In contemporary philosophy, the phrase ‘the contents ofperception’ means, roughly, what is conveyed to the subject byher perceptual experience. For example, suppose you are looking into apiano at the array of hammers and strings. There will be a way thesethings look to you when you see them: they will look to have a certainshape, color, texture, and arrangement relative to one another, amongother things. Your visual experience conveys to you that the piano hasthese features. If your experience is illusory in some respect thenthe piano won’t really have all those features. But even then,there will still be something conveyed to you by your experience.Three main questions will be addressed in this entry. First (Sections1–2), what are contents and what is their relation toexperiences? Second (Sections 3–7), which contents are contentsof experience? Finally (Section 8), in virtue of what do experienceshave contents, when they do?

• 1. What are Perceptual Experiences?
• 2. Contents as Accuracy Conditions
  • 2.1 Do Experiences Have Accuracy Conditions?
  • 2.2 Beliefs and Experiences
  • 2.3 Content and Attitude
• 3. Varieties of Content
  • 3.1 Russellian Content
  • 3.2 Possible-Worlds Content
  • 3.3 Fregean Content
  • 3.4 Indexical Content
  • 3.5 Multiple Contents
• 4. The Representation of Properties
  • 4.1 Which Color Properties are Represented?
  • 4.2 Which Spatial Properties are Represented?
  • 4.3 Are High-Level Properties Represented?
• 5. The Representation of Objects
  • 5.1 Object-Involving, Gappy, and Existential Contents
  • 5.2 Causal and Non-Causal Contents
  • 5.3 Are Ordinary Objects Represented in Experience?
  • 5.4 Mind-Independence
• 6. Concepts and Content
  • 6.1 Conceptualist Views of Perceptual Contents
  • 6.2 Varieties of Nonconceptual Content
• 7. Content and Phenomenology
  • 7.1 Representationalism about Phenomenology
  • 7.2 Does Phenomenology Determine Content?
• 8. Theories of Intentionality in Experience
  • 8.1 Externalism vs. Internalism about Content
  • 8.2 Functionalist Informational Theories of Experience
  • 8.3 Phenomenal Intentionality
• 9. Directions for Future Research
• Bibliography
• Academic Tools
• Other Internet Resources
• Related Entries

1. What are Perceptual Experiences?

This entry will focus exclusively on the contents of perceptualexperiences. It is definitional of experience, as the term isused here, that they have some phenomenal character, or more briefly,some phenomenology. The phenomenology of an experience is what it islike for the subject to have it. At any given waking moment, onenormally has experiences in some (perhaps all) of the five sensemodalities, along with proprioceptive experience of some sort. Likeoccurrent emotions and background moods, these contribute toone’soverall experience. Each of five sense modalitiesseems to be accompanied by a distinctive kind of phenomenology.Sometimes the boundaries between them can be hard to draw, for examplebetween olfactory and gustatory phenomenology, kinesthetic experienceand haptic-tactile experience (that is, tactile experiences involvingmovement rather than just pressure sensation), and perhaps, in somecases, between the latter two and visual experience (on the last pointsee Presset al. 2004, Kennettet al. 2001,Fotopoulouet al. 2009; on distinguishing the senses, seeStokes, Matthen and Briggs 2014, Wilson and Macpherson 2018). Despitethese difficulties, some rough distinctions can be drawn: thedifference between seeing red and seeing blue is not a difference inhow things sound, or in how one’s body feels.

In what follows, “experiences” will be restricted toexperiences in any of the five modalities or kinesthetic experience orsome combination. In the analytic tradition, most of the questionsdiscussed in this entry have been addressed with regard to visualexperience and pain. In contrast, the phenomenological traditionexplored similar questions with respect to other modalities. Forinstance, Husserl’s exploration of auditory experience ofhearing a melody is central to his discussion of the experience oftime (see the entries onphenomenology,auditory perception, andtemporal consciousness). In principle, however, the questions addressed here could be raisedwith respect to any kind of experience.

2. Contents as Accuracy Conditions

When one speaks of the contents of a bucket, one is talking about whatis spatially inside the bucket. An analogous use of “thecontents of perception” would pick out what is ‘in themind’ when one has a perceptual experience. In contrast, whenone speaks of the contents of a newspaper, one is talking about whatinformation the newspaper stories convey. Many contemporary uses of“the contents of perception” take such contents to beanalogous to the contents of a newspaper story, rather than thecontents of a bucket. This notion of content can straightforwardlyaccommodate the idea that there is such a thing as the‘testimony of the senses’.

One influential version of the idea that the contents of perceptionare analogous to the contents of a newspaper story holds that thecontents of an experience are given by the conditions under which itis accurate. What an experience conveys to the subject, according tothis conception, is that those conditions are satisfied.

On the conception of contents as given by accuracy conditions, thereis a broad analogy between the contents of experience and the contentsof thoughts and utterances, in that both contents are assessable foraccuracy. Suppose I utter the sentence “Dogs are lively, loyalcreatures” and thereby express my belief that dogs are lively,loyal creatures. The content of my utterance is what I assert, and thecontent of my belief is what I believe—in both cases, that dogsare lively, loyal creatures. Usually these contents are thought of assome kind of proposition—an abstract object that is the kind ofthing that can be true or false (see the entry onstructured propositions). Notice that both the utterance and the belief are assessable fortruth: their content is true just in case dogs really are lively,loyal creatures.

In the case of beliefs and utterances, it is not just their contentsthat are assessable for accuracy (where accuracy is truth); thebeliefs and utterances themselves are as well. A belief or utteranceinherits its truth-value from the truth-value of its content. Thisstands in contrast to desires or hopes, which may have contents thatare assessable for accuracy, but are not themselves so assessable. Ifexperiences have accuracy conditions, then they are like beliefs andutterances in this respect, and unlike desires and hopes.

The conception of the contents of experience as given by its accuracycondition can be motivated by the idea that one can be misled byone’s experiences. Suppose you see a fish while unwittinglylooking in a mirror. It may look as if there is a red fish in front ofyou, when in fact the red fish you see is behind you and there is nofish at all in front of you. Similarly, in auditory or olfactoryhallucinations, one may seem to hear voices when in fact no one isspeaking, or to smell an odor when in fact nothing is emitting thatsmell. In phantom limb pain, one feels pain as located whereone’s limb used to be but is no longer. These are cases of beingmisled by one’s senses, and it is natural to say that in thesecases things are not as they appear to be.

Once one recognizes the datum that one can be misled by one’ssenses, it is a further claim that experiences themselves can beassessed for accuracy. It is a further claim, because one might bemisled by something that is not itself assessable for accuracy: e.g.,one might be misled by the car in the driveway into thinking that yourneighbor is home. So the fact that something plays the role ofmisleading you does not show that it is assessable for accuracy. Butin the case of experiences, this further claim is motivated in thefollowing way. Given an experience—either one we actually have,or a hypothetical one—we at least sometimes have intuitionsabout whether the experience is accurate (“veridical”) orinaccurate (“falsidical”). To this extent, we seem to beable to assess experiences for accuracy.

When we assess an experience as accurate or inaccurate, we considerhow things are in the world. This suggests that experiences which canbe assessed for accuracy are associated with accuracyconditions: conditions under which the experience isaccurate. When we have intuitions about whether an experience isaccurate, we at least sometimes also have intuitions about theconditions under which it would be accurate. (For discussion of suchintuitions, see Siewert (1998), chapter 7, Schellenberg (2018). Goingwith this, we seem to have intuitions about when experiences areincorrect.

In large part, our intuitions about when experiences are accurateconcern objects and properties. For instance, many have the intuitionthat the following experience is falsidical: you see a fish thatappears blue to you, but is actually red. According to the intuition,the fish’s failing to be blue suffices for the experience to befalsidical. Intuitively, for the experience to be accurate, the fishin question must really be blue. It is an open question how specificand how extensive such intuitions are, hence to what extent suchintuitions can reveal which accuracy conditions experiences have,assuming that they have accuracy conditions at all. But so long asthere are some such intuitions, there is a prima facie case forexperiences having accuracy conditions.

2.1 Do Experiences Have Accuracy Conditions?

Some philosophers deny that experiences have accuracy conditions. Suchphilosophers must offer a different account of what is happening incases where one is misled by one’s senses, and they must dismissor explain away putative intuitions about the accuracy or inaccuracyof experiences. A natural way to do so is to confine accuracy andinaccuracy to the level of belief, further downstream fromexperience.

There are a number of positions that deny that experiences haveaccuracy conditions (Wilson and Locatelli 2017). All these positionsdeny this claim on the grounds that it is not supported by either themetaphysics of experience or by its phenomenology. The positionsdiffer, however, on their positive views of the metaphysics andphenomenology of experience.

The first position is the view that experiences are raw feels, orsensory affectations of the subject, that do not purport to representthe world in any way at all. The eighteenth-century philosopherThomas Reid held a view of this sort: Reid argued that experiences have a sensorypart that is a mere raw feel, and that is independent of judgmentsabout how things are in the environment. (He took the latteressentially to involve concepts). One might, on this view, speak morebroadly about experiences as having two parts, sensory part and ajudgment part; or one might speak more narrowly about experience asbeing limited to the sensory part that is typically accompanied byjudgment. Which way one speaks is merely a matter of terminology. Thecore feature of the view is its cleavage between the sensory aspect ofexperiences, on the one hand, and assessability for accuracy, on theother. Given a judgement that occurs simultaneously with a sensation,one could in principle make that very same judgment without having anysensation, and one could in principle have those very sensationswithout making any judgments. That is the force of saying that thesetwo aspects, sensation and judgment, are independent. If experiencesare understood narrowly as being limited to the sensory part, thenthey are not assessable for accuracy; and if they are understoodbroadly as having both a sensory part and a judgmental part, then itis only the latter that is assessable for accuracy.

The second position is a version ofadverbialism about visualexperiences, according to which such experiences are nothing more thanmodifications of a subject that can be characterized by specializedadverbs specifying how one is visually ‘appeared-to’.E.g., normally, when one sees a red tomato, one is appeared to red-lyand round-ly. Visual experience, on this view, does not by itself evenpurport to present one with objects or their properties. There arethus no accuracy conditions associated with such experiences. Instead,on this view, to be misled by one’s sense is to end up with afalse belief that is caused by one’s experience. (For furtherdiscussion of adverbialism, seethe problem of perception; also Chisholm 1957, Ducasse 1942, Tye 1984, Sellars 1975, and forcritical discussion, Jackson 1977). Strictly speaking, it should benoted that so long as the core thesis of adverbialism is thatexperiences are modifications or properties of the subject,adverbialism leaves open whether experiences are assessable foraccuracy or not. For all that core thesis says, being appeared-toF-ly could be a way of representing that something isF. Historically, adverbialists’ main motivation was tooppose the act-object conception of experience, according to whichexperiences are structured by an act of awareness, on the one hand,and an object of awareness, on the other. Early formulations of theview presented it as a view according which experiences were notassessable for accuracy.

The third position is that at least some perceptual experiences comein degrees that correspond to degrees of uncertainty. Compare: thereis straight answer to the question “Is the sound I hear acoyote?”, when you are uncertain whether the sound comes from acoyote or a human baby. You cannot simply answer Yes or No to expressyour opinion; you have to hedge. If you are in such a state ofuncertainty, it might seem, then there no way the world could be inorder for your state to be accurate or inaccurate. By analogy, on thethird position, experiences sometimes reflect uncertainty in a waythat leaves them with no accuracy conditions (Madary 2012, Morrison2015, Munton 2016, Vance 2020, Raleigh and Vindrola 2020). There is afourth position that denies that experiences have accuracy conditions.This position starts with the claim that some experiences consist in achunk of the environment—for instance, a lavenderbush—being perceptually presented to a subject, so that both thelavender bush and the perceptual relation between it and the subjectare constituents of the experience. (Note that this is more specificthan the view that says merely thatperception of publicobjects is ‘direct’. That view is silent on what thenature of perceptual experience is). According to one version of thisview, when you see a lavender bush, some of its properties arepresented to you, and your experience consists in your being sorelated to the bush and those of its properties that are presentingthemselves to you. You could have an experience that wasindistinguishable from one of so seeing that very lavender bush: as itmight be, an experience of seeing a different lavender bush, or ahallucination that was indistinguishable from the original experience.But neither of these experiences, according to this view, would be thesame sort of experience as the experience of seeing the originallavender bush. Martin (2006) and Smith (2002) call this view“naïve realism”. Another version of this approach,developed by Johnston (2006) and (2014), develops the idea thatperceivers are presented with instances of properties rather than withuniversals. Note that both Martin and Smith’s versions aresimply silent on the nature of hallucinatory experiences. (For more onnaïve realism see the entry onthe disjunctive theory of perception).

As stated so far, naïve realism is neutral on whether anyexperiences have accuracy conditions. A version of the view, however,explicitly denies this. This version holds that experiences areanalogous to chunks of the environment that clearly lack accuracyconditions. The lawn chair on the deck is not the kind of thing thatcan be accurate or inaccurate (though it may figure in the accuracyconditions of an utterance, such as an utterance of “that is alawn chair”). Travis (2004, 2013) develops naïve realism injust this way. He also offers further reasons for denying thatexperiences have accuracy conditions. Travis agrees that in having anexperience one may represent that such-and-such is the case. He holds,however, that there is no unique ‘taking’ associated withan experience, and experience is only ever an input to representation,never the result of it. One of Travis’s examples is this: if yousee Sid and Pia touch hands at the dinner table, you make take them tobe trysting. But considered independently of this‘taking’, your experience simply consists in beingpresented with an unrepeatable, particular chunk of the environment,namely Sid and Pia at the table touching hands. (The notion of being‘presented with’ is left as primitive). Accuracyconditions come into the picture only after you take the environmentas it is presented to you to be some specific, potentially repeatableway. So according to Travis, the experience does not map onto anyunique set of accuracy conditions. Since illusions involve inaccuracy,by Travis’s lights illusory experiences are experiences thathave been ‘taken’ in such a way that the‘taking’ is false.

A fifth position that denies that experiences have accuracy conditionssays that experiences are raw feels, but also holds that allexperiences have the same structure as perception of objects. On thisview, experience consists in immediately perceiving private objectsknown assense-data. According to sense-datum theories, whenyour experience is a case of perceiving a public object, such as awhite wall that looks red, youindirectly perceive the wallby immediately perceiving a red sense-datum. In general, sense-datareally have the properties that public objects look to have. In thisway your having (i.e., immediately perceiving) the sense-data you dois supposed to account for what it is for a public object to appear toyou to be a certain way; and you have sense-data whether or not thereis any public object that you indirectly perceive by immediatelyperceiving them.

It should be noted that if the core commitment of sense-datum theoriesis that one immediately perceives mental objects in experience, thensense-datum theories are neutral on whether experiences are assessablefor accuracy. From the fact that one perceives (or immediatelyperceives) a particular object (be it a public object or a privatemental one), nothing at all follows about whether the experience of soperceiving is assessable for accuracy or not (for more on sense-datumtheories, see the entries onsense data and theproblem of perception). Jackson (1977) defends a version of the sense-datum view according towhich experiences arenot assessable for accuracy.

All of these positions except the third one could be seen as offeringconstruals of the contents of the perceiver’s mind, where theseare analogous to the contents of a bucket. Such construals contrastwith the newspaper model of perceptual contents, and more specificallywith the conception of contents as the accuracy conditions ofexperience. Henceforth ‘the contents of experience’ willbe restricted to a notion that says that the contents of an experienceare the conditions of accuracy associated with it. The claim thatexperiences have contents in this sense is substantive, and, as wehave seen, is denied by some philosophers. But many philosophersaccept that experiences have contents in this sense, and thisconception of the contents of experience dominates the recentphilosophical literature on perception. Defenses of the thesis thatexperiences have contents can be found in Byrne 2009, Pautz 2010,Schellenberg 2018, and Siegel 2010a and 2010b, and criticisms of thosedefenses in Breckenridge 2007 and Wilson and Locatelli (2017).

2.2 Beliefs and Experiences

There are many ways of developing the idea that experiences areassessable for accuracy. One idea is that the contents of experiencederive in some fashion from the contents of beliefs, so thatexperiences bear some constitutive link to beliefs. Three sorts ofconstitutive links to belief have been discussed in the literature.The first is that experiences areacquisitions of beliefs;the second is that they aredispositions to form beliefs; thethird is that they aregrounds of dispositions to formbeliefs. A fourth position simply identifies experiences with beliefsabout how things look (Gluer 2009), or even more simply, with beliefswhose content characterizes the way things look (Bryne 2016).

First, suppose experiences areacquisition of beliefs formedby a standard use of a perceptual apparatus. When applied to visualexperiences, this view says that for as long as you see something thatlooks like a red, shiny fish in front of you, you are acquiring abelief that there is a red, shiny fish in front of you. The content ofexperience, on this view, will be the same as the content of thebeliefs with whose acquisition the experience is identical. And whatmakes a content the content of an experience, as opposed to some otherkind of mental state, will be whatever makes it the content of thosebeliefs. (This view is critically discussed by Pitcher in his 1971,chapter 2.)

