The Case for Phenomenal Externalism

William G. Lycan

University of North Carolina


Since Twin Earth was discovered by American philosophical-space explorers in the 1970s, the domain of “wide” or externalist or extrinsic properties has expanded.Putnam (1975) originally focused just on the meanings of people’s linguistic utterances, arguing that two perfect twins, an Earthling and that person’s molecular duplicate on Twin Earth in parallel physical surroundings, can utter identical sentences that nonetheless have different meanings.Fodor (1980) and Stich (1978) extended this argument to cover the contents of people’s propositional attitudes, so far as those contents comprised concepts of certain sorts, such as natural-kind concepts.Stich (1980) then argued that Fodor had underestimated the range of the wide (“…what Fodor sees as a bit I see as the tip of an iceberg” (p. 97)).

That far, the discussion had confined itself to standard cognitive and conative propositional attitudes.Block (1990, 1994, 1996) widened the wide further by granting a sense in which perceptual contents are wide—in which, for example, a same-colored object might look red to me but blue to my molecular twin.Yet Block insisted that there is another and perhaps more important sense in which the perceptual contents are “narrow,” i.e., in which the object must look the same color to any molecular duplicate of mine.He calls these narrow perceptual contents “qualia,” in a special neologistic sense of that fraught word.

I do not believe in “qualia” in Block’s special sense,[1] though I insist on the existence of qualia in a more customary and familiar sense.My purpose in this paper is to argue that qualia in the latter sense are wide.I made that claim and defended it a bit in Lycan (1996), but since then a number of arguments have been given for the contrary thesis that qualia are narrow, so it is time to rejoin the debate.


1.The Representational theory of qualia.[2]



Qualia in my sense are the introspectible qualitative phenomenal features that characteristically inhere in sensory experiences.The color of an after-image would be a good example, as is the color occupying such-and-such a region of your ordinary visual field right now; other examples are the pitch or the volume of a heard sound, and the smell of a smell.I emphasize that these are phenomenal properties.They may be part of normal, veridicasensory experience; or they may be hallucinatory or otherwise purely subjective, and correspond to nothing in the subject’s actual environment.

In previous works I have called qualia “introspectible monadic properties of apparent phenomenal individuals,” meaning the characteristic properties of what Russell thought were actual phenomenal individuals, i.e., of what he called sense-data; Russell took qualia in my sense to be first-order properties of sense-data.But my reference to “apparent phenomenal individuals” has sometimes confused readers, since it may be hard to think about “characteristic properties of” sense-data without falling in with the assumption that there really are sense-data.Yet it is crucial to see that one need not believe in sense-data to believe in qualia.Most of us reject Russell’s idea that an after-image is an actual and nonphysical individual thing, but if you are experiencing a green after-image, you cannot deny that in some sense there is greenness in your visual field; greenness is somehow visually exemplified and present to you.If the greenness is (indeed) not a first-order property of an immaterial sense-datum, then of what is it a property, and/or, what kind of property is it?We must relocate it ontologically.And I maintain (Lycan (1987a, 1987b, 1996)) that this is a very difficult metaphysical problem.

To see the difficulty, suppose that while you are experiencing the after-image, there is in reality no green physical object in your environment.[3]Suppose also that there is no green physical object in your brain (if there were one, you would be in big trouble).But there is a green something before you, Russellian sense-datum or not.If there is no green physical object outside your head and no green physical object inside your head, it follows that the green something is a nonphysical object.So much for materialism; it seems that the green something might as well be a sense-datum after all.

Enter the Representational theory, to carry us between the horns of the foregoing dilemma:Notice that each disjunct assumes that the green “something” is an actual thing (else we could not derive the unwanted conclusion that there actually exists a nonphysical green thing).The Representational theory affords a third alternative, by supposing that qualia are intentional contents of sensory states, properties of intentional objects, represented properties of representata.Of course it is characteristic of intentional contents that they may or may not actually exist; and that is how we evade the dilemma.Your visual system quite often portrays, alleges something green.But, vision being not entirely reliable, on a given occasion the green thing may or may not actually exist.

Suppose Ludwig is seeing a real tomato in good light, and naturally it looks red to him.He is visually representing the actual redness of the tomato, and veridically so.But suppose Bertie is experiencing a green after-image as a result of seeing a red flash bulb go off.According to the Representational theory, that is for Bertie to be visually-representing a filmy green blob located at such-and-such a spot in the room. The representation has a truth-condition involving greenness.

