that causes us to have sense-data The representational or intentional theory of perceptual can be inferred as it is the best explanation of our experiences holds that perceptual experiences represent experience, and such inference is considered to be a the world. Reasons to think that they do include: (1) ex- source of knowledge in other Welds of enquiry, particu- periences seem to present the world to us, (2) ascribing larly science. Another attempt would be to hold that so perceptual experiences that represent is often the best long as having a perceptual experience is a reliable way way to explain and predict the behaviour of people and of forming true beliefs about the world then it is a way animals, (3) experiences seem to have correctness con- of gaining knowledge. ditions, that is, there is a way the world could be that

506 perception, philosophical perspectives would make what is represented true and a way that ent objects and properties that we are perceiving, rather would make it false, and (4) experiences are similar in than attending to other distinctive mental features of some respect to beliefs, which are the paradigm experience or any apparent non-representational proper- of representational states. If one believes that the sun ties of experience. If such a view were true, then perhaps is shining then the ‘that’ clause speciWes what is repre- what it is like to have an experience is exhausted by the sented (equivalently, speciWes the content) to which we experience’s contents. It might be tempting to think that take the attitude of belief. Similarly, if one seemed to see an exhaustive description of what it was like to have an that the sun is shining then the ‘that’ clause would experience would be an exhaustive account of the appar- specify the content of the perceptual experience. Of ent worldly scene before us. However, whether any course while the subject of a belief, by deWnition, takes version of the *transparency claim is true and, if it is, the content to be true, the subject of a perceptual what it shows about the mind is a topic of much recent experience need not do so, if, for example, they have debate. Further, there are a battery of examples in the reason to think that their senses are deceiving them. recent literature in which, it is claimed, there are experi- The most common form of representational theory is ences that have diVering phenomenal character yet the that which adheres to the common-kind commitment same representational content and vice versa. Examples and the causal theory of perception. On such a view, of this kind would provide counter-examples to the when perceiving in a non-illusory manner the experience representationalist’s identity or supervenience claim. with content will accurately represent the world and be Whether any of these constitute successful counter-ex- caused in the right way by it. Unlike the sense-datum amples to representationalism is an open question. view, this view of perception of the mind-independent Representationalism is often conjoined with a natural- world holds that it is direct and occurs partly in virtue of istic theory of representation and it is hoped that a natur- being in a state with content. However, other versions of alistic theory of phenomenal character or consciousness the representational view are possible, in particular one will be the result. This view is hotly disputed in the might think of experiences as representational states but current literature, with many people claiming that no nonetheless reject the common-kind view. naturalistic theory of representation can account for the There are very many debates concerning the nature of the phenomenal character of certain unusual experiences content of perceptual experiences. (See, contents of con- known to exist, or our intuitions about the phenomenal sciousness; externalism; intentionality; non-conceptual character of experience in various hypothetical cases. content; representationalism.) Objections to representational theories. Returning now to Representation and phenomenal character. Representational consider representational views more generally, it was views diVer on the relationship that they believe holds stated above that representational views can be between representational content and phenomenal conjoined with either a common-kind and causal view character. Some views hold that there are some or a disjunctivist view. Some objections to these views phenomenal aspects of experience that are not are therefore simply versions of objections to common- representational at all. Such views therefore are faced kind and causal views or disjunctivst views, some of with supplementing their account of experience to which have been mentioned above. explain these aspects. Some theorists become More particular objections to representational views physicalist about such aspects, usually identifying them of perception focus on what account can be given with states of the brain (see physicalism). Other of hallucinatory states. Consider the content of a per- theorists could treat such aspects as mental primitives ceptual experience involved in an accurate perception of that cannot be given further explanation. a starWsh (the good case) and one involved in an indis- Other representational theorists maintain that phe- tinguishable non-veridical hallucination (the bad case). nomenal character either supervenes on or is identical Both states would seem to represent the same thing: a to the representational content of experience. This view starWsh, or a starWsh-shaped thing. Thus, it is tempting is often called representationalism. (Unhelpfully, this term to think that the content is the very same. However, in is sometimes used to refer more broadly to what I have the good case if the content is the object perceived—the been calling ‘representational theories’.) It is often held starWsh—then it is not obvious that that starWsh is the not only about perceptual experiences but also about all content in the bad case, for that starWsh might not exist states with phenomenal character, such as sensations and in the bad case. Two