A standard objection to so identifying experiences with theacquisition of beliefs is that one may not believe that things are theway they appear. (This objection is discussed by Pitcher, op cit, andby Armstrong 1968). For instance, if you have background knowledgethat despite appearances there is no red shiny fish in front of you,then you will not believe that there is one. To accommodate thisobjection, the relation between experiences and the acquisition ofbeliefs must be something other than identity.

Second, suppose that experiences aredispositions on the partof the subject of the experience to form beliefs. This view isdefended by D. Armstrong (1968), G. Pitcher (1971), D. Dennett (1991).Which disposition an experience is, on this view, depends onhow things appear to the subject in the experience. For instance, ifyou seem to see a red shiny fish in front of you, then (on this view)you have a disposition to believe that there is a fish in front ofyou, and your experience is identical with this disposition.

If experiences are dispositions on the part of the subject to formbeliefs about how things appear to her, a natural accompanying viewabout the contents of experience is that these just are the contentsof the beliefs one is disposed to form. According to this view, whatmakes some contents contents ofexperience, as opposed tobeing contents of some other kind of mental state, is in part that oneisdisposed to believe those contents.

One could then develop the idea that experiences are assessable foraccuracy in the following way:

1. Having an experience consists in being disposed to form certainbeliefs.
2. The conditions under which the experience is accurate (i.e., itscontents) are the same as the conditions under which the beliefs oneis constitutively disposed to form in having the experience areaccurate. (Equivalently, the contents of experience are the contentsof the beliefs one is constitutively disposed to form in havingit).

Pitcher’s view in his 1971 is (a). Claim (b) is a naturalextension of (a), given the assumption that beliefs have accuracyconditions. (Pitcher doesn’t talk explicitly about accuracyconditions).

One difficulty with (a), and so with (a) and (b), is that it leavesopen how the relevant dispositions must be grounded. You could bedisposed to believe that there was music playing next door becausesomeone told you it was, and yet, it seems, not hear or even seem tohear any music: you might be totally deaf, for instance. In this caseyou would not have any auditory experience of music, despite beingdisposed to believe that music is playing. So proponents of (a) seemwell-advised to specify an appropriate ground for the dispositionsthey identify with experiences, so that they do not misclassify asexperiences events that intuitively are not experiences.

This brings us to the third position on the relation betweenexperience and belief: that experiences are perceptual grounds ofdispositions to form beliefs. Just what a specifically perceptualground is has to be spelled out—a natural suggestion is that itrequires a certain etiology involving sensory organs, or functionalequivalents. Given this third position, a natural accompanyingconception of the contents of an experience is that they are the sameas the contents of beliefs that one is disposed to form in having theexperience (claim (b) above). However, another version of the viewmight hold that although the experience itself is constitutivelylinked to beliefs, it does not havecontents that areconstitutively linked to the contents of the beliefs.

Claim (b) above, which identifies the contents of an experience withcontents of beliefs one is disposed to form in having the experience,assumes that the contents of both kinds of state can be the same. Somephilosophers have argued that the contents of such states must differ.If this is correct then none of the views here mentioned can becorrect as stated. (For discussion of how the views might be modified,seeSection 6.2 below, and the entry onnon-conceptual content.)

Another objection to views that posit constitutive links betweenexperiences and beliefs, or between experiential contents and beliefcontents focuses on primitive creatures that can have experiences,even though they lack the capacity and disposition to form beliefs.Such creatures would seem to be impossible if experiences wereconstitutively linked to beliefs, as the views discussed here hold.This objection seems to apply to the position that identifiesexperiences with beliefs about the ways things look (defended byGlüer 2009). In contrast, Glüer’s (2009) view seemsimmune from the objection that assimilating perceptual experience tobelief would entail that subjects of known illusions havecontradictory beliefs, and so are irrational. This objection appliesto Byrne (2016), who argues that this consequence is plausible.

2.3 Content and Attitude

Instead of holding that the perceptual content derives from thecontent of belief, one might instead hold that the contents ofperception and the contents of belief are in some sense analogous, andmore generally that experiences and beliefs are similarly structuredby contents on the one hand, and a relation to those contents, on theother.

Beliefs are often thought to be relations to contents, and contents ofbeliefs are often thought to be contents that one can stand in otherrelations toward, besides belief. Such relations are sometimes called‘attitudes’, where examples of these are supposed to benotions from folk-psychology such as hoping, wanting, or supposing.However, it is useful to construe the notion of ‘attitude’more broadly, so that it is more generally a way of having content.Subpersonal states of visual processing, if they are informationalstates with accuracy conditions (and so with contents in our sense),will instantiate ways of having content; so would the propositionalattitudes as they are traditionally conceived. From now on‘attitudes’ will be relations in which one can stand tocontents, when one has a contentful mental state. Both philosophersand psychologists have found it useful to mark such a distinction.^[1] More recently, Crane (2003) calls attitudes in this broader sense‘intentional modes’. (See also Searle 1983, Chapter 1.)These are all distinctions between aspects of mental states, where onefeature (akin to content) characterizes the subject-matter of themental state, and the other (akin to attitude) characterizes a mode ofentertaining it.

If experiences have contents, then they will be structured by attitudeand content: the experiences will have content in an experiential way.Several substantive questions arise about what experiential attitudesare.

First, it is an open question how finely grained the experientialattitudes may be. A natural proposal is that there are differentexperiential ways of having content corresponding to the differentsense modalities. But it is far from given that the modalities markout natural kinds of sensory experience, let alone the only naturalkinds of such experience, since it is a complicated question how todistinguish the different kinds of sensory experience from oneanother. Furthermore, there are multi-sensory phenomena, such assynesthesia, which defy straightforward categorization as a singlekind of sensory experience. Another problem case may be the experienceof bodily actions, in which (say) visual, kinesthetic and tactileexperiences are intertwined. (See Presset al. 2004, Kennettet al. 2001 and Fotopoulouet al. 2009 fordiscussion of cases in which visual experience enhances tactile acuityand kinesthetic experience generally).

Second, it is a substantive question what constitutes anexperiential way of having content. Tye (1995, 2000) hasproposed that what makes a state have a content in the experientialway is that it has a certain functional role: it is‘poised’ for a certain kind of use in the cognitivesystem; and in addition it has non-conceptual content (seeSection 6.2) and is not individuated by any objects that it may represent. Anotheroption is that what makes a state have content in an experiential wayis that it has the appropriate sort of phenomenology (see Chalmers2004). A third option is that the experiential way of entertainingcontent overlaps with the belief way of entertaining it—whateverexactly that is. This last option seems to underlie a very influentialidea developed in Armstrong and Pitcher’s proposals, thatexperiences are belief-like in an important way. (Cf.Pendelbury’s discussion of sensuous belief in his 1986;Heck’s discussion of ‘assertoric force’ in his 2000;Martin’s discussion of ‘coerciveness’ in his 2002).These options are not exclusive.

Third, it is a substantive question to what extent attitude andcontent are independent from one another. Once we fix on anexperiential way of entertaining content, what if anything limits therange of contents that can be entertained in that way? Going in theother direction, once we fix on a content, are there limits on whatattitudes may be taken toward it? How this question is answered bearson several neighboring issues. If any content that can be entertainedexperientially can also be entertained unconsciously, then it will notbe contents alone that give an experience its phenomenality. Thequestion also bears on the question of whether there isnon-conceptual content (seeSection 6.2 and the entry onnon-conceptual content).

Fourth, it is sometimes assumed that if experiences are structured byattitude and content, then experience types will be individuated bytheir specific attitude plus their contents, so that if an experienceE₁ with contentC is type-identical to anexperienceE₂, thenE₂ willalso have contentC, andE₁ andE₂ will instantiate the same experientialattitude. The source of this idea may be an analogous claim aboutbeliefs: namely, the claim that if beliefB₁ withcontentC is the same belief-type asB₂,thenB₂ will also have contentC.However, it should not be taken as a consequence of the claim thatexperiences have content that their contents individuateexperience-types. Rather, this is a substantive claim that needs bothrefinement and defense. Even if beliefs were individuated by theircontents, that would not suffice to show that experiences are as well.Some additional reason would be needed to think that experiences andbeliefs are alike in this respect.

If one holds that experiences are not constitutively linked to beliefsbut are similarly structured by attitude and content, then this is onemotivation for thinking there are experiential attitudes distinct frombelief. However, one could hold that experiences are so structured (byattitude and content), while also holding that experiences have theircontents in virtue of their connections to beliefs.

3. Varieties of Content

So far, we’ve reviewed two broad ways of developing the ideathat experiences have accuracy conditions: on the first way, thesederive from the contents of beliefs; on the second way, theydon’t. On either of these two approaches, the question arises ofwhat kinds of content experiential contents are.

Whichever of the two approaches are taken, a basic constraint on thecontents of an experience is that these should reflect (at least some)intuitions about when experiences are accurate or inaccurate. Anothercommonly-held constraint is that the contents must be adequate to itsphenomenology: any proposal for what contents a given experience hasmust in some way reflect the phenomenology of the experience. Thenotion of phenomenal adequacy has considerable intuitive force.Theories of the contents of experience are often defended on thegrounds that they “do justice to the phenomenology” (forexample, see the discussion inSection 6.2 of the ‘richness’ motivation for non-conceptual content,the discussion inSection 3.4 of indexical content, and Chalmers (2006)) or are criticized on thegrounds that they “get the phenomenology wrong” (e.g.,Smith 2003, 259ff; Dainton 2000, 115; Johnston 2004, 120). Theseappeals presuppose that the contents of experience have to in some wayreflect the phenomenology of the experience.

This section reviews the main theories of what kinds of contentsexperience contents are.

3.1 Russellian Contents

Some philosophers of language defend the view that an utterance of theform ‘N isF’, where‘N’ is an ordinary proper name, expresses astructured propositional content consisting of the referent of thename ‘N’ and the property expressed by thepredicate ‘F’. For instance, if the sentenceuttered is ‘Frida is cold’, then the associated contentwould consist in Frida herself and the property of being cold.Contents consisting of referents of singular terms and propertiesexpressed by predicates are known asRussellian, because atone time Russell advocated the view that singular terms contributeonly their referents to the proposition expressed by sentences inwhich they occur (or utterances of such sentences). Our example usesthe ordinary proper name ‘Frida’; but Russell himselfthought that such names were not genuine singular terms, and deniedthat public objects ever got into Russellian contents. In contemporarydiscussions, such contents are also calledsingular contents,because they include the referent of a singular term itself ratherthan a description that picks out different objects depending on whichthing in the world satisfies it.

One of the simplest theories of the contents of experiences says thatexperiences have contents that are Russellian, and that these contentsconsist of the very things that appear to have certain properties andthe very properties that they appear to have. For instance, if one hasa visual experience representing a red cube, the contents ofexperience on this view would be the structured proposition containingthe object one sees that looks like a red cube and the property ofbeing a red cube. This content would have the form[o,P], whereo is the very object one isperceiving in having the experience, andP is the propertythato appears to have. Note that this is a view both aboutthe nature of the contents themselves, and about the conditions underwhich such experiences have them: an experience has such contents whenthe subject isseeing an individual that appears to have theproperty that occurs in the content.

The strongest Russellian contents include both objects and properties.In practice, the Russellian content of a complex visual experiencemight involve a number of different objects having a conjunction ofdifferent properties: location properties, shape properties, colorproperties, and so on.

Instead of strongly Russellian contents just described, one coulddefine weaker Russellian contents that involve the very propertiesthings appear to have, but don’t involve the very objects thatappear to have them. These weaker Russellian contents are said to beexistentially quantified contents, which would be naturallyexpressed by saying something of the form ‘There is a red cubeat locationL’. Most philosophers who defend the viewthat experiences have Russellian contents hold that these contents areweaker Russellian contents. (For general discussion, see Chalmers2004; for discussion of the case of color, see Thau 2002, Shoemaker1990, 1994, 2005, Holman 2002, Maund 1995, Wright 2003.) One reasonfor weakening contents is this way is to respect the ideas that twoexperiences can represent the world as being the same way, even ifthey are experiences of perceiving numerically different objects, suchas twins. Another reason sometimes given for positing weak as opposedto strong Russellian contents is that two experiences can representthe world as being the same way, even if one is a hallucination inwhich no object is perceived at all. (see e.g., Tye 1995, chapter 5,and Tye 2000, chapter 3). Here, though, it is important to distinguishthe view that contents are individuated by the objects they comprise,from the view that an experience has a Russellian content thatincludes the object o only if the subject is perceiving the o. Thislatter view is an optional extra (built in to the simple Russellianview discussed above), not entailed by the claim that experience hasRussellian content. (Logical space has room for the view thathallucinations can sometimes have strong Russellian contents, as whenone hallucinates one’s father, or some other particular person.For discussion of this issue see McLaughlin 1989, Smith 2002, ch. 7and Johnston 2004).

Some philosophers have argued that experiences can’t havecolor-involving Russellian contents and can’t havelocation-involving contents on the grounds that this would allowphenomenally different experiences of colors and locations to have thesame Russellian contents. These criticisms illustrate the appeal tothe phenomenal adequacy constraint on contents. More exactly, theyappeal to a way of specifying that constraint by holding that when twoexperiences are phenomenally the same, or are phenomenally the same ina respect, there should be some contents that they have in common.These criticisms are discussed inSection 4.1 (color) andSection 4.2 (location).

3.2 Possible-Worlds Content

Another simple theory says that the contents of experiences are setsof possible worlds in which things are as they appear to be in theexperience. For instance, if one has a visual experience representinga red cube, the contents of experience on this view would be a set ofpossible worlds in which there really is a red cube in front of you.Or perhaps they might be a set of possible worlds in which the verycubeC that is actually in front of you is red. The centralidea behind this view is that to represent a situation—whetherin language, in thought, or in experience—is to distinguishbetween two ways the world might be: first, ways the world might be ifthe situation does obtain, where these include the actual world;second, ways the world might be if the situation does not obtain. Forinstance, if a subject’s experience represents that there is ared cube in front of her, she thereby (on this general view ofrepresentation) represents the actual world as being among the worldsin which there is a red cube in front of her. Stalnaker (1984) andLewis (1986) think this view best reflects the nature ofrepresentation.

One potential problem with possible-worlds contents concernsexperiences that represent impossible scenes. Consider the staircasedepicted in the Escher drawing, on which one seems to be able to makefour consecutive left turns, ending where one begins, yet walkingdownstairs (or upstairs) the whole time. Another example ofcontradictory contents comes from the waterfall illusion, when onelooks at a stationary object after looking at a waterfall, the objectappears to move, yet appears to stay still relative to its background.Similar illusions of movement can be achieved by looking at spirals onpaper. These experiences differ phenomenally from one another.Moreover, although all seem to represent both that something is movingand that the very same thing is not moving, they represent much morespecific scenes as well, and these specifics clearly differ. It seemsthat possible-worlds contents could not reflect these differences.Since the set of possible of worlds in which things are as theexperience represents them to be is the null set, the view that saysthat the only contents of experience are possible worlds would positthe same content (i.e., the null set) for all of them.

3.3 Fregean Contents

A third theory says that the contents of experience are composed ofmodes of presentation of objects and properties, rather thanobjects and properties themselves. The key feature of modes ofpresentation, on this view, is that a single object or property canhave multiple modes of presentation. A Fregean content will be trueonly if the object presented by a mode of presentation in the contenthas the property (or properties) presented by a mode of presentationin the content. (See Chalmers 2004).

The Fregean theory is motivated in part by the same considerationsthat motivate analogous views in the philosophy of language. The keysort of example in the philosophy of language involves a pair ofsentences, such as “Hesperus is the morning star” and“Hesperus is the evening star”, that are the same exceptfor containing different terms that refer to or denote the same thing(both “the morning star” and “the eveningstar” denote Venus). Suppose a subject rationally accepts one ofthese sentences while rationally rejecting or remaining neutral on theother. Then, it is argued, though the sentences cannot differ in theirtruth-value, they must play different roles in the subject’sreasoning: for example, the subject will list Hesperus among theheavenly bodies visible in the evening, but she won’t listHesperus among the heavenly bodies visible in the morning. Fregeantheorists take there to be an aspect of meaning that is linked tocognitive role in this way.

Fregean contents of experience are supposed to play the same role, andthus are motivated by cases with the same structure. Recall ourexample of two people with experiences that represent the same redcube at the same locationL, but from different perspectives(Schellenberg 2018). The Fregean addresses this sort of case byholding that the experiences contain different modes of presentationof the cube. An example of such a view is defended in Burge(1991).

Another example that motivates the view that experiences have Fregeancontents involves perceptual constancies. In the case of colorconstancy, a table can look such a way that one would readily judgethat it is uniformly brown, even though one can discern variations inlightness and darkness. Sometimes the variations are brought intofocus by considering how many different shades of paint one would haveto use to paint a realistic picture of the table. The central problemthese cases pose for the theory of experiential content is to sayexactly what ‘lightness and darkness’ are lightness anddarknessof: since there is a good sense in which the tablelooks brown, it is not variation in the table’s color; but thevariations seems nonetheless to be closely related to color. Sizeconstancy illustrates the same problem: a departing train does notappear to shrink as it moves away from you; but there is somethinglike size (even if it is not the size of the train) that does appearto lessen, and the problem for the theory of content is say what thisis. Accounting for the constancies is a problem for the theory ofexperience content, because both the constant features and the varyingones are reflected in phenomenology, so by the phenomenal adequacyconstraint, the contents should reflect them. The problem presented bythe constancies is to show how it is that the constant and varyingfeatures don’t conflict in such a way as to make the experiencefalsidical (for having contradictory contents).