There visually appears to Ludwig to be a red tomato; there visually appears to Bertie to be a green blob.The tomato is real and so is its redness, but the blob is unreal, an intentional inexistent, and so is its greenness.The greenness is the color of an illusory, nonexistent thing.If that seems weird to you, think of hallucinating pink rats.Perhaps you know the rats are not real, but they are unquestionably pink.The pinkness is the color of the nonexistent rats.(I take that to be uncontroversial.[4])Construing intentionality as representation, we can say that the pinkness is the represented color of the represented but nonactual rats, i.e., it is the color they are represented as having.[5]

Nor should it be surprising that Bertie’s blob is a nonactual, nonexistent thing.Since in reality there is no green blob in the room with Bertie, his visual experience is unveridical.Remember that after-images are illusions; it looks to Bertie as though there is a green blob before him, when there is not anything green before him.Moreover, vision science has an extensively worked out explanation of this particular illusion.It is well understood why it can look to someone as though there is something green even when there is not anything green before that person.

I contend that Russell mistook a nonactual physical individual for an actual nonphysical one.


There are a number of arguments in favor of the Representational theory, but I will (briefly) state only the three most accessible ones.

First:The theory is the only very promising way to preserve materialism while accommodating qualia.For the only viable alternative exegesis of Bertie seems to be belief in actual Russellian sense-data or at least in immaterial properties.(Jackson (1977), Butchvarov (1980) and Lycan (1987a, 1987b) argue at length that the frantic old “adverbial” dodge of Chisholm (1957, Chapter 5) and Sellars (1967) is hopeless.)

Second:We distinguish between veridical and unveridical visual experiences.Ludwig accurately sees a real tomato, but George Edward hallucinates pink rats.I would take it to be fairly uncontentious that George Edward’s experience is, falsely, as of pink rats, and has rats and their pinkness as illusory intentional objects.Likewise, Bertie’s experience is as of a green blob and has greenness as an intentional object, and what the experience reports is false.Moreover, the experience’s veridicality condition, i.e., there being a green blob where there seems to Bertie to be one, seems to exhaust not only the experience’s representational content but its qualitative content.Once the greenness has already been accounted for, what qualitative content is left?

Third (Harman (1990)):We normally “see right through” perceptual states to external objects and do not even notice that we are in perceptual states; the properties we are aware of in perception are attributed to the objects perceived.“Look at a tree and try to turn your attention to intrinsic features of your visual experience. I predict you will find that the only features there to turn your attention to will be features of the presented tree, including relational features of the tree ‘from here’” (p. 39).

Harman’s transparency argument can be extended also to the purely hallucinatory case.Suppose you are looking at a real, bright yellow lemon in good light.Suppose also that you then hallucinate a second, identical lemon to the right of the real one.(You may be aware that the second lemon is not real.)Phenomenally, the relevant two sectors of your visual field are just the same; the appearances are just the same in structure.The yellowness involved in the second-lemon appearance is exactly the same property as is involved in the first.[6]But if we agree that the yellowness perceived in the real lemon is just the yellowness of the lemon itself, then the yellowness perceived in the hallucinated lemon--the yellow quale involved in the second-lemon appearance--is just the yellowness of the hallucinated lemon itself.


On pain of circularity, the Representational theory requires color realism, for it explicates color qualia in terms of the real- (and unreal-)world colors of physical objects; “yellow” means the objective, public property inhering in physical objects.One could not then turn around and explicate the ostensible colors of physical objects in terms of color qualia (e.g., as the disposition to produce yellow qualia in normal human percipients).Of course, color realism has been a minority position in the history of philosophy, so this must be counted as a liability of the Representational theory.What physical property of a lemon is it yellowness supposed to be?For the record, I buy into D.M. Armstrong’s Disjunctive Realism, the view that a color is a wildly disjunctive but perfectly real physical property.(See Armstrong and Malcolm (1984).Lycan (1996, Chapter 4) rebuts some objections to the Disjunctive Realist view.)


2.Internalism and externalism about qualia.


As I have mentioned,Ned Block has defended properties of a type he calls “qualia,” distinct from qualia in my more familiar sense, that are narrow properties.But he has also offered a prominent objection to the Representational theory of qualia in my sense, his (1990) Inverted Earth argument (to which we shall return).And that argument depends on assuming that qualia in my sense are narrow also.In print or in conversation, a number of philosophers (including even Dretske (1996)) have since called that assumption a “deep / powerful / compelling intuition.”

I reject that judgment utterly.Until the past few years, the assumption that qualia are narrow has been just that, a gratuitous and unexamined assumption.And only recently, to my knowledge, has any philosopher actually tried to motivate it.[7]In the rest of this paper, I shall first defend the opposing thesis, that qualia in my sense are wide, and then survey and rebut the arguments I have heard for their being narrow.