consequences follow. First, we emotions. Representationalists are often motivated by seem now to be denying that the content is really the their belief that experience is transparent, that is to say same, certainly the same in all respects, in both cases. their belief that when we introspect we Wnd that we are Second, we still need to say what the content is in the only paying attention to the seemingly mind-independ- bad case. Some representationalists have adverted to 507 perception, unconscious holding that the contents of hallucinatory experiences conscious representation. The idea is to eliminate pos- are intentional inexistents. That is to say, the contents sibly confounding conscious perceptual inXuences en- are objects that do not exist. However, now an account tirely, thereby enabling strong conclusions that of intentional inexistents is required. If they are to be obtained eVects, if any, reXect purely unconscious per- treated really as ‘things’ then there is the worry that ceptual processes. Usually, one task is used to index the view seems to collapse into a view with as unpalat- conscious perception (e.g. identiWcation, which assesses able ontological commitments as the sense-datum view. the direct, intentional use of stimulus information), If such a view does not treat intentional inexistents in while another task (often *priming, wherein the unin- this way then the worry is that such talk is just a way tended, indirect inXuence of an initial stimulus on the of labelling the problem that in hallucination subjects processing of a later stimulus is examined) is used to appear to be confronted by objects when none are index unconscious perceptual eVects. If successful, such there. In this case, the problem of accounting for the paradigms could potentially reveal much about not *phenomenology of hallucinatory experience, in which only unconscious mental processes, but even fundamen- one appears to have contact with objects, remains. tal aspects of consciousness itself. Many believe, for One way to get round this problem is to deny that in example, that consciousness somehow enables more the good case the content is the object perceived—the complex and Xexible mental processes than are possible starWsh. Instead, one might hold that the content is an purely unconsciously. By varying whether stimuli are abstract object—say the proposition that a starWsh is unconsciously vs consciously perceived, we can empir- before one—and that this is the content in the bad ically test such hypotheses. case too. This solution is tempting, but it is resisted by For unconscious perception paradigms to serve this some representationalists who want to give a naturalis- role, however, requires solving a deceptively simple but tic theory of content and who think that giving a nat- surprisingly tenacious methodological problem: How can uralistic explanation of how the mind grasps, or stands we really be sure that putatively unconscious eVects are in relation to, such an entity is problematic. Other not, instead, actually weakly conscious after all? Given that representationalists believe that existing naturalistic the- unconscious perception currently enjoys relatively broad ories of content can meet this challenge. Of course, if acceptance, one might think that some deWnitive meth- one is not motivated by naturalism, then one might odological breakthrough had been achieved. Unfortu- happily hold that it is simply a primitive fact that ex- nately, however, this is not the case. Consequently, until periences seem to present objects to their subjects these core issues are satisfactorily addressed, the currently and thus that the notions of content and of representa- positive consensus runs the risk of simply perpetuating the tion are not to be given further explanation. boom and bust cycle of critical acceptability that has FIONA MACPHERSON plagued unconscious perception throughout its controver- Armstrong, D. M. (1961). Perception and the Physical World. sial history. Moreover, and just as importantly, careful Chisholm, R. (1957). Perceiving: A Philosophical Study. consideration of these issues is substantively informative Crane, T. (1992). The Contents of Experience: Essays on Perception. in its own right—raising, for example, fundamental ques- Gendler, T. S. and Hawthorne, J. (2006). Perceptual Experience. tions about how consciousness should be indexed, how Gibson, J. (1979). The Ecological Approach to Visual Perception. conscious and unconscious processes interact, and the role 1965 Pro- Grice, H. P. ( ). ‘The causal theory of perception’. In of volition. Indeed, vigorous debate on these issues con- ceedings of the Aristotelian Society, Supplementary Volume 35 , . tinues even among unconscious perception proponents, Jackson, F. (1977). Perception: A Representative Theory. yielding sharp disagreement on which data are valid and Haddock, A. and Macpherson, F. (2008). Disjunctivism: Percep- tion, Action, Knowledge. their appropriate interpretation. Hamlyn, D. W. (1961). Sensation and Perception: A History of the 1. How should consciousness be indexed? Philosophy of Perception . 2. Modern unconscious perception models 2003 Perception Maund, B. ( ). . 3. Concluding remarks Noe¨,A.(2004). Action in Perception. —— and Thompson, E. (2002). Vision and Mind: Selected Read- 1. How should consciousness be indexed? ings in the Philosophy of Perception. Almost everyone agrees that conscious perception cov- Robinson, H. (1994). Perception. aries with stimulus intensity, typically manipulated Smith, A. D. (2002). The Problem of Perception. 1995 Ten Problems of Consciousness by varying stimulus duration, masking intensity, or Tye, M. ( ). . both. Whereas strong stimuli are plainly visible, con- scious perception diminishes as stimulus strength is re- perception, unconscious Unconscious perception duced—Wnally disappearing altogether when stimuli are paradigms seek to examine the eVects of sensory stimuli weak enough. But exactly how should this threshold be (typically visual) that are rendered too weak to achieve deWned? It turns out that there are two basic alternatives. 508