One attempt to account for the variation in these cases is to invokeFregeanmodes of presentation of the constant features. Inthe case of color, such a Fregean proposal would say that experiencerepresents the brownness of the table under modes of presentation thatvary with the variations in lightness and darkness. The appeal of thisstrategy is that it does not treat the constant property and varyingones on a par: the constant feature is represented at the level ofreference; the varying features at the level of sense. The table doesnot look to be both brown and shiny white or grey; it looks to bebrown, by looking shinier or brighter in some parts than it does inothers.

It should be noted that the other historically influential attempts toaccount for the constancies was the sense-datum theory, which (in thecase of the table) would posit a sense-datum that really is shiny andwhite, in addition to the brown table itself. The Fregean attemptclaims to acknowledge both constant and varying features, withoutpositing such additional entities.

The main challenges for Fregean views (of utterance content, beliefcontent, and experience content) is to make precise in each case whatmodes of presentation are such that they can reflect these variations.(For criticisms of Fregean views about the contents of both beliefsand experiences, see Thau 2002, ch. 2. For defense of the Fregean viewdrawing on epistemological considerations, see Silins 2011).

3.4 Indexical Contents

Many contentful experiences seem to have contents that must bespecified by the use ofindexical expressions, such as‘over there’, ‘to the left’,‘here’, ‘in front of/behind me’, ‘just asecond ago’, ‘since a few second ago’, and so on.For instance, some auditory experiences seem to present sounds ascoming from a direction relative to the ears; some proprioceptiveexperiences seem to present pressure sensations as located in specificparts of the body; visual experiences seem to present things inlocations relative to the eyes and the rest of the body. Moregenerally, contentful experiences present the world from thesubject’s perspective both in space and in time. If one tried tocharacterize the perspective aspect of contents in language, it wouldbe natural to use indexical expressions for spatial properties, forthe subject herself, and for times.

To begin with spatial content, suppose you hear a sound as coming fromyour left. Someone else facing you, due to different orientation,hears the same sound as coming from the right. Moreover, a thirdperceiver situated halfway around the world could hear a qualitativelyidentical sound as coming fromher left. The area to thefirst perceiver’s left is the very same spot as the area to thesecond perceiver’s right, whereas the area to the left of thethird perceiver, who is very far away from the other two, is acompletely different part of space. Now, if contents are to reflectwhich side of the perceiver the soundseems to come from,then it is not enough simply to include the very area from whichsounds are coming in the content. That would not reflect what’sin common between you and the third perceiver, who is halfway aroundthe world. It would also fail to reflect what’s differentbetween you and the second perceiver, who is differently oriented thanyou. So it seems that some more specific representation of locationand direction is needed.

If locations and directions are represented in experience, then itseems they are represented relative to the subject, in which casethere has to be some way of representing the subject herself (oritself)—or at least the perceptual organs or body parts that areon the subject-side of the spatial relations. This suggests thatbodily awareness of some sort may be involved in contentfulexperiences, even when those experiences are otherwise visual orauditory, so that experiential attitudes may have to be cross-modal,rather than being individuated simply by sensory modality. Anotherquestion is how self-representation works in experience. Some forms ofself-awareness are cognitively demanding; yet representation ofperceiver-relative locations and directions are clearly involved inexperiential contents. There is a theoretical need, then, for anaccount of self-representations that could be had by cognitivelyprimitive creatures, such as animals and human infants (fordiscussion, see Peacocke 1994).

Peacocke has argued (1992 chapter 3) that to adequately reflect thespatial and first-personal perspectives in experience, a special kindof content is needed. He calls his sort of contentscenariocontent. Scenario contents are ways of filling out the spacearound the perceiver, consistent with the experience being correct.They include a point of origin and orientation on axes, in relation towhich things are located. In our example, the sound comes from theleft, relative to the origin in the subject’s body, whichPeacocke proposes is the middle of the chest.

Indexical contents (spatial, first-personal and temporal) requiremodification or further specification of all three views about thecontents of experience mentioned so far. In the case of possible-worldcontents, one way to have such an axis is to appeal to centeredpossible worlds: worlds centered on a subject, and possibly also atime and/or a location. The further features of the worlds can then bespecified in relation to the point of origin: in the example, comingfrom the left of the center (or being such-and-such distance anddirection from it). (See discussion of Peacocke’s‘positioned scenarios’ in his 1992 ch. 3 and below.)

Somewhat similarly, Russellian contents may be modified in order toreflect first-personal indexical contents by introducing contents thatare akin to open sentences, with a gap in place of a componentcorresponding to the subject. In the example involving the red cube,the contents might be something like a structured propositioncontaining the cube, the non-locational properties it appears to have(e.g., being red, being cubical, and so on), and then a componentexpressed by ‘two feet in front of ____’, where the gapcan be filled in with different subjects in different circumstances.To assess this special semantic entity for accuracy, the gap wouldneed to have a value. Whether its gap is filled or not, this entity issomething that in principle two experiences had by different subjectscould share. For discussion of these contents for experience, see Bach1997 (in the Other Internet Resources section below); for a Fregeanversion of gappy contents see Burge 1991 and Schellenberg 2010; for aRussellian version of gappy contents for belief and utterances, seeBraun 1993.

How to treat indexicality is one of the main challenges to Fregeancontents of utterances as Frege himself understood these. (See Perry1977, Evans 1990, Burge 1991.) One of the roles of modes ofpresentation according to Frege was to determine reference, so thatany two expressions (or occurrences thereof) with the same sense wouldhave the same reference. Another role is to reflect cognitivesignificance of sentences such as “Now it is 5pm, October 1st2005”. The occurrence of “now” and of “5pmOctober 1st 2005” seem to refer to the same time, yet anutterance of the sentence could be informative to someone who did notknow the date and time. One and the same time seems in some intuitivesense to have different ‘modes of presentation’: oneexpressed by ‘now’, the other expressed by ‘5pmOctober 1st 2005’.

In the case of indexical expressions, it is hard to see what featureof the meaning of ‘now’, if any, plays both of these tworoles. This poses a problem for the Fregean doctrine of senses. Latertheorists have rejected the feature of sense or modes of presentationwhereby they determine reference, while retaining the idea a singleobject or property can have multiple modes of presentation. (SeeChalmers 2002). In rejecting the idea that modes of presentationdetermine referents by themselves, they say that reference isdetermined by mode of presentation plus select facts about theenvironment in which the utterance or experience are had.

Proponents of these new Fregean modes of presentation invoke them toaccount for indexical features of experience contents. In the exampleof sounds coming from one’s left, the new Fregeans aboutexperience contents invoke a mode of presentation that is shared bythe experiences of the two perceivers in different places who hear thesound coming from their left, and a different mode of presentation forthe experience of the perceiver who hears the sound as coming from theright. The former mode of presentation (roughly the same, they wouldsay, as the mode of presentation associated with the expression‘to the left’) refers to different places for eachperceiver. The latter mode of presentation (roughly the same, theywould say, as the mode of presentation had by the expression ‘tothe right’), when it’s in the contents of the experienceof the perceiver who hears the sound coming from the right, refers tothe same place as the mode of presentation ‘to the left’when it is in the contents of the experience of the originalperceiver.

Modifications would also be needed for Russellian, possible-worlds, orFregean contents ofbeliefs; and proponents of each of theseviews have discussed what those might be in the case of belief. (Fordiscussion of indexicals and Russellian contents, see Kaplan 1977,Schiffer 1981, Austin 1990, and the entry onindexicals; for discussion of indexicals and possible-worlds content, seeStalnaker 1981; for discussion of indexical beliefs and Fregeancontents, see Perry 1977; Evans 1981, McDowell 1984.) It is an openquestion whether modifications that succeed for the case of belief canalso succeed for the case of experience. An influential approach thathas been developed for both beliefs and experiences appeals to aspecial kind of possible world: centered worlds, where these arepossible worlds marked with a designated subject and perhaps adesignated time or other parameters. The centered-world framework canprovide a level of content that can be shared by two perceivers whoperceive qualitatively identical but numerically distinct cubes indifferent locations as being nearby: each experience is true in worldswhere there is a cube nearby the designated center of that world.Features such asbeing nearby the center are called“centering features” (see Egan 2006a, 2006b, 2010), andcan be used to describe both the phenomena that indexicals are apt fordescribing, as well as perceptual constancies and secondary qualities.For discussion see Brogaard 2010 and 2018, Chalmers 2006, Egan 2006a,2006b, 2010, GReenberg 2018. Centered worlds are implicit inPeacocke’s 1992 discussion of scenario content.

3.5 Multiple Contents

If experiences have at most only one kind of content, then there willbe a substantive debate about whether experience contents areRussellian, unstructured, Fregean, or versions of these modified totake account of indexicality. Three sorts of considerations aretypically invoked to decide between these: (i) which best reflectintuitions about veridicality of experiences; (ii) which best reflectdistinctions to which experiences are sensitive; (iii) which contentsbest reflect the phenomenology of experience.

Some philosophers have proposed that there are different explanatorypurposes for contents to serve, and one kind of content cannot serveall of them. (On the general strategy of dividing contents in this waysee Fodor 1987 and White 1991.) Such philosophers think experienceshave multiple contents. One example of such a position holds thatexperiences have contents that are bothconceptual andnon-conceptual (for more on this distinction, seeSection 6.2 below. This kind of duality is explored by Reiland and Lyons 2015,and defended by by Lyons 2005a, 2005b and Bengson, Grube and Korman2011). Another holds that experiences have both ‘gappy’contents (somewhat akin to open sentences), but have object-involvingcontents as well when the experience stands in the right causalrelations to environment (e.g., Burge 1991 and Schellenberg 2010; fordiscussion, see Martin 2003). Here the idea is that the gappy contentreflects the phenomenology of the experience and the commonality incontent between indistinguishable hallucinations andnon-hallucinations, whereas the object-involving contents reflectintuitions about accuracy conditions for experiences that areadequately connected to the environment. A related idea is thatcontents of experience include both general and object-involvingcontents (Siegel 2010b).

Another example of a view positing multiple contents for multipleexplanatory purposes is Shoemaker’s theory of color experience,according to which experience represents both color properties, partlyin virtue of representing associated ‘appearanceproperties’ (seeSection 4.1 for discussion). The representation of appearance properties issupposed to explain thephenomenal sameness of certainexperiences with inverted qualia; whereas the representation of colorproperties is supposed to underlie intuitions about the veridicalityof the experiences. (This is explained more fully inSection 4.1). A somewhat similar rationale is given for a form of‘two-dimensionalism’ about experience contents, whichholds that experiences have both indexical Fregean contents, andnon-Fregean contents (either Russellian or unstructured) determined bythe Fregean contents in conjunction with select facts about theenvironment in which the experience is had (Chalmers 2004, Thompson2003). The indexical Fregean contents are supposed to reflect theinferential role and phenomenology of experience, while thenon-Fregean contents are supposed to reflect the properties that wetake experiences to attribute to objects. A related two-dimensionalistposition (Chalmers 2006) adds a third kind of Russellian content toexperience, ‘Edenic’ content, involving specialnon-instantiated ‘Edenic’ properties, which are supposedto reflect phenomenology perfectly—unlike indexical Fregeancontent, which is supposed to reflect it only imperfectly.

4. The Representation of Properties

If experiences have accuracy conditions, then it is natural to thinkof those conditions as conditions under which certainobjectshave certainproperties. For instance, experiences seem toattribute properties to ordinary objects such as people, fish,bottles, shoes and the like: the bottle feels round to haptic touch;the shoes make clicking sounds as they hit the floor; a person’sface looks to be only 3 feet away and angry; the fish smells andtastes fishy. Taking ‘object’ in a broader sense, theattribution of properties to objects seems even more widespread:auditory experience can presentsounds as loud;proprioception can presentpains as throbbing; kinestheticexperience can presentone’s toes as cold and wiggling;visual experience can presentan entire expanse as brightpink; haptic-tactile experience can presenta surface assmooth. Contents of experience, then, often seem to have a predicativestructure, involving objects (in either narrow or broad senses) andproperties. It then seems possible to theorize separately about theaccuracy conditions that concern theproperties things appearto have, on the one hand; and the accuracy conditions that concern theobjects or things that appear to have the properties. Therest of section 4 will focus on accuracy conditions associated withproperties; accuracy conditions associated with objects will bediscussed inSection 5.

When the accuracy-conditions of an experience require that a certainproperty is instantiated, the experience represents that property. Forinstance, if the experience one has when seeing a grape is accurateonly if the grape is green, then the experience represents theproperty of being green.

4.1 Which Color Properties are Represented?

The literature on how properties are represented in experience hasfocused largely on the case of visual experience and color properties.The question of how color is represented in experience is a good casewith which to review debates about what sorts of contents experienceshave. But one should generalize from the case of color experience withcaution: views that seem strongest for the case of color propertiesmay seem weaker for cases of other properties; views that seemstrongest for the case of visual experience may seem weaker for othermodalities, or for combinations of modalities.

The central issue here—which color-related properties arerepresented in experience—is closely tied up with the moregeneral question of how the phenomenology of experience relates to itshaving contents. These more general issues are discussed inSection 7; but because the general discussion to date has focused so narrowly oncolor and visual experience, the considerations overlap greatly.

It is widely held that visual experiences represent color properties,attributing them to objects, surfaces, volumes and locations. However,there is substantial debate about what experiences represent when theyrepresent colors. LetR-experiences be experiences withphenomenal characterR, whereR is the phenomenalcharacter that normal perceivers have when they see red things. LetG-experiences be experiences with phenomenal characterG, where G is the phenomenal character that normal perceivershave when they see green things.

According to a simple theory,G experiences representsgreenness, andR experiences represent redness. An objectionto this theory arises from a putative case ofinverted qualiawithout illusion. The case goes like this. One perceiver, callher Nonvert, typically hasR-experiences when she veridicallyperceives and is in causal contact with red things, and she typicallyhasG-experiences when she veridically perceives and is incausal contact with green things. Another perceiver, call him Invert,typically hasR-experiences when he veridically perceives andis in causal contact with green things, and hasG-experienceswhen he veridically perceives and is in causal contact with redthings. So, in the thought experiment, when Invert, with hisG-experiences, sees a red tomato, his experience representsredness; when Nonvert sees the same tomato with hisR-experiences, his experiences also represents redness.

On the assumption that experiences with the same color phenomenologywill have the same color content, this thought experiment challengesthe simple theory we began with, since it purports to describe a casein which a G-experience is an experience of veridically seeing red.Defenders of the simple theory may reject the assumption.Alternatively, they may challenge the part of the thought experimentthat purports to describe inverted qualiawithout illusion.That is, they may challenge the idea that Invert’sG-experiences veridically represent redness, rather thannon-veridically representing greenness. In response to this challenge,most interpretations of this thought experiment build in substantiveassumptions about what sorts of conditions are sufficient for makingit the case that experiences represent one color property rather thananother. It is standardly said that the phenomenally different colorexperiences come to represent the same property due toexternalism about color contents of experience, whereby whena type of experience stands in the right sort of causal relations to,say, redness, this suffices to make experiences of that type representredness.

If the thought experiment describes a genuine possibility, there areat least three ways to modify the simple view: (i) Deny that contentsupervenes on phenomenology, allowing that two experience with thesame color phenomenology can nonetheless differ in content. Oneversion of this response says that the phenomenal difference is a‘raw feel’ or ‘qualia’ (see Block 1990, 1996);another version holds that a single color property can present itselfto a perceiver in phenomenally different ways. (ii) Hold that thereare phenomenal differences betweenG-experiences thatrepresent redness andG-experiences that represent greenness,so that ‘G-experience’ does not pick out a fullydeterminate phenomenal type An example of this response isexternalism about phenomenology, defended by Lycan 2001 andDretske 1995. (iii) Hold that there are contents shared byG-experiences that represent redness andG-experiences that represent greenness. This response isdeveloped by Thau 2002, who offers a Russellian version of it;Chalmers 2004, who offers a Fregean version of it; and Shoemaker 1994,2005, who offers another Russellian version of it.

Responses (i) and (ii) will be discussed inSection 7. Regarding response (iii), one Fregean theory about this thoughtexperiment is that it is a case in which the experiences of Invert andNonvert have different modes of presentation of the same colorproperty (say, redness). On this view, redness would be presented inInvert’s experience under aG-mode, and it would bepresented in Nonvert’s experience under aR-mode, sothat the phenomenal difference between Invert and Nonvert’sexperience would also be a difference in Fregean content.

A Russellian theory is that in addition to color properties, Invertand Nonvert’s experiences of the red tomato also represent aproperty closely associated with redness but distinct from it, and isrepresented by and only byR-experiences. Shoemaker callsthese properties ‘appearance properties’. Call this theShoemaker proposal. (Shoemaker 1994, Thau 2002, though Thaudenies that colors are represented in experience, claiming that onlythe other properties are—so he does not accept the thoughtexperiment exactly as stated. For another denial that colors arerepresented in experiences see Tolliver 1994).

A final theory about this thought experiment combines aspects of theFregean and Russellian proposals. It says that Invert’s andNonvert’s experience each has two contents associated with it.One set of contents involves objects and properties represented by theexperiences; the other involves modes of presentation of these objectsand properties. On this proposal, the object-property contentsassociated with Invert’s experiences are the same asNonvert’s, but their modes of presentation are different.

The issue between the Fregean and Shoemaker proposals is subtle. Bothagree (contrary to response (i), which includes the qualia view) thatthe difference between Invert and Nonvert’s experience ofredness is a representational difference of some sort, as opposed to araw feel. And both views posit entities associated with thisdifference: appearance properties for the Shoemaker proposal; modes ofpresentation for the Fregean proposal. The difference concerns thestatus of the entities associated with yet distinct from colorproperties: are they modes of presentation of the colors, or are theysimply other properties, represented alongside the colors (or insteadof them, as Thau holds). In Shoemaker’s 1994 version, colorproperties are represented in experienceby representingappearance properties. The issue is sometimes described by saying thatthe Shoemaker proposal predicts that there aretwocolor-associated ways that the tomato looks to be to each perceiver:it looks to be red; and it looks to have the appearance property. Incontrast, the Fregean view predicts that there is just onecolor-associated way that the tomato looks to be: it simply looks tobe red.

A related question is which properties color experience representsthings as having when it represents them as being colored. Since colorexperiences in principle could be systematically in error, this isdistinct from the ontological question of which properties the colorsare, though the same options suggest themselves. According to oneposition, when an experience represents that an apple is red, itrepresents that it has a disposition to cause certain kinds ofexperiences. (See Langsam 2000 and McDowell 1985) According to anopposed position, when an experience represents that an apple is red,it represents that it has a non-dispositional property (McGinn 1983,Johnston 1992, Boghossian and Velleman 1992). One version of thelatter says that experience represents a primitive property thatcannot be specified in non-chromatic terms (Broackes 1992, Campbell2003). Yet another says experience is neutral on whether redness is orisn’t primitive. (Byrne and Hilbert 2003) Another question iswhether colors are represented as being properties of externalobjects, or of sense-data (where these are private mental objects), orboth.

4.2 Which Spatial Properties are Represented?

Many contentful experiences represent properties as being instantiatedat certain locations: one hears sounds as coming from certaindirections; one experiences pains in specific parts of the body; onesees different colors in different places in the scene before theeyes. It is more controversial whether gustatory and olfactoryexperiences represent locations. We certainly end up attributingtastes and odors to things in the external world, but this alone doesnot show that the associated experiences represent any locations. (Fordiscussion of olfactory experience see Lycan 2000, Smith 2002, ch 5,Batty 2010, Richardson 2013, Coveden-Taylor 2018; for discussion ofgustatory experience see Smithibid.) How are locationsrepresented in experience, when they are?

One debate about the nature of representation of space concernswhether the positional properties represented in visual experience arerelational or monadic. For discussion, see Casullo 1986.

A different debate concerns which contents are best to reflectimportant similarities in spatial experiences. Consider an experiencewhose contents include those that are expressed by an utterance of‘There is a red cube at location L’. A simple view is thatnothing more finely grained than locationL itself isincluded in these contents. Peacocke (in his 1992, chapter 3)criticizes this view. Suppose that the experience in the example isveridical, so that there really is a red cube atL, AndsupposeL is to the left of the original perceiver, but thatthere is a second perceiver who is seeing the same red cube, exceptgiven her position relative to the cube it is on her right. Theexperiences differ: the perceivers see the cube as being in differentpositions relative to their bodies, and the experiences will affectactions of the perceivers differently, in that one would have to moveto her right while the other will move to her left to get to the cube.Going with these differences, there seems to be a phenomenaldifference between the experience of seeing a red cube to your leftand seeing a red cube to your right. If the contents of experience areto reflect such differences, then there must be some more specificrepresentation of locationL than is allowed by simplyincluding the area itself in the contents. The potential problemdiscussed here for location suggests that the contents of experienceinclude indexical contents of some sort.

Peacocke has arguedscenario contents are needed toadequately reflect how locations are represented in experience. Ascenario is a spatial type, with a fixed origin in the humanbody and axes given by directions front/back, left/right, and up/down.Once the origin and axes are fixed, the scenario specifies the way thespace is filled in at each point-type identified by its distance anddirection from the origin. Apositioned scenario is ascenario that has specific origin and axes at a specific time, and isthus assessable for accuracy. Positioned scenarios are supposed toreflect the way in which locations are represented in experience.

According to another proposal about spatial experiences, made byCussins (1990), experiences of space consist in the subject’shaving abilities to move in certain ways. On this view, to experiencea sound as coming from the left is to know how to locate the place itis coming from, where this knowledge is practical knowledge about howto move through the environment. Having these abilities, according toCussins, enables the subject to form beliefs and other mental statesthat are assessable for accuracy (or that have contents in our sense);but the abilities themselves, which are supposed to constituteexperiences, are not; and even a creature who hadn’t yet orcouldn’t form such states could still have the abilities. Tohave the abilities, according to Cussins, is to be presented withlocations in experience. In his 1990 he calls this way of beingpresented with locationsconstruction-theoretic content(where this kind of ‘content’ is not assessable fortruth).

Peacocke has objected to Cussins’ construction-theoretic contentthat there appear to be cases where subjects lack the abilities tomove to the locations presented in experience, for instance due toparalysis. Whether this objection has force depends on the notion of‘abilities’; one might think the paralyzed have abilitiesthat they can’t exercise.

Peacocke’s and Cussins’ theories differ in that scenariocontents are assessable for accuracy and construction-theoreticcontents aren’t, and both purport to be contents of the sameexperiences. However, something like construction-theoretic contentscan be combined with scenario contents. One outstanding question forPeacocke’s theory is how an experience comes to have scenariocontents, and a potential answer to this is that an experience comesto have the scenario contents it does by the subject’s havingthe abilities Cussins describes. More generally, the question of whatmakes it the case that one has the experience contents one does is anoutstanding question for many theories of content (this issue isdiscussed further in theSection 8).

Another debate about the representation of spatial properties concernsthe phenomenal difference between the experiences of seeing afour-sided equilateral figure as a square and seeing it as a regulardiamond. (The example is from Mach 1914).

In one version of this contrast, there are two figures differing onlyin their orientation to the viewer (on a page, the top and bottom ofthe square is parallel with the top and bottom of the page; while thediamond is just like the square except rotated 45 degrees around itscenter.) Assuming that the property of being a square is the same asthe property of being a regular diamond (an assumption defended byPeacocke in 1993, contested by Tye 2003, but granted by Tye 2004),this appears to be a case in which there is a phenomenal differencewithout any difference in the properties represented. The options,then, are similar to the ones listed in the discussion of the invertedcolor qualia: (i) deny that there is a difference in contentcorresponding to the phenomenal difference; (ii) hold that there aredifferences in theway the property of being square isrepresented, i.e., invoke different modes of presentation of thisproperty (Peacocke 2001a, 2001b); (iii) find some other property thatis represented in one experience but not the other, such as theviewer-relative property of resting on a side, or the property havingtwo horizontal sides (properties allegedly attributed by experience tothe square but not the diamond); and the property of standing on apoint or the property of having inclined sides (properties allegedlyattributed by the experience to the diamond but not to the square)(Tye 2005). Here we have another case of a debate between Fregean andRussellian positions on the nature of experience content.

Another topic concerning spatial representation in experience is therelation between representing locations of things, on the one hand,and representing things as mind-independent, on the other. Accordingto some philosophers, for the case of visual experience, representingthings as occupying three dimensions of space suffices forrepresenting them as mind-independent (see Strawson 1958, Peacocke1983, Masrour 2011; for dissent, see Smith ch 5, and 2000). Accordingto others, visual experience represents spatial locations as locationsfrom which one could act (Schellenberg 2007).^[2]

4.3 Are High-Level Properties Represented?

Positions on which properties are represented in experience can belocated on a rough continuum, with low-level properties (where theseinclude color, shape, illumination, and depth) at one end, andhigh-level properties (where these include kind properties, agentialor other emotional properties, and semantic properties) at the other.The issue is which of these properties can be perceived as beinginstantiated: e.g., whether one can have a visual experience thatrepresents that someone istrying to do something. The debateabout this is closely related to the earlier debate in the philosophyof science about whether one can distinguish between observational andtheoretical properties (see Introduction to Suppe (1979)). In thephilosophy of mind, defenders of the view that only low-levelproperties are represented in experience include Tye 1995, Dretske1995, Clark 2000, Price 2009, Brogaard 2013, Bryne 2017; defenders ofthe view that high-level properties are represented in experienceinclude Peacocke 1992, Siewert 1998, Bayne 2009, Masrour 2011, Nanay2011, Block 2014, Siegel 2006 and 2017). It is worth noting thatanalogous debates can be pursued for views of experience that denythat experiences have accuracy conditions: for instance, if (withnaive realism of the sort championed by Martin) one thinks thatveridical experiences consist in being perceptually presented with anobject and its properties, the debate will concern which properties ofobjects one can be perceptually presented with; if one is a puresense-datum theorist, the debate will concern which propertiessense-data can have.

According to one type of argument that high-level properties can berepresented in experience, experiences can be theory-laden, in thesense that being disposed to recognize an object or propertyinfluences the phenomenology and content of the experience. The mainpremise of the argument is a claim about phenomenology: a place, suchas your neighborhood, or a person, such as a close friend, may lookdifferent to you than it did when you first saw them (these examplesare discussed by Siewert 1998); written text and spoken words in aninitially unfamiliar language may look and sound different before andafter you come to know their meaning (the auditory example isdiscussed by Peacocke and Siegel). These are cases in which backgroundknowledge or abilities putatively influences what is represented inexperience. If this can happen, then even if a distinction in thephilosophy of science between ‘observation and‘theory’ can be drawn, it will not align with therepresentation of low-level properties, on the one hand, andhigh-level properties on the other.

One type of case for denying that high-level properties arerepresented in visual experience appeals to two main ideas. First,that the only experiential representations are those that result in alawlike way from retinal stimulation. Second, retinal stimulation, inturn, is said to give rise to experiences independently of othercognitive apparatus on the part of the subject, and is insulated fromany information processed cognitively. A case of this sort is given byO’Shaughnessy (2000), chapter 17.

Some debates about which properties can be represented in perceptionfocus on specific properties. It is important to distinguish between,first,perceiving causation, on the one hand, and, second,representing causal relations in experience, on the other. It is asubstantive question whether the former can happen only if the latterdoes. The first issue concerns whether one can perceive causation, orwhether all one can perceive is a sequence of events, minus its causalnature (assuming there is such a thing). For discussion see Ducasse1965, Armstrong 1997, Fales 1990, Pietroski 2000, Beebee 2003,Butterfill 2011. For discussion in psychology of cognition regardingcausation, see Michotte (1963) and essays in Premack, Premack andSperber (1995.) The second issue concerns whether experiences canrepresent causal relations, in the same way they represent colors,shapes and whatever else is plausibly represented in experience.

Another example of this sort of debate concerns absences. As before,one can distinguish the question whether you perceive that somethingis absent, from the question whether absences can be represented inexperience. Putative examples of perceiving absences include hearingpauses, sensing the emptiness of your stomach, seeing the totaldarkness of a cave, or feeling the holes in old-fashioned computerpaper by running your finger down the sides of the page (theseexamples are from Sorensen 2004; see also Sartre 1958 and Farennikova2013 and 2015). Another putative example comes from Taylor (1952),which contains a diagram consisting of two circles, one with a dot init, the other empty. Arguably, you can see that there is no dot in theempty one; this is a putative example of seeing that a dot is absentfrom the empty circle, once one is primed to see the absence by seeingthe dotted circle next to the empty one. This interacts withdiscussions of the role of causation in seeing (for discussion of therole of causation in perception, see entry on the causation andperception). Turning to questions about whether absences arerepresented in experience, one question is whether the absence of someparticular thing, or of some kind of thing, can be represented in anykind of experience. Perhaps one could represent (or, if one isfortunate, misrepresent) that one’s limb is missing, or that noone is home. Some phenomena involving absences may be tough tocategorize. For instance, if, while seeing the empty circle, by somenon-hallucinatory process you came to seem to see a dot in the circle,would you thereby misperceive the absence of a dot as a dot? Or wouldyou merely falsidically represent that there is a dot in the circle?

A final example of debate concerning specific properties concernsaffordances. Affordances are possibilities of action of a creature(Gibson 1977). Some philosophers argue that we perceive affordances(Nanay 2011) or that affordances can be presented in visual experience(Prosser 2011, Siegel 2014).

5. How Are Objects Represented In Perception?

When properties are represented in experience, they are oftenattributed to ordinary objects: for instance, one may have a visualexperience that attributes redness and sphericality to a tomato, or ahaptic tactile experience that represents smoothness andcylindricality of a cup in one’s hands, or an olfactoryexperience attributing scents to flowers, or a gustatory experienceattributing deliciousness to something one is chewing. In order forproperties to be attributed to objects in experience, the object mustbe represented in some way. A variety of different ways have beenproposed for the case of visual experiences, so we will focus onthese. The debate is best introduced by considering an example.

Suppose you are seeing a cube that looks red and looks to be atlocationL. To simplify matters, let us ignore properties itlooks to have besides color and location, and let us ignore as wellall the other issues discussed about how properties includingsubject-relative locations are represented in experience, inparticular whether the contents include modes of representation ofproperties or properties themselves as constituents. All that mattersfor this discussion is that there be some properties attributed to theobjects.

5.1 Object-Involving, Gappy, and Existential Contents

One proposal is that when you have this experience, you entertaincontents that have the cube itself as a constituent, and attribute toit whatever properties the cube looks to have.

According to this proposal, when you represent an object inexperience, you entertain contents such as (1):

(1)o is red and at L, whereo is the very cube thatyou see.

Here, the object itself is a constituent of the content. McDowelldefends a less simple view that an object-involving Fregean sense is aconstituent of the content, rather than the object by itself. For nowit will be simplest if we focus on the simpler version.

A second proposal is a bit more complicated. On this view,hallucinations can share the same phenomenology as cases ofobject-perception, and this phenomenology suffices for the subject toexperientially entertain a special semantic entity akin to an opensentence. This might be a thought of as a structured proposition withthe same form as (1), except that in place of the perceived object,there is an unfilled position in a structure. This can be symbolizedas follows:

(2) _____ is a red cube atL.

On this view, in cases of object-perception, the properties expressedby the special semantic entity that is experientially entertainedcharacterizes the way the perceived object appears. It is thus writteninto this proposal that the gap gets a value only in cases ofperceptual contact with an object. It thus posits multiple contents toexperience. Versions of this view have been defended by Bach 1997 (inthe Other Internet Resources section below) and Burge 1991.Bach’s proposal is closer to the one stated. Burge holds that inplace of the gap, there is a demonstrative mode of presentation of anobject. See also Schellenberg 2018, Matthen 2019.

A third proposal (a version of which is defended by McGinn 1981) isthat when you see the cube, you entertainexistentially quantifiedcontents, such as (3):

(3) There is a red cube at L.

Some philosophers argue against this proposal on the grounds that itis not adequate to the phenomenology of seeing objects, wherein onealways (it is said) seems to see a particular object. The simpleexistentially quantified contents, it is said, cannot reflect thephenomenology of being visually presented with a particular object.(For this line of criticism, see Campbell 2002 ch. 6, Soteriou 2000,Martin 2003, Burge 1991.) Others go further and say that only contentsthat could naturally be expressed using a demonstrative expressionputatively referring to the seen object will be adequate to thisphenomenology. (Campbell 2002 ch. 6)

Another consideration that has been brought to bear on the debateabout how objects are represented in experience is the predictionsthat the proposals make about the accuracy (veridicality orfalsidicality) of experiences. One kind of case that brings thedifferent predictions into focus is a case discussed by Grice (1961).Suppose you see a cube (Grice’s example involves a pillarinstead of a cube). Unbeknownst to you, you are seeing the cube in amirror, so that the cube is behind you, though it looks to be atlocationL in front of you. In addition, the cube you areseeing is orange, though due to strange lighting it looks red. Nowsuppose that behind the mirror, at the exact location where the orangecube appears to be, there really is a red cube. The situation is thenthis: there are two cubes, one orange and one red; you see the orangeone but it’s not at locationL, whereas you don’tsee the red one but it is at locationL. With respect to thiscase, the view that says that objects are represented in experiencevia object-property contents such as (1) will predict that yourexperience is inaccurate (‘falsidical’), because the cubeyou see is not red and atL. In contrast, the simpleexistentially quantified contents, (3), predict that the experience isveridical, because the way things appear (on that view) is that thereis a red cube atL, and there is a red cube atL. Ifone of these verdicts is intuitively stronger than the others, thenthat counts in favor of the view(s) that predict it.

What about the gappy contents, proposal (2)? Since it is written intothe proposal that the gap gets a value when the subject of theexperience sees an object that appears to have the properties attachedto the gap, the experience in the cube case will be falsidical, sinceeven though there is a red cube atl, that cube is not theone you see. The cube you see is not red and atL, so thecontents are false, hence the experience is falsidical. Soteriou(2000) argues that this is the right verdict (intuitively, theexperience is falsidical), and so cube cases count against theproposal (3) which predicts that the experience is veridical.

A third consideration bearing on the choice between (1), (2) and (3)concerns phenomenal sameness. If two experiences that are phenomenallythe same have the same content, then proposal (1) will not work,unless it is part of a more complex proposal that experiences ofseeing objects have multiple contents: gappy ones shared byphenomenally indistinguishable experiences, and object-involving onesthat are not so shared.

5.2 Causal and Non-Causal Contents

Of the three proposals considered so far, the two that positobject-involving contents ((1) and the multiple-content proposal (2))predict that the cube experience will be falsidical. Searle (1983)defends a fourth proposal about how objects are represented inexperience that makes the same prediction, but without positingobject-involving contents. He proposes that when you see the cube, youexperientially entertain complex contents of the following form:

(4) There is a red cube atL, and the fact that there is ared cube atL is causing this experience.

Searle’s proposal predicts that the cube experience isfalsidical, because the fact that there is a red cube atL isnot causing your experience, and the second clause of hiscomplex existentially quantified contents says that it is. Accordingto Searle, this proposal also reflects the phenomenology ofparticularity.

Searle’s proposal entails that in experiences of seeing objects,the experience itself is represented (note the occurrence of‘this experience’ in (4)). Armstrong, Burge, Soteriou andothers have objected to Searle’s proposal on the grounds thatthese contents are too cognitively sophisticated to be contents ofexperiences of seeing. They argue that only sophisticated creatureshave the cognitive resources needed for this form of self-reference,and plenty of creatures can have visual experiences of seeing objectseven while lacking these resources; so such self-reference cannot be anecessary condition for visually experiencing objects. Anotherobjection, also raised by Armstrong, Burge and Soteriou, is thatcausal relations between experiences and facts are not represented inexperience. Armstrong claims that this too is overly sophisticated;Burge claims that it is not part of the phenomenology of visualexperience. The debate turns on exactly what cognitive resources areneeded for an experience to represent (a) itself, and (b) causalrelations between itself and other things. If one denies that causalrelations of any sort can be represented in experience, then one willthink Searle’s view is false.

In principle several replies to these objections are open to defendersof Searle’s position. First, they can claim that what is neededfor representations of (a) and (b) is not so cognitively sophisticatedafter all. Second, they can deny that the cognitively sophisticatedand cognitively unsophisticated creatures have experiences with thesame contents. Third, they can argue that such causal relations reallyare present in visual phenomenology, for example by contrasting visualexperiences of seeing objects with visual experiences of imaginingthem. Searle himself pursues the last strategy in his 1989.

5.3 Are Ordinary Objects Represented?

The discussion so far has assumed that ordinary objects arerepresented in at least some experiences. Some philosophers assumethat ordinary objects are represented in visual experience, but thisis a matter of some controversy. According to theories on which visualexperience represents only very low-level properties, such as color,the shapes of facing surfaces, their illumination properties andnothing more, visual experience does not carry information aboutwhether any of the facing surfaces belong to the same ordinary object,and does not carry information about whether there are any ordinaryobjects in the immediate environment at all. Instead, visualexperience represents that low-level properties are instantiated atcertain locations, without taking a stand on whether ordinary objectsare instantiating them. A version of this view is defended by AustenClark (2000, ch. 5), who argues that experience represents colors (andperhaps other low-level properties) instantiated in regions ofspace-time around the perceiver, and nothing else. In Clark’sterms, reference within experience is limited to reference to places:these are the only things to which properties are attributed inexperience. Clark refines his view in 2004.

5.4 Mind-Independence

Some perceptual experiences arguably represent objects as existingindependently of the subject’s mind (where something ismind-independent only if it does not depend for its existence on beingperceived). One might think that representing that something existsindependently of being perceived is a precondition for representingthe occurrence of more mundane happenings in the external world, suchas a tomato’s being round, or a cat’s sitting on a mat.Representing mind-independence is thus arguably a fundamental kind ofcontent. In what different ways might experiences represent things asmind-independent?

One debate in this area (already mentioned in connection with spatialrepresentation) is whether a visual experience’s representingthings as occupying three dimensions suffices to represent things asmind-independent (see Strawson 1958, Peacocke 1983, Smith 2002).

Another idea, prominent in the writings of Merleau-Ponty 1945 andHusserl 1900, is that some experiences are about things in the worldexternal to the mind in virtue of representing possible interactionbetween the subject and the thing represented. For instance, accordingto one version of this view, a visual experience’s representing(say) a bird as mind-independent consists (at least in part) in itsrepresenting that if the subject moves relative to the bird, differentparts of the bird will come into view, and that if something opaquemoves between the subject’s eyes and the bird, then the birdwill be occluded. More generally, the idea is that an object looksmind-independent if it looks as if different perspectives on it may betaken.

It is clear that subjects have some sort of expectations to thiseffect. But what concerns Merleau-Ponty and Husserl is what bearingthese expectations have on experience itself. These two philosophersagreed that experience would not be what it is without suchexpectations; but they disagreed about how exactly they areincorporated into experience. According to Merleau-Ponty, the‘expectations’ took the form of ‘readiness’ onthe part of the subject to move her body relative to an object inorder to get a better view, if she so wished. Rather than posit mentalstates that represent the possibility and possible results ofsuch interaction, Merleau-Ponty thought that such representation wasimplicit in the subject’s own dispositions. This idea has beendeveloped in the literature on so-called ‘embodiedcognition’ (see Clark 1997, Hurley 2000, Noe and O’Regan2001). Husserl, in contrast, held that the subject’sexpectations about how phenomenology would change withchanges in the perceiver’s apparent movement were presented inexperience along with other properties, such as location and shape.Similar ideas are developed in Smith 2002 and Siegel 2006b.

A third idea in the area is that tactile experiences themselves, byvirtue of their phenomenology alone, represent things as external tothe subject’s body. Consider the experience of grasping arounded bottle. In experiencing the limits of one’s body, oneseems simultaneously to experience the surfaces of things external toone. This seem to yield two ways of describing tactile phenomenology:it involves both a feeling of pressure, located roughly at the placeson the hands where the glass is touching; and a feeling of somethingspatially external to the body, located at the point where it touchesthe body. Moreover, tactile experiences of grasping a rounded bottleseems to represent the surfaces touching each part of the hand assurfaces that are part of the same object, so that it is also part ofwhat the experience represents that the surfaces extend through spacenot occupied by the body. If so, then tactile experiences representboth happenings inside the body and things outside the body.

Now, in principle something could exist in the space outside asubject’s body, yet not exist independently of thatsubject’s mind. But assuming that things outside the body arenot also internal to the mind, this would be a way of representingthings outside the mind as well. For further discussion of therelation between tactile experience and mind-independence, seeO’Shaughnessy 1989, Martin 1992b, Smith 2002, Condillac1947.

6. Concepts and Content

Many philosophers hold that beliefs andconcepts are relatedin the following way: in order to have a belief thatx isF, for any objectx and any propertyF, asubject must have concepts of the object and the property, and mustdeploy those concepts in the belief. When concepts andbeliefs are related in this way, belief content is said to beconceptual. Call this viewbelief conceptualism(endorsed by Dretske 1981 and Martin 1992a, among others). Beliefconceptualism is typically defended on the grounds that beliefs differfrom one another according to which concepts one has. One can believethat whales swim without believing that large sea-dwelling mammalsswim, even though whales are large sea-dwelling mammals. According toa common defense of belief conceptualism, the relevant beliefs differin which concepts the thinker is deploying, and that explains how itis that someone can have one of these beliefs without the other.According to a related defense of belief conceptualism, it explainswhy some thinkers cannot have any thoughts about some subject matter.Suppose the thinker lacked the concept ‘whale’: she had noidea what whales were, nor even that there was such a thing. Arguablysuch a thinker could not believe that whales swim, that whales areblue, or anything else about whales. According to beliefconceptualism, this is because she lacks the concept of a whale. Afull specification of belief conceptualism would require explicatingwhat concepts are, and what it is to possess and deploy one.

If beliefs have conceptual content, a question arises about whetherexperiences do as well. Here it is useful to distinguish between twotheses about experiences. The first concerns concepts explicitly, andis the straightforward analog of the thesis above.

Experience conceptualism: For any objectx and anypropertyF, a subject has an experience as ofxbeingF only if she has concepts ofx andF, and deploys those concepts in the experience.

The second thesis says that the contents of experiences are the samekind of content as belief contents. This thesis does not mentionconcepts explicitly.

Same-content: For any experience as of an objectxhaving a propertyF, if the experience has contentp, then it is possible to have a belief with contentp.

In principle two strands to the debate about the role of concepts inexperience contents can be distinguished: whether experiences have anycontents that cannot be believed, and whether concepts play the samerole in the contents of beliefs as they play in the contents ofexperience. Some philosophers accept the same-content thesis andbelief-conceptualism but deny experience-conceptualism, e.g., Tye(2005). Speaks (2005, 2009) defends same-content thesis but deniesexperience-conceptualism.

6.1 Conceptualist Views of Content

Experience conceptualism and the same-content thesis go naturally withthe views (discussed inSection 2.2 above) that link the content of experience to the contents ofbeliefs, either by identifying experiences with beliefs or withdispositions to form beliefs. However, McDowell (1994), Sedivy (1996)and Brewer (1999) endorse both theses while denying that experiencesand beliefs are linked in either of these ways. They defend boththeses on the grounds that experiences can provide justification forbeliefs only if these theses are true. Peacocke (2001a), Byrne (1996),Heck (2000) and others have objected to these ‘epistemic’defenses of experience conceptualism on the grounds that experiencecan stand in inferential relations with beliefs if they have accuracyconditions, and they can have accuracy conditions even if experienceconceptualism is false. (Earlier defenses of experience-conceptualismare found in Craig (1976) and Peacocke (1983)).

Dretske and Martin have argued against experience conceptualism on thegrounds that it cannot reflect theinformational richness ofexperience (see Dretske 1981, Martin 1992a, see also Bermudez andMacpherson 1999) and Chuard (2007). The main idea that experiences can(and typically do) convey information about so many objects,properties and relations in the environment that it is implausible tosuppose in those cases that the subject could possess and deployconcepts for every object, property and relation that experiencerepresents. The richness argument thus has two main parts: the firstpart defending the alleged phenomenal and informational richness ofexperience; the second defending that inference from richness to thefalsity of experiential conceptualism.

Martin in his 1992a makes a version of the richness argument againstexperiential conceptualism. His strategy is to consider cases in whichone remembers a perceived object or property that one did not noticeat the time of the experience itself. In Martin’s example,someone is looking for cufflinks in a drawer, and fails to see themeven though they are right there in front of him; but later on, whenremembering the look of the drawer’s contents, realizes that thecufflinks were among them. The fact (if it is a fact) that one canhave such memories suggests that the cufflinks were perceivedconsciously after all.

How is this sort of case supposed to counter experientialconceptualism? If deploying a concept of an object or property issufficient for noticing it, then the fact that one can perceivesomething without noticing it implies that one can perceive somethingwithout forming a concept of it. And that in turn suggests that onecan represent an object or property in experience without forming aconcept of it—contra experiential conceptualism.

Dretske develops the richness argument slightly differently. Hedistinguishes betweendigital andanalog ways ofencoding information, where the difference between these is analogousto the difference between the way statements and pictures(respectively) encode information. When information thats isF is encoded in digital form (as in a sentence, for example),no extra information is carried in the encoding; but when informationthats ifF is encoded in analog form (as in aphotograph, for example), extra information will usually be carried inthe encoding. After explicating the difference between analog anddigital encodings of information, Dretske (in his 1981) writes,“The traditional idea that knowledge, belief and thought involveconcepts while sensation (or sensory experience) does not isreflected in this coding difference”. Dretske thinks thatexperiences encode information in the analog way, whereas beliefsencode information in the digital way. This view targets both thesame-content thesis, and experiential conceptualism.

A different argument against experiential conceptualism appeals to theidea that the contents of experience are sometimes veryfinelygrained. The idea here is that experiences often providedetailed, determinate information, in contrast to thought. Considerthe case of color experience. If you have an experience as ofsomething red, you will experience it as being a determinate shade ofred. In contrast, if you think that something is red, there need notbe any determinate shade of red such that you think that it is thatshade of red. It is worth noting that an experience could befinely-grained without being informationally rich—as a visualexperience would if it represented nothing but a uniform field(sometimes called a ‘ganzfeld’) of a single color.

In his 1995, Tye attacks experience-conceptualism using the example ofdeterminate color properties, which he says can be represented inexperience even by subjects who lack concepts of them. A similar pointis made by Raffman (1995) who discusses evidence that humansperceptually discriminate more shades of color than they remember andhave concepts of. It is supposed to be independently plausible thatthe discriminated shades are represented in experience; so if thesubjects really do lack concepts of them, then experientialconceptualism is false. A similar line of thought is hinted at byEvans (1982), who asks (apparently rhetorically) “Do we reallyunderstand the proposal that we have as many colour concepts as thereare shades of colour that we can sensibly discriminate?”(229).

In response to this line of thought, McDowell (1994) has proposed thatthe subjects really do possess fine-grained color concepts, albeitdemonstrative ones. This proposal has sparked debate aboutwhat counts as possessing a demonstrative concept. One point of debateconcerns whether a subject counts as having a demonstrative concept ofa shade of color, only if she can re-identify the shade on subsequentoccasions. If so, then McDowell’s argument fails, since this isprecisely what human subjects cannot always do. It is a matter ofcontroversy, however, whether having demonstrative concepts doesrequire this (as Kelly 2001 argues), or whether there can be one-offdemonstrative concepts. Another point of debate raised byMcDowell’s concerns whether it is possible to form demonstrativeconcepts of the shade represented in experience in cases of illusion,when the shade represented in experience differs from the shade of thething seen. If demonstrative concepts of color shades can pick outonly shades actually had by the thing demonstrated (as Heck 2000contends), then again McDowell’s argument fails. However, it isagain a matter of controversy whether demonstrative concepts arelimited in this way. Yet another point of debate in this area iswhether experience itself would be needed to anchor demonstrativeconcepts in the first place — in which case, it is said, theycould not already be constituted by them (Heck defends this view).

6.2 Varieties of Non-Conceptual Content

This section reviews some proposals about the contents of experiencethat use the label ‘non-conceptual content’.

Scenario Content

One way to argue against the same-content thesis is to assume, orindependently defend, a theory of belief content, and then offer apositive account of experience content that differs from it. This isPeacocke’s strategy in his 1992. (Note the change in view fromexperiential conceptualism, which Peacocke defends in his 1983).Peacocke accepts belief conceptualism, and has the view that beliefcontents are Fregean, composed of Fregean senses. Peacocke describes akind of content calledscenario content that is not composedof Fregean senses, but is the content of experiences. Scenario contentis a set of ways of filling out the space around the perceiver,relative to an origin and axes marking directions, that is consistentwith the perceiver’s experience being veridical. Peacockestresses that although spatial concepts may be needed to specify ascenario, those concepts need not be possessed by subjects whoseexperiences have that scenario content. Thus, according to Peacocke,spatial properties (such as coming from the left of the perceiver, orbeing approximately three feet away) can be represented inexperiences, even when the subject lacks the concept ‘to theleft’ or ‘being approximately three feet away’. Ifexperiences have scenario content, and if Peacocke is right that therelevant concepts are not required for experiences to have scenariocontent, then experience conceptualism is false (barring the necessaryaccompaniment to scenario content of other contents that can be hadonly if the subject possesses and deploys concepts of the very ways offilling out the space around the perceiver that are specified byscenario content).

According to Peacocke, one role of scenario content is to make itpossible for the subject to form demonstrative concepts, such as onesthat might naturally be expressed bythat figure,thatcolor,that shape, where the figure, color or shapepicked out by the concept is one that is represented in experience.Like demonstrative expressions, demonstrative concepts pick outdifferent things on different occasions, and as Peacocke sees it therole of experience is to anchor a demonstrative concept to itsreferent. Positioned scenario content is the content of experiencethat does this.

Proto-Propositional Content

In another example involving Mach figures (discussed inSection 4.2), one focuses on a single four-sided equilateral figure, and contraststhe experience of seeing it as a square and seeing it as a diamond.The experiences supposedly differ phenomenally. According to Peacocke,there can be no difference in the scenario content of theseexperiences, because they represent the space around the perceiver asfilled in in the same way. Rather, according to Peacocke, thedifference is in theway in which the very same shapeproperty is represented in experience. The central claim in support ofthis view is that the properties represented do not differ. In his(2001a) Peacocke offers two other examples in support of this claim.First, there are supposedly two ways of hearing the interval thatsounds when middle C and F-sharp just above it are played on a piano:one and the same interval, he says, can be heard as an augmentedfourth, or as a diminished fifth. Second, a matrix of evenly spaceddots can be seen as a set of rows or a set of columns.

The multiple ‘ways’ in which the same property can berepresented, according to Peacocke, are not modes of presentation ofthe sort that constitute the contents of beliefs. In the threeputative examples of single properties that are represented indifferent ways (the four-sided figure, the middle C/F-sharp interval,and the arrangement of dots), there are (according to Peacocke)multiple concepts in the vicinity that pick out the same shape,interval, or arrangement. The concepts ‘regular diamond’and ‘regular square’, or simply two occurrences of‘that shape’, one for each way of seeing the four-sidedfigure; the concepts ‘augmented fourth’ and‘diminished fifth’ in the case of the interval; and theconcepts ‘row’ and ‘column’ in the case of thedots. These concepts differ from the perceptual ‘ways’ ofrepresenting the shape (or the interval or the dot-arrangement),according to Peacocke, in that the perceptual ‘ways’ makeit possible to acquire the concepts, and so cannot be identical withthem. In the case of the two demonstrative shape concepts, these areanchored to each shape by the experience of it, and in order to playthis role, they cannot be the same as demonstrative concepts (a pointechoed by Heck). In his (1992), Peacocke uses the label‘proto-propositional content’ for contents just described.The discussion of these examples in his (2001) is intended to clarifythe 1992 discussion (see 2001, fn 14).

The State View

According to thestate view, an experienceE hasnon-conceptual content just in case (i)E has accuracyconditions; and (ii) the subject ofE need not possess any ofthe concepts used to specifyE’s correctnessconditions. This thesis that experiences have non-conceptual contentin this sense is sometimes called thestate view ofnon-conceptual content (Heck 2000), because it is only a condition onthe possession of concepts, rather than being a condition on the kindof content had byE. Note that while the state view isincompatible with experiential conceptualism (assuming that theconcepts required by the latter are those used in a canonicalspecification), it is neutral on the same-content thesis. Consider asubject who doesn’t possess the concept RED29 (where red29 is aspecific shade of red), but whose experience represents something asbeing red29. Nothing in the state view prevents some other subjectfrom believing the contents of the first subject’s experience.Some philosophers (e.g., Crane 1992) have defended the state viewusing arguments from richness and fine-grainedness of experience(reviewed inSection 6.1). Other defenses of the state view include Stalnaker (1998a).

Another kind of content that is called ‘non-conceptualcontent’ is Cussins’construction-theoreticcontent; this is introduced inSection 4.2.

7. Content and Phenomenology

As the term ‘experiences’ is used here, it is definitionalof experiences that they have a phenomenal character. If they havecontents as well, then a question arises as to how these two featuresare related. It is useful to contrast different views on the relationbetween the content of an experience and its phenomenology byconsideringphenomenally indistinguishable experiences, orexperiences with the ‘same phenomenology’, on the onehand; and experiences with thesame content, on the other. Apreliminary remark about this notion is in order.

The notion of phenomenal indistinguishability of unrepeatableexperiences cannot be defined as a statistical notion. Exactly how itshould be defined is a matter of controversy. Gottlob Frege and MoritzSchlick raised doubts about the coherence of the notion ofinterpersonal phenomenal indistinguishability, (these‘Frege-Schlick’ doubts are discussed by Stalnaker 2000,Shoemaker 1981 and 1996) In addition, though phenomenalindistinguishability is commonly taken to be a sort ofphenomenalsameness, some philosophers use the expression topick out a merely epistemic notion, involving experiences thatsubjects cannot distinguish (for discussion, see T. Williamson (1990),Martin (2004), Siegel (2004), and entry onthe disjunctive theory of perception.) Martin and Williamson regard it as a substantive claim thatexperiences said to be ‘phenomenally the same’ haveanything more than an epistemic property of indistinguishability incommon.

There are several relations in which the content of an experience hasbeen thought to stand to its phenomenology. One position holds thatcontent suffices for phenomenology. A stronger position holds that thecontent of experience is identical with its phenomenology. A thirdposition is that the phenomenology of an experience determines itscontent. These positions and others are discussed below.

7.1 Representationalism about Phenomenology

Representationalism (orintentionalism) is thethesis that phenomenal properties are determined by representationalproperties. The thesis is sometimes stated as the supervenience claimthat phenomenal properties supervene on the representationalproperties of an experience, so that any two experiences that have thesame representational properties have the same phenomenology.

Representationalism comes in many versions. One way in which versionsdiffer is that they may be claims about all sensory experience, orjust some kinds of sensory experience. (Tye 1995 defends a version ofrepresentationalism about all sensory experience; most discussions todate have focused on visual experience, as does Pautz 2010.)

A challenge to across-the-board representationalism comes fromexperiences in modalities where it is unclear whether the experiencesrepresent happenings in the space outside the body, such as gustation,olfaction, and experiences of afterimages. However, the notion ofaccuracy conditions that defines the notion of content we’vebeen working with leaves it open whether those conditions pertain tothe world outside the mind or not. Consider whether olfactoryexperiences represent anything about the smells or sources of smellsoutside the body (for discussion see Smith (2002), chapter 5, Batty2010, 2011, Richardson 2013). There is a coherent view on which anolfactory experience can have accuracy conditions, even if they are ofthe form ‘an odor like so is present to me’. To takeanother example, consider afterimages. According to Loar (2003),experiences of afterimages represent ‘luminous happenings instrange spaces’, which suggests that an experience of anafterimage could include a content roughly like ‘there is areddish luminous occurrence over there’, where‘there’ indicates a position in the ‘strangespace’. (Perhaps the experience is neutral about whether thespace is inside or outside the mind). The notion of accuracyconditions also leaves it open whether there could be accuracyconditions that were met whenever a subject had an experience withthose conditions. For instance, suppose that sense can be made of thenotion of ‘mental space’ or ‘subjectivespace’, and that some experiences have accuracy conditionsconcerning smells, tastes or afterimages that reside in such space.Given these suppositions, one might think that whenever an experiencerepresents that something occurs in such spaces, the experience isaccurate. In opposition to this idea, Bengson (2013) argues that thevisual experiences of pink glow or seeing blackness, and the auditoryexperience of ringing in the ears fail to attribute color, luminance,or auditory properties to anything at all, and so resist analysis interms of accuracy conditions.

Another potential problem for across-the-board representationalismconcerns experiences in different sense-modalities. As stated,representationalism requires that any two experiences with the samecontent have the same phenomenology, even if the experiences occur indifferent sensory modalities. Seeing something square and touchingsomething square are phenomenally different. If representationalism istrue for both haptic-tactile and visual experiences, then theseexperiences cannot have the same content. Normally two suchexperiences would probably not have exactly the same content anyway,for instance if the visual experience represented properties (e.g.,colors) that the tactile experience didn’t. But abstracting awayfrom those differences, there is a strong intuition that the visualphenomenology correlated with visually representing squareness differsfrom the tactile phenomenology correlated with feeling something to besquare. It seems that across-the-board representationalism must either(i) deny the intuition that the visual phenomenology correlated withvisually representing squareness differs from the tactilephenomenology correlated with feeling something to be square, or else(ii) deny that the contents related to representations of squarenessreally are the same after all, so that each modality (or naturalgrouping of modalities) has a special kind of content to call its own,and one cannot experience by one modality exactly the same thing asone can experience by another. Response (ii) is illustrated by oneanswer to Molyneux’s question (see entry onMolyneux’s problem) whether someone born blind and was familiar with the feel of a cubecould recognize a cube upon seeing it for the time, without alsotouching it. If the answer to Molyneux’s question is that thepreviously blind person could not visually recognize cubes, one mightconjecture that this is because the shape properties represented byvisual experience on the one hand and by haptic tactile experience onthe other differ.

Another response to this potential problem revises the thesis ofrepresentationalism, construing the representational properties thatdetermine phenomenology as attitude-content complexes rather thancontents alone. For example, Chalmers 2004 distinguishespurerepresentational properties—the property of having acertain content—fromimpure representationalproperties—the property of having a certain contentrepresented in a certain way. Examples of impure representationalproperties include the property of having a certain contentrepresented in modalityM, and that of having attitudeA to a certain content.Impure representationalismis then the thesis that impure representational properties determinephenomenology, so that any two experiences with the same impurerepresentational properties have the same phenomenology. This positionallows that experiences in different modalities, or experiences thatinstantiate different attitudes, could have the same content, yetdiffer phenomenally because, say, one is visual and the otherkinesthetic. For instance, perhaps both represent that one’shand is moving, or that something is round. Versions of impurerepresentationalism are held by Lycan 1996, Chalmers 2004, 2006, Pautz2010.

7.2 Does Phenomenology Determine Content?

The converse of representationalism is the thesis (put roughly) thatthe phenomenology of an experience determines its content. The thesisis often more precisely formulated as a supervenience claim. Twotheses here should be distinguished:

(1) For any two phenomenally identical experiences, there are someaccuracy conditions that they share.

(2) Any two phenomenally identical experiences have the same accuracyconditions.

Claim (2) says that all accuracy conditions are determined byphenomenology, claim (1) says that at least some are.

Claim (1) constrains which contents experiences have. Shoemakerdevelops this idea in a series of papers (1994, 2005). See alsoChalmers (2004, 2005).

The two theses above are both entailed by a strong version ofrepresentationalism, which combines representationalism with thethesis that any two experiences with the same phenomenal propertieshave the same representational properties. This sort ofrepresentationalism is sometimes held in the guise of an identityclaim: the claim that phenomenal properties are identical torepresentational properties. Versions of this view are defended byDretske 1995, Tye 2000, and Carruthers 2000. But theses (1) and (2)can be held even if one rejects these strong representationalisttheses. Some philosophers hold (1) on the grounds that therepresentational content of an experience is grounded in itsphenomenology. According to Horgan and Tienson 2002, for example,there is a kind of content that experiences have in virtue of theirphenomenology, so that their phenomenology has explanatory priorityover the content itself. This view is suggested by Siewert (1998) anddefended by Kriegel (2002; 2013), and see entry onphenomenal intentionality.

Some philosophers deny both (1) and (2) on the grounds that theaccuracy-conditions of an experience are determined in part byextra-cranial features, while the phenomenology of an experience isnot. These philosophers are internalist about phenomenology, butexternalist about experience content. (Block 1990 and 1996, Speaks2009).

8. Theories of Intentionality for Experience

So far, we’ve discussed some reasons to think that there arecontentful experiences, and some questions about what sorts ofcontents there might be. A different cluster of questions concernswhat makes it the case that an experience has the contents it does.How is representation of any property is possible? Two sorts oftheories addressing this question are addressed inSection 8.2 andSection 8.3, and some possible constraints on such theories are reviewed inSection 8.1.

8.1 Externalism vs. Internalism About Perceptual Content

Externalism about experience content (orexperience-content externalism) is the thesis that subjectsalike in properties that are intrinsic to them need not share the sameexperience contents. If experience-content externalism is true, thenwhat makes it the case that an experience has the content it doesdepends on external relations to things outside the subject’sskin, such as social relations between the subject and others, orcausal relations to things in the environment. Defenders ofexperience-content externalism include Dretske, Lycan, and Tye, all ofwhom argue for experiential externalism on the grounds that thealleged best naturalistic theories of intentionality for experienceare externalist theories. (For discussion of externalism about thecontent of other mental states, see the entry onexternal theories of content.)

A thesis opposed to experience-content externalism isexperience-contentinternalism. Put roughly, this is thethesis that experience content is determined entirely by theindividual’s intrinsic properties.Total experience-contentinternalism is the thesis thatall experience content isdetermined by the individual’s intrinsic properties (i.e., thattwo subjects alike in all intrinsic respects will have exactly thesame experiential contents).Partial experience-contentinternalism is the thesis that there is asort ofexperiential content that is determined by a subject’s intrinsicproperties (i.e., two subjects alike in all intrinsic respects willhave the same experiential contents of this sort. This thesis allowsthat experiences may also have some externally determined contents.Total experiential internalism is defended by Segal 1991 and Matthews1985; partial experience-content internalism is defended by Chalmers2004.

Externalism and internalism about contents are distinct theses fromexternalism and internalism about phenomenology.Internalism aboutphenomenology is the thesis that intrinsic duplicates arephenomenal duplicates.Externalism about phenomenology is thedenial of this thesis, on the grounds that phenomenology is determinedin part by factors outside a subject’s skin. Externalism aboutphenomenology is sometimes motivated by representationalism aboutphenomenology combined with externalism about content. If one isexternalist about experiential content, and if one holds that thecontent and the phenomenology of an experience co-vary, then one mustalso be an externalist about phenomenology. This thesis is defended byDretske (1995) and Lycan (2001), and criticized by Block (1990) and(1996). For further discussion, see the entry onexternal theories of content.

One of the main motivations for experience-content internalism is theidea that only internally individuated states can figure inpsychological explanations. This mirrors one of the main motivationsfor internalism about the contents of other mental states. The generalissue is reviewed in the entry onexternalism. In the case of experience, philosophers have pursued this debate withrespect to Marr’s computational theory of vision, one of theleading empirical theories of visual processing. The debate is overwhether Marr’s theory itself posits any externally-individuatedcontents, or whether the contents it posits are all individuatedinternalistically. If Marr’s theory is correct and explanatory,and if the representational contents it posits are externallyindividuated, then it provides a counter-example toindividualists’ claims that only internally-determined contentsfigure in psychological explanation. The debate thus focuses onexactly what commitments Marr’s theory makes about theindividuation of perceptual content.

Though Marr’s theory posits representational states, it does notmake use of the notion of a visual experience, and it is an openquestion how experiences map onto the stages of computational visualprocessing posited by the theory (see Marr (1982), section 7.2 forMarr’s own brief discussion of this issue). Marr’s theoryseeks to explain how the visual system arrives at representations of athree-dimensional array given retinal stimulation as input. Theretinal stimulations, or ‘gray arrays’, are computationaldescriptions that represent light and intensity values in atwo-dimensional coordinate system. According to Marr, the visualsystem computes a description of a three-dimensional scene in threestages. First, given the gray arrays, it computes a ‘primalsketch’ of reflectance changes on the surface outside theviewer. Next, with the primal sketch as input, it computes a‘2.5-D sketch’ of this surface, a sketch that explicitlyrepresents information concerning depth and contours of the surfacesrelative to the viewer. (The depth information concerns only facingsurfaces, so that the 2.5 sketch contains no information—andtherefore no misinformation—about how surfaces continue out ofview.) Finally, given the 2.5 sketch as input, the visual systemcomputes information concerning all three dimensions of objects, onethat describes their shapes and orientations. According to the theory,the computation at each stage combines the input from the previousstage with assumptions that are supposedly built into the visualsystem.

The debate is over whether Marr’s theory allows two subjects,call themV andV*, to have different visualcontents, whenV is a normal perceiver on earth, andV* is an intrinsic duplicate ofV, with the samegray arrays as those had byV, and undergoes the samesequence of computations, viewed internally.V*’s grayarrays, however, have different distal causes. Borrowing an examplefrom Burge (1986), suppose thatV’s gray arrays aretypically caused by shadows, and occasionally caused by cracks;whereasV* lives in a world where the optical laws aredifferent, and there are no instances of the sort of shadows thatcauseV’s gray arrays. According to Burge, Marr’stheory says that the contentful visual states that are computed fromV’s gray arrays always represent shadows, and thus theones that arise from gray arrays caused by cracks are misperceptions;while the contentful visual states that are computed fromV*’s gray arrays represent cracks. NeitherVnorV* can discriminate the relevant sorts of cracks from therelevant sort of shadows, but according to Burge this is notsufficient to make their contents the same. According to GabrielSegal, in contrast, the contentful visual states computed from bothV andV*’s gray arrays represent a propertymore general than the property of being a crack or the property ofbeing a shadow—a ‘crackdow’. (For furtherdiscussion, see Matthews 1988, Burge 1988, Davies 1991, Shapiro 1993,Egan 1991, and Patterson 1996.)

8.2 Functionalist Informational Theories of Experience

Dretske and Tye have each proposed that an experience has the contentsit does in virtue of having certain functional properties.

According to Dretske, an experience has its content fixed by thebiological, phylogenetically determined function of the system (orcombination of systems: (1995), 22) of which it is a state. Thefunction of the system is to provide information: for instance, thefunction of the auditory system is to provide information about pitch,intensity, timbre, and direction of sound, and it does this byoccupying different states corresponding to different values for eachof these properties. The different states are all the differentauditory experiences that a given system is capable of being in. Thesame main idea is developed by Ruth Millikan, who focuses on mentalcontent generally rather than on experience contentperse.

Tye’s account of experience content derives from the theories ofintentionality for belief developed by Stampe (1977) and Stalnaker(1984). According to Tye, an experienceS of creaturec has the content thatp just in case: “Ifoptimal conditions were to obtain,S would be tokened increaturec if and only ifP were the case; moreover,in these circumstances,S would be tokened incbecauseP is the case” (2000, 136, note omitted; cf.1995, 101). For the case of visual experiences, Tye, like Dretske,appeals to the design of the visual system to explain what“optimal conditions” are: “In the case of evolvedcreatures, it is natural to hold that such conditions for visioninvolve the various components of the visual system operating as theywere designed to do in the sort of external environment in which theywere designed to operate” (Tye 2000, 138).

Both kinds of theories areexternalist: according to them,what makes it the case that a subject’s mental state has thecontent it does is in part facts about the subject’senvironment. Any objections to externalism about mental content arethus also objections to these theories of intentionality. Forinstance, suppose there was a creature, Swampman, who suddenly poppedinto existence due to random conglomeration of particles, and happenedto be an intrinsic molecular duplicate of, say, Jerry Fodor. Pureexternalist theories (that is, theories that hold that the only waymental states can have content is by standing in the right externalrelations to the subject’s environment) are committed to denyingthat Swampman would have any contentful mental states. Somephilosophers have found this commitment counterintuitive. For furtherdiscussion of this objection and others, see the entry onexternal theories of mental content.

A central challenge facing Dretske’s and Millikan’s theoryis to specify the relevant notions ofinformational systemandfunction in an informational system. One question in thearea is whether informational systems and functions within them arecontained within an organism, or whether they rather encompassmultiple organisms in a community. Much work by Millikan has beendevoted to this topic. According to her, an informational systemencompasses both creatures who produce representations, and creatureswho use or receive them; and the relevant notion of function issupposed to be defined in terms of both parts of such a system. Forinstance, in a bee dance, what makes it the case that the bee dancerepresents the location of nectar is in part that other bees respondto it by flying to the nectar. A analogous challenge for co-variationtheories inspired by Stampe and Stalnaker is to specify what therelevant ‘becausal’ relations are. Until these centralnotions are pinned down, the theories remain programmatic.

One objection that has been raised against both kinds of informationaltheories of experience (those that invoke biological functions, andthose that invoke co-variation) is that these theories areinsufficiently general. According to the objection, there can beproperties represented in a subject’s experience even if she hasnot been in causal contact with instances of those properties before,and which it is no part of the biological function of the visualsystem to represent. For example, consider impossible scenes depictedin Escher drawings, or a hallucination of a color that one has neverseen before (such as Hume’s missing shade of blue). Supposeexperiences can represent that one is standing before an Escherstaircase, or that something is the missing shade of blue. Accordingto Tye’s theory, this can happen only if in optimal conditionsthat type of experience is tokened because these things are the case.But since it is logically impossible to be standing before an Escherstaircase, this condition can never be met. And if it is nomologicallyimpossible to be seeing anything that is the missing shade of blue,then even if the proposed necessary condition in principle can be met,the worlds in which it is met would seem to be irrelevant ones for thepurpose of the theory. So, the objection concludes, Tye’s theorymust predict that Escher staircases and missing shades of color cannotbe represented in experience. For further discussion of objections toinformational theories, see the entry oncausal theories of mental content.

8.3 Phenomenal Intentionality

According to some philosophers (e.g., Horgan and Tienson 2002; Kriegel2002, 2013; Siewert 1998), there is a some content of experience thatis constitutively determined by the experience’s phenomenologyalone (Horgan and Tienson 2002, p. 524). Horgan and Tienson call thisthesisphenomenal intentionality (See entry onphenomenal intentionality). A consequence of the thesis is that phenomenology enjoys explanatorypriority over intentionality. Proponents of phenomenal intentionalitytake phenomenology as a primitive notion, which cannot be reduced tocontent, and then try to explain (at least some kinds of) content interms of it.

If the phenomenal intentionality thesis is true, then phenomenologycannot be reduced or identified with content, or with the having ofcontent. The thesis thus stands in contrast to views defended byDretske and Tye according to which the content of experience is whollyexplained in causal or functional terms, where one does not need toinvoke phenomenology to explain why an experience has the content ithas. These latter views leave open the possibility that the content ofan experience enjoys explanatory priority over its phenomenology, sothat one can explain phenomenology in terms of content.

Together with the assumption (made by Horgan and Tienson and byKriegel) that phenomenology is narrow (i.e., shared by intrinsicduplicates), the thesis of phenomenal intentionality yields aninternalist theory of experience content. This introducesanother respect in which phenomenal intentionality stands opposed tosome versions of functionalist informational theories of experiencecontent—namely, those versions that hold that theonlycontents experiences can have are those had in virtue of thefunctional properties of the sort discussed inSection 8.2. If the thesis of phenomenal intentionality is true and phenomenologyis narrow, then there are narrow contents that are constitutivelydetermined by phenomenology. Any objections to narrow content willthus also apply to this thesis. For further discussion of narrowcontent, see the entry onnarrow content.

9. Directions for Future Research

There is much more research in philosophy to be done on the contentsof experience. A great deal of theorizing has focused on visualexperiences. Even so, none of the questions discussed above are fullysettled, and so there is plenty of opportunity for further research onany of those topics. Additional topics ripe for future researchinclude questions such as: what is the relation between perceptualcontent and perceptual contact with reality?; how do experiences indifferent modalities interact?; what role does experiential contentplay in reasoning, in intentional action and more generally in beingan agent?; and how are perceptual constancies reflected in thecontents of experience?; what notion of contents of experience is mosttheoretically useful for unconscious perception, and how may thelatter be distinguished among the varieties of subpersonal informationprocessing?

Bibliography

• Anscombe, E., 1965. “The Intentionality of Sensation: AGrammatical Feature,” in herCollected Papers (Volume2),Metaphysics and the Philosophy of Mind, Minneapolis:University of Minnesota Press, 1981, pp. 3–20.
• Armstrong, D.M., 1961.Perception and the Physical World,London: Routledge and Kegan Paul.
• –––, 1968.A Materialist Theory ofMind, London: Routledge and Kegan Paul.
• –––, 1997.A world of states ofaffairs, Cambridge: Cambridge University Press.
• Austin, D.F., 1990.What’s the Meaning of“This”? Ithaca, NY: Cornell University Press.
• Batty, C., 2010. “Olfactory Experience I: The Content ofOlfactory Experience,”Philosophy Compass, 5(12):1137–1146.
• –––, 2011. “Smelling lessons,”Philosophical Studies, 153 (Mar): 161–174.
• Bayne, T., 2009. “Perception and the reach of phenomenalcontent,”Philosophical Quarterly, 59 (236):385–404.
• Bayne, T. and Levy, N., 2006. “The Feeling of Doing,”in W. Prinz, & N. Sebanz (eds.),Disorders of Volition,Cambridge, MA: MIT Press, 49–68.
• Beebee, H., 2003. “Seeing Causing,”Proceedings ofthe Aristotelian Society, 103: 257–280.
• Bengson, J., 2013. “Presentation and Content. ACritical Study of Susanna Siegel,The Contents of VisualExperience,”Noûs, 47(4):795–807.
• Bermudez, J. and Macpherson, F., 1999. “NonconceptualContent and the Nature of Perceptual Experience,”TheElectronic Journal of Analytic Philosophy, Issue 6. [Available online.
• Block, N., 2003. “Mental paint,”Reflections andReplies: 10 Essays With Responses From Tyler Burge, M. Hahn andB. Ramberg (eds.), Cambridge, MA: MIT Press, 165–200.
• –––, 1990. “Inverted earth,”Philosophical Perspectives, 4: 53–79.
• –––, 1996. “Mental Paint and MentalLatex,” in E. Villanueva (ed.),Philosophical Issues 7,Atascadero, CA: Ridgeview.
• Boghossian, P. and Velleman, J. D., 1989. “Color as aSecondary Quality,” reprinted in A. Byrne and D. Hilbert (eds.),Readings on Color (Volume 1), Cambridge, MA: MIT Press, 1997,81–104.
• Boring, E., 1929.A History of Experimental Psychology,New York: Appleton Century Crofts.
• Braun, D., 1993. “Empty Names,”Noûs,27: 449–69.
• Breckenridge, W., 2007. “Against one reason for thinkingthat visual experiences have representational content,”Philosophical Perspectives, 21(1): 117–123.
• Brewer, B., 1999.Perception and Reason, Oxford: OxfordUniversity Press.
• Broackes, J., 1992. “The Autonomy of Color,” in A.Byrne and D. Hilbert,Readings on Color (Volume 1),Cambridge, MA: MIT Press, 191–226.
• Brogaard, B., 2010. “Centered Worlds and the Content ofPerception,” in S. Hales (ed.),A Companion toRelativism, New York: Wiley Blackwell, pp. 137–158.
• –––, 2018.Seeing and Saying: The languageof perception and the representational view of experience, NewYork: Oxford University Press.
• Burge, T., 1986. “Individualism and Psychology,”Philosophical Review, 95: 3–45.
• –––, 1988. “Authoritative Self-knowledgeand Perceptual Individualism,” in R. Grimm and D. Merrill(eds.),Contents of Thought, Tucson: University of ArizonaPress; reprinted inFoundations of Mind: Philosophical Essays(Volume 2), Oxford: Oxford University Press, 208–220.
• –––, 1991. ‘Vision and IntentionalContent,’ in E. LePore and R. Van Gulick (eds.),John Searleand his Critics, Oxford: Blackwell, 195–214.
• Byrne, A., 1996. “Spin control: comment on JohnMcDowell’s Mind and World,” in E. Villanueva (ed.),Philosophical Issues 7, Atascadero, CA: Ridgeview.
• –––, 2001. “Intentionalismdefended,”Philosophical Review, 110: 49–90.
• –––, 2005. “Perception and ConceptualContent,”Contemporary Debates in Epistemology, E. Sosaand M. Steup (eds.), Oxford: Blackwell, pp. 231–250.
• –––, 2001. “Do Colors Look LikeDispositions? Reply to Langsam and others,”PhilosophicalQuarterly, 51: 238–45.
• –––, 2009. “Experience and Content,”Philosophical Quarterly, 59: 429–51.
• –––, 2016. “The epistemic significance ofexperience,”Philosophical Studies, 173(4):947–967.
• Byrne, A. and Hilbert, D., 2003. “Color RealismRevisited,”Behavioral and Brain Sciences, 26:791–4.
• Campbell, J., 2002.Reference and Consciousness, Oxford:Oxford University Press.
• –––, 2003. “A Simple View of Color,”in J. Haldane and C. Wright (eds.),Reality Representation andProjection, Oxford: Oxford University Press.
• Carruthers, P., 2000.Phenomenal Consciousness: A NaturalisticTheory, Cambridge: Cambridge University Press.
• Casullo, A., 1986. “The Spatial Structure of PerceptualSpace,”Philosophy and Phenomenological Research, 46(4): 665–71.
• Cavedon-Taylor, D., 2018. “Odors, Objects andOlfaction,”American Philosophical Quarterly, 55 (1):81–94.
• Chalmers, D.J., 2004. The Representational Character ofExperience, inThe Future for Philosophy, B. Leiter (ed.),Oxford: Oxford University Press, 153–181.
• –––, 2006. “Perception and the Fall fromEden,” inPerceptual Experience, T. Gendler and J.Hawthorne (eds.), Oxford: Oxford University Press, 49–125.
• –––, 2002. “On Sense and Intension,”Philosophical Perspectives 16, Atascadero: Ridgeview,135–182.
• Chisholm, R., 1957.Perceiving: A Philosophical Study,Ithaca: Cornell University Press.
• Chuard, P., 2007. “The Riches of Experience,” in R.Genarro (ed.),The Interplay between Consciousness andConcepts, Special Issue,Journal of ConsciousnessStudies, 14 (9–10): 20–42.
• Clark, Andy, 1997.Being There, Cambridge, MA: MITPress.
• Clark, Austen, 1993.Sensory Qualities, Oxford:Clarendon.
• –––, 2000.A Theory of Sentience,Oxford: Oxford University Press.
• –––, 2004. “Sensing, Objects, andAwareness: Reply to Commentators,”PhilosophicalPsychology, 17 (4): 553–579.
• Condillac, E., 1947.Traite des senations, OeuvresPhilosophiques de Condillac (Volume 1), Paris: PressesUniversitaires de France, pp. 219–314.
• Craig, E., 1976. “Sensory Experience and the Foundations ofKnowledge,”Synthese, 33 (8): 1–24.
• Crane, T., 2003. “The Intentional Structure ofConsciousness,” in Q. Smith and A. Jokic (eds.),Consciousness: New Philosophical Essays, Oxford: OxfordUniversity Press.
• –––, 1998, “The Waterfall Illusion,”in Y. Gunther (ed.),Essays on Non-conceptual content,Cambridge, MA: MIT Press, 231–36
• ––– (ed.), 1992.The Contents ofExperience, Cambridge: Cambridge University Press.
• –––, 2001.Elements of Mind, Oxford:Oxford University Press.
• –––, 1992. “The Non-conceptual Content ofExperience,” in T. Crane (ed.),The Contents ofExperience, Cambridge: Cambridge University Press,136–57.
• Cussins, A., 1990. “Content, Conceptual Content, andNonconceptual Content,” inEssays on NonconceptualContent, Y. Gunther (ed.), Cambridge, MA: MIT Press, pp.133–163.
• Dainton, B., 2000.Stream of Consciousness, London:Routledge and Kegan Paul.
• Davies, M., 1996. “Externalism and experience,”Philosophy and Cognitive Science: Categories, Consciousness, andReasoning, A. Clark, J. Ezquerro, and J. M. Larrazabal (eds.),Dordrecht: Kluwer. Reference to the reprinting inThe Nature ofConsciousness, N. Block, O. Flanagan, and G. Güzeldere(eds.), Cambridge, MA: MIT Press, 1997, 309–28.
• –––, 1991. “Individualism and PerceptualContent,”Mind, 100: 485–94.
• –––, 1992. “Perceptual Content and LocalSupervenience,”Proceedings of the AristotelianSociety, 92: 21–45.
• Dennett, D., 1991.Consciousness Explained, Boston:Little, Brown and Company.
• Dretske, F., 1981.Knowledge and the Flow of Information,Cambridge, MA: MIT Press.
• –––, 1995.Naturalizing the Mind,Cambridge, MA: MIT Press.
• –––, 1993. “Conscious Experience,”Mind, 102 (406): 263–283.
• Ducasse, C. J., 1942. “Moore’s ‘The Refutationof Idealism,’” in P. Schlipp (ed.),The Philosophy ofG.E. Moore, Evanston, IL: Northwestern Press.
• –––, 1965. “Causation: Perceived or OnlyInferred?”Philosophy and Phenomenological Research,26: 173–79.
• Egan, A., 2006a. “Appearance Properties?”Noûs, 40 (3): 495–521.
• –––, 2006b. “Secondary Qualities andSelf-Location,”Philosophy and PhenomenologicalResearch, 72 (1): 97–119.
• –––, 2010. “Projectivism withoutError,” in B. Nanay (ed.),Perceiving the World, NewYork: Oxford University Press, 68–96.
• Egan, F., 1991. “Must Psychology be individualistic?”Philosophical Review, 100: 179–203.
• Evans, G., 1982.The Varieties of Reference, Oxford:Oxford University Press.
• –––, 1981. “UnderstandingDemonstratives,” in H. Parret and J. Bouveresse (eds.),Meaning and Understanding, Berlin: W. de Gruyter,280–303.
• Fales, E., 1990.Causation and Universals, London:Routledge.
• Francescotti, R., 1991. “Externalism and Marr’s Theoryof Vision,”The British Journal for the Philosophy ofScience, 42: 227–38.
• Fodor, J., 1987.Psychosemantics, Cambridge, MA: MITPress.
• Fotopoulou, A.et al., 2009. “Self-observationreinstates motor awareness in anosognosia for hemiplegia,”Neuropsychologia, 47: 1256–1260.
• Gibson, J.J., 1977. “The Theory of Affordances,” in R.Shaw and J. Bransford (eds.),Perceiving, Acting, and Knowing.Toward an Ecological Psychology, Hillsdale: NJ, Lawrence ErlbaumAssociates, 67–82; revised and republished as Chapter 8 of J.J.Gibson,The Ecological Approach to Visual Perception, Boston:Houghton Miflin, 1979.
• Gluer, K., 2009. “In Defence of a Doxastic Account ofExperience,”Mind and Language, 24: 297–373.
• Greenberg, G., 2018. “Content and Target in PictorialRepresentation,”Ergo, 5(33).doi:10.3998/ergo.12405314.0005.033
• Grice, H.P., 1961. “The Causal Theory of Perception,”Proceedings of the Aristotelian Society (SupplementaryVolume), 35: 121–53.
• Heck, R. G., 2000. “Nonconceptual content and the‘space of reasons’,”Philosophical Review,109: 483–523.
• Holman, E., 2002. “Color eliminativism and colorexperience,”Pacific Philosophical Quarterly, 83:38–56.
• Horgan, T. and J. Tienson, 2002. “The intentionality ofphenomenology and the phenomenology of intentionality,” in D.Chalmers (ed.),Philosophy of Mind: Classical and ContemporaryReadings, Oxford: Oxford University Press, 520–33.
• Hurley, Susan, 2000. “Consciousness in Action”European Journal of Philosophy, 8 (1): 106–110.
• Husserl, Edmund, 1900–01.Logical Investigations,J.N. Findlay (trans.), revised edition by D. Moran, London: Routledge2001.
• Jackson, F., 1977.Perception: A Representative Theory,Cambridge: Cambridge University Press.
• Johnston, M., 2004. “The Obscure Object ofHallucination,”Philosophical Studies, 103:113–83.
• –––, 1992. “How to Speak of theColors,”Philosophical Studies, 68: 221–63.
• –––, 2006. “Better than mere knowledge?The function of sensory awareness,” in T.S. Gendler & JohnHawthorne (eds.),Perceptual Experience, Oxford: OxfordUniversity Press, 260–290.
• –––, 2014. “The Problem with the ContentView,” in B. Brogaard (ed.),Does Perception HaveContent?, Oxford: Oxford University Press.
• Kaplan, D., 1989. “Demonstratives,” and“Afterthoughts,” inThemes from Kaplan, J. Almog,J. Perry, and H. Wettstein (eds.), New York: Oxford UniversityPress.
• Kelly, S. D., 2001. “The non-conceptual content ofperceptual experience: situation dependence and fineness ofgrain,”Philosophy and Phenomenological Research, 62:601–8.
• Kennett, S., Taylor-Clarke, M., and Haggard, P., 2001.“Noninformative vision improves the spatial resolution of touchin humans,”Current Biology, 11: 1188–1191.
• Kriegel, U., 2002. “Phenomenal content,”Erkenntnis, 57: 175–198.
• –––, 2013. “The Phenomenal IntentionalityResearch Program,” in U. Kriegel (ed.),PhenomenalIntentionality, New York: Oxford University Press, pp.1–26.
• Langsam, H., 2000. “Why colors do look likedispositions,”Philosophical Quarterly, 50:68–75.
• Lewis, D.K., 1986.On the Plurality of Worlds, Oxford:Blackwell.
• Locatelli, R. and Wilson, K., forthcoming. “Perceptionwithout Representation,”Topoi: An International Review ofPhilosophy (special issue).
• Loar, B., 2002. “Transparent Experience and the Availabilityof Qualia,” in Q. Smith and A. Jokic (eds.),Consciousness:New Philosophical Perspectives, Oxford: Oxford University Press,pp. 77–96.
• Lycan, W., 2000. “The Slighting of Smell (with a brief wordon the slighting of chemistry),” in N. Bhushan and S. Rosenfeld(eds.),Of Minds and Molecules, Oxford: Oxford UniversityPress, pp. 273–89.
• –––, 2001. “The Case for PhenomenalExternalism,”Philosophical Perspectives, 15:17–35.
• –––, 1996.Consciousness andExperience, Cambridge, MA: MIT Press.
• Lyons, J., 2005a. “Clades, Capgras, and PerceptualKinds,”Philosophical Topics, 33(1):185–206.
• –––, 2005b. “Perceptual belief andnonexperiential looks,”Philosophical Perspectives, 19(1): 237–256.
• Madary, M., 2012. “How would the world look if it as it wereencoded as an intertwined set of probability densitydistributions?”Frontiers in Psychology, 3:419.doi:10.3389/fpsyg.2012.00419
• Marr, D., 1982.Vision, San Francisco: FreemanPress.
• Martin, M.G.F., 1992a. “Perception, Concepts andMemory,” reprinted in Y. Gunther (ed.),Essays inNon-Conceptual Content, Cambridge, MA: MIT Press,237–250.
• –––, 1992b. “Sight and Touch,” inThe Contents of Experience, T. Crane (ed.), Cambridge:Cambridge University Press, pp. 196–215.
• –––, 2002. “The Transparency ofExperience,”Mind & Language, 17 (4):376–425.
• –––, 2003. “Particular Thoughts andSingular Thought,” in A. O’Hear (ed.),Logic, Thoughtand Language, Cambridge: Cambridge University Press.
• –––, 2004. “The Limits ofSelf-Awareness,”Philosophical Studies, 120:37–89.
• –––, 2006. “On Being Alienated,” inIn Perceptual Experience, T. Gendler Szabo and J. Hawthorne(eds.), Oxford: Oxford University Press, 354–410.
• Masrour, F., 2011. “Is Perceptual Phenomenology Thin?”Philosophy and Phenomenological Research, 83 (2):366–397
• –––, 2015. “The geometry of visual spaceand the nature of visual experience,”PhilosophicalStudies, 172 (7): 1813–1832.
• –––, forthcoming. “Space Perception,Visual Dissonance and the Fate of Standard Representationalism,”Noús, 50 (1); online February 2016, doi:10.1111/nous.12139
• Matthen, M., 2019. “Objects, seeing, andobject-seeing,”Synthese, 198: 3265–3288.
• Matthews, R., 1985, “Burge on PerceptualIndividualism,” in P. Grimm and D. Merrill (eds.),Contentsof Thought, Tucson: University of Arizona Press.
• Maund, J.B., 1995.Colours: Their Nature andRepresentation, Cambridge: Cambridge University Press.
• McDowell, J., 1994a.Mind and World, Cambridge, MA:Harvard University Press.
• –––, 1984. “Values and SecondaryQualities,” inMorality and Objectivity, T. Honderichand Paul Kegan (eds.), London: Routledge, 110–129.
• –––, 1998. “Reply to commentators,”Philosophy and Phenomenological Research, 58:403–31.
• –––, 1994b. “The Content of PerceptualExperience,”Philosophical Quarterly, 44:190–205.
• –––, 1982. “Criteria, Defeasibility andKnowledge,”Proceedings of the British Academy, 68:455–79.
• –––, 1995. “Knowledge and theInternal,”Philosophy and Phenomenological Research, 55(5): 877–93.
• –––, 1984. “De Re Senses,” in C.Wright (ed.),Frege: Tradition and Influence, Oxford:Blackwell Books.
• McGinn, C., 1989.Mental Content, Oxford: BasilBlackwell.
• –––, 1983.The Subjective View, Oxford:Oxford University Press.
• –––, 1996. “Another Look at Color,”Journal of Philosophy, 93: 537–53.
• McLaughlin, B., 1989. “Why Perception is not SingularReference,” inCause, Mind and Reality, J. Heil (ed.),Dordrecht: Kluwer, 111–20.
• Merleau-Ponty, M., 1945.The Phenomenology of Perception,Colin Smith (trans.), London: Rooutledge 2002.
• Michotte, A., 1963.The Perception of Causality, NewYork: Basic Books.
• Morrison, J., 2015. “Perceptual Confidence,”Analytic Philosophy, 57 (1): 15–48.
• Munton, J., 2016. “Visual confidence and perceptualjustification,”Philosophical Topics, 44 (2):301–326.
• Nanay, B., 2011. “Do we see apples as edible?”Pacific Philosophical Quarterly, 92 (3): 305–322.
• Noe, A. and O’Regan, K., 2001. “What it is like tosee: a sensorimotor theory of perceptual experience,”Synthese, 129 (1): 79–103.
• Raleigh, J and Vindrola, F., 2020. “Perceptual experienceand degrees of belief,”Philosophical Quarterly, 2:378–406.
• O’Shaughnessy, B., 1989. “The Sense of Touch,”Australasian Journal of Philosophy, 67: 37–58.
• –––, 2000.Consciousness and the World,Oxford: Clarendon.
• Patterson, S., 1996. “Success-orientation and individualismin Marr’s theory of vision,” E. Akins (ed.),Perception, Oxford: Oxford University Press.
• Pautz, A., 2010. “Why explain visual experience in terms ofcontent?” inPerceiving the World, B. Nanay (ed.), NewYork: Oxford University Press, 254–309.
• Peacocke, C., 1983.Sense and Content, Oxford: OxfordUniversity Press.
• –––, 1992.A Study of Concepts,Cambridge, MA: MIT Press.
• –––, 1998. “Nonconceptual contentdefended,”Philosophy and Phenomenological Research,58: 381–8.
• –––, 2001a. “Does perception have anonconceptual content?,”Journal of Philosophy, 98:239–64.
• –––, 2001b. “Phenomenology andNonconceptual content,”Philosophy and PhenomenologicalResearch, 62: 609–15.
• Pendelbury, M., 1987. “Perceptual Representation,”Proceedings of the Aristotleian Society, 87:91–106
• Perry, J., 1977. “Frege on Demonstratives,”Philosophical Review, 86 (4): 474–97
• Pietroski, P., 2000.Causing Actions, New York:OxfordUniversity Press.
• Pitcher, G., 1971.Perception, Princeton: PrincetonUniversity Press.
• Premack, A., Premack, D., and Sperber, D. (eds.), 1995.CausalCognition, Oxford: Clarendon.
• Press, C., Taylor-Clarke, M., Kennett, S., and Haggard, P., 2004.“Visual enhancement of touch in spatial bodyrepresentation,”Experimental Brain Research, 154:238–45.
• Prosser, S., 2011. “Affordances and Phenomenal Character inSpatial Perception,”Philosophical Review, 120 (4):475–513.
• Raffman, D., 1995. “On the Persistence ofPhenomenology,” in T. Metzinger (ed.),ConsciousExperience, Ferdinand-Schoningh: Paderborn, 293-302.
• Reiland, I. and Lyons, J., 2015. “The Disunity ofPerception: An Introduction,”Pacific PhilosophicalQuarterly, 96 (4): 443–445.
• Richardson, L., 2013. “Sniffing and smelling,”Philosophical Studies, 162 (2): 401–419.
• Sartre, J.P., 1958.Being and Nothingness, H. E. Barnes(trans.), New York: Washington Square Press.
• Schellenberg, S., 2010. “The Particularity and Phenomenologyof Visual Perception,”Philosophical Studies, 149 (1):19–48.
• –––, 2011. “Perceptual ContentDefended,”Noús, 45 (4): 714–750.
• –––, 2018.The Unity of Perception,Oxford: Oxford University Press.
• Schiffer, S., 1978. “The Basis of Reference,”Erkenntnis, 13: 171–206.
• Searle, J., 1983.Intentionality: An Essay in the Philosophyof Mind, Cambridge: Cambridge University Press.
• –––, 1989. “Replies,” in E. LePoreand R. Van Gulick (eds.),John Searle and His Critics,Oxford: Blackwell, 181–192.
• Sedivy, S., 1996. “Must Conceptually-informed ExperienceInvolve Non-conceptual Content?”Canadian Journal ofPhilosophy, 26: 413–31.
• Segal, G., 1991. “Defense of a ReasonableIndividualism,”Mind, 100 (4): 485–494.
• –––, 1989. “Seeing What is the notThere,”Philosophical Review, 98: 189–214
• Sellars, W., 1975. “The Adverbial Theory of the Objects ofPerception,”Metaphilosophy, 6: 144–60.
• Shapiro, L., 1993. “Contents, Kinds and Individuation inMarr’s theory of vision,”Philosophical Review,102 (4): 489–513.
• Shoemaker, S., 1994. “Self-knowledge and ‘innersense’,” (Lecture III: the phenomenal character ofexperience),Philosophy and Phenomenological Research, 54:219–314.
• –––, 1990. ‘Qualities and Qualia:What’s in the Mind?’Philosophy and PhenomenologicalResearch, 50 (Supplement): 109–31.
• –––, 1981. “The Inverted Spectrum,”Journal of Philosophy, 74 (7): 357–81.
• –––, 1996.“Intersubjective/Intersubjective,” inThe First-PersonPerspective and Other Essays, Cambridge: Cambridge UniversityPress, pp. 141–54.
• –––, 2006. “On the Ways ThingsAppear,” inPerceptual Experience, T. Gendler andJ.Hawthorne (eds.), Oxford: Oxford University Press,461–480.
• Siegel, S., 2006a. “Which Properties are Represented inPerception?” inIn Perceptual Experience, T. GendlerSzabo and J. Hawthorne (eds.), Oxford: Oxford University Press,481–503.
• –––, 2006b. “Subject and Object in theContents of Visual Experience,”Philosophical Review,115 (3): 355–88.
• –––, 2004. “Indiscriminability and thePhenomenal,”Philosophical Studies, 120:90–112.
• –––, 2010a. “Do Experiences HaveContents?” in B. Nanay (ed.),Perceiving the World, NewYork: Oxford University Press, 333–68.
• –––, 2010b.The Contents of VisualExperience, New York: Oxford University Press.
• –––, 2014. “Affordances and the Contentsof Perception,” in Berit Brogaard (ed.),Does PerceptionHave Content?, Oxford: Oxford University Press, 39–76.
• Siewert, C., 1998.The Significance of Consciousness,Princeton: Princeton University Press.
• Silins, N., 2011, “Seeing Through the Veil ofPerception,”Mind, 120(478): 329–367.
• Smith, A.D., 2002.The Problem of Perception, Cambridge,MA: Harvard University Press.
• –––, 2000. “Space and Sight,”Mind, 109 (435): 481–518.
• Sorensen, R., 2004. “We see in the dark,”Noûs, 38 (3): 456–480.
• Soteriou, M., 2000. “The Particularity of VisualPerception,”European Journal of Philosophy, 8 (2):173–89.
• Speaks, J., 2005. “Is there a problem about nonconceptualcontent?”Philosophical Review, 114 (3):359–98.
• –––, 2009. “Transparency, intentionalism,and the nature of perceptual content,”Philosophy andPhenomenological Research, 79 (3): 539–573.
• Spener, M., 2003.Gilding or Staining the Mind: Phenomenologyand the Metaphysics of Visual Experience, Ph.D. thesis,Philosophy Department, University of London.
• Stalnaker, R., 1984.Inquiry, Cambridge, MA: MITPress.
• –––, 1998a. “What Might NonconceptualContent Be?” in E. Villanueva (ed.),PhilosophicalIssues 9, Atascadero, CA: Ridgeview, 339–52.
• –––, 1981. “Indexical Belief”reprinted inContext and Content, Cambridge, MA: MIT Press, 1999,pp. 130–49.
• –––, 2000. “Comparing Qualia AcrossPersons.”Philosophical Topics, 26: 385–405.
• Stampe, D., 1977. “Towards a causal theory of linguisticrepresentation,”Midwest Studies in Philosophy, 2:42–63.
• Stokes, D., M. Matthen and S. Biggs, 2014.Perception and itsModalities, Oxford: Oxford University Press.
• Strawson, P. F., 1985. “Causation and Explanation” inEssays on Davidson, B. Vermazen (ed.), Oxford: Clarendon, pp.115- 135.
• –––, 1958.Individuals, London:Methuen.
• Suppe, F. (ed.), 1977.The Structure of ScientificTheories, Urbana: University of Illinois Press, 2nd edition.
• Szabó, Z., 2003. “Believing in Things,”Philosophy and Phenomenological Research, 66:584–611.
• Taylor, R., 1952. “Negative Things,”Journal ofPhilosophy, 19 (13): 433–449.
• Travis, C., 2004. “The Silence of the Senses,”Mind, 113 (449): 57–94.
• Thau, M., 2002.Consciousness and Cognition, Oxford:Oxford University Press.
• Thompson, B., 2003.The Nature of Phenomenal Content,Ph.D. Dissertation, Philosophy Department, University of Arizona.
• Tolliver. J.T., 1994. “Interior Colors,”Philosophical Topics, 22: 411–14.
• Travis, C., 2013. “Susanna Siegel, The contents of visualexperience,”Philosophical Studies, 163(3):837–846.
• Tye, M., 1995.Ten Problems of Consciousness, Cambridge,MA: MIT Press.
• –––, 2000.Consciousness, Color, andContent, Cambridge, MA: MIT Press.
• –––, 1984. “The Adverbial Approach ofVisual Experience,”Philosophical Review, 93:195–226.
• –––, 2005. “Non-conceptual content,Richness, and Fineness of Grain,” in T. Gendler and J. Hawthorne(eds.),Perceptual Experience, Oxford: Oxford UniversityPress, 504–26.
• White, S., 2004. “Skepticism, Deflationism, and theRediscovery of the Self,”The Monist, 87:275–98.
• –––, 1991.The Unity of the Self,Cambridge, MA: MIT Press
• –––, 1991. “Partial Character and theLanguage of Thought,” inThe Unity of the Self (Chapter1), Cambridge, MA: MIT Press.
• Wilson, K. and F. Macpherson, 2017. “Introduction:Perception without Representation,”Topoi, 36 (2):197–212.
• –––, 2018. “The Senses,”OxfordBibliographies in Philosophy.doi:10.1093/OBO/9780195396577-0368
• Williamson, T., 1990.Identity and Discrimination,Oxford: Blackwell.
• Wright, W., 2003. “Projectivist representationalism andcolor,”Philosophical Psychology, 16(4):515–29

Academic Tools

	How to cite this entry.
	Preview the PDF version of this entry at theFriends of the SEP Society.
	Look up topics and thinkers related to this entry at the Internet Philosophy Ontology Project (InPhO).
	Enhanced bibliography for this entryatPhilPapers, with links to its database.

Other Internet Resources

• Bach, Kent, 1997, “Searle Against the World”, manuscript.
• Horgan, T., and Tienson, J., 2005. “The Phenomenology ofEmbodied Agency,” in J. Joao Sàágua (ed.),Mind and Action 3, [Preprint available online]
• Sandlot Science.com, an interesting collection of visual illusions.
• Contemporary Philosophy of Mind: An Annotated Bibliography, compiled by David Chalmers (University of Arizona)
• The Contents of Perception, in Philpapers.org.
• Wilson, K. and F. Macpherson, 2018. The Senses, OxfordBibliographies in Philosophy. DOI: 10.1093/OBO/9780195396577-0368

Related Entries

consciousness |consciousness: representational theories of |consciousness: temporal |externalism about the mind |indexicals |mental content: causal theories of |mental content: narrow |Molyneux’s problem |perception: auditory |perception: experience and justification |perception: the disjunctive theory of |perception: the problem of |phenomenology |propositions: structured |qualia: inverted