(By the way, I do not myself need phenomenal externalism for my overall Representationalist project, for I have at least two ways of accommodating narrow qualia if I must (Lycan (1996), Chapter 6).But I would prefer to avoid that, and I also enjoy defending phenomenal externalism out of amor belli.)


3.Arguments for phenomenal externalism.


The arguments I shall give are far from conclusive.I believe they create a presumption; but I will be happy if they only show that qualia internalism is not obviously correct.


(A)If the Representational theory is correct, then qualia are determined by whatever determines a psychological state’s intentional content; in particular, the color properties represented are taken to be physical properties instanced in the subject’s environment, features detected by our visual systems.What determines a psychological state’s intentional content is given by a psychosemantics, in Fodor’s (1987) sense.But every known plausible psychosemantics makes intentional contents wide, explicating them in terms of relations to things external to the subject.Therefore (still assuming the Representational theory), very probably qualia are wide.Of course, the Representational theory is itself contentious; but if one grants that it is plausible or at least defensible, the further step to externalism is not a giant step. 


(B)Suppose qualia are narrow.Then Block’s Inverted Earth argument is plausible.The argument is based on an example in which a victim is unknowingly transported to a planet which is exactly like Earth except that its physical colors are inverted with respect to ours.But the victim is fitted with compensating lenses so that s/he does not notice the color switch; s/he is now misperceiving, because s/he sees green objects as red, if it is green Inverted objects that correspond to red Earthly objects under the inversion in question.The Twin Earthlings’ speech sounds just like English, of course, so they do not notice anything odd about the victim.But their intentional contents in regard to color are inverted relative to ours: When they say”red,” they mean green, and green things look green to them even though they call those things “red.”

After some period of time, Block maintains, the victim’s embedding in the physical and linguistic environment of Inverted Earth would come to dominate, since s/he would be interacting with the local color environment itself, and speaking the Inverted Earthlings’ language.Her/his wide intentional color contents would gradually shift to match those of the surrounding population on Inverted Earth.But the victim’s qualia (being supposedly narrow) would not shift.If all the intentional properties would shift but the qualia would not, then the qualia are not intentional properties.

Block’s argument would show that either qualia are narrow functional properties or they are narrow properties of a very weird kind whose existence is suggested by nothing else we know (see Chapter 6 of Lycan (1996), in which the weird properties are called “Strange Qualia”and their weirdness is exhibited).But qualia are not functional properties, at least not narrow ones.First, qualia are (ostensibly) monadic properties, while functional properties are all overtly relational.Second and more importantly, Functionalism does nothing to resolve the Russellian dilemma aforementioned.So, either qualia are wide or weirdness is multiplied beyond necessity.[8]


(C)Dretske (1996) argues much as follows:So far as qualia are intentional contents, we must concede that they are distinct from belief contents and other conceptually characterized contents (since one can have qualia themselves without possessing the corresponding concepts).Therefore, one can have qualia without being able to know or even have beliefs about what qualia one has.Suppose qualia are narrow.Then two molecular twins whose intentional contents are all different owing to different environmental surrounds must nonetheless have the same qualia.But neither the twins nor anyone else could ever know or even have reason to think that there were those shared qualia, because neither twin has any concept that applies to them.So there is no reason we should believe in them either.

Dretske’s argument is flawed, in that it assumes all knowledge is conceptual.Suppose instead that the twins have a more primitive kind of introspective acquaintance with their qualia.Perhaps they, and we, have a nonconceptual “inner sense” that does the acquainting??“nonconceptual” at least in that it does not mobilize concepts possessed by ordinary whole persons.(Unlike Dretske, I believe that we are equipped with an organ of this kind (Lycan (1995; 1996, Chapter 2)).)

But this would not alleviate the epistemic difficulty.As Moore, Harman and others have pointed out, what introspection reveals in the way of sensory qualities are only the qualities imputed (rightly or wrongly) by the senses to external objects.E.g., when I introspect my visual field in regard to the yellow ellipsoidal patch associated with a lemon that is in front of me, the sensory quality that I find in the visual field is the very yellowness that the experience presents as inhering in the lemon, nothing more.If asked, I would report it using the English word “yellow.”Of course, introspection presents a much more specific shade, intensity and saturation of yellow than anyone could express in ordinary English words, and in this sense it is non- or subconceptual.But the same is true of vision itself; nonconceptualness of this sort lends no support to the idea that the yellowness is narrow.And notice that the same argument goes for the yellowness of an hallucinated lemon; the argument does not exploit veridicality.

Thus, as before, even if there were shared qualia underlying the twins’ differing intentional contents, neither the twins nor anyone else could ever know or even have reason to think so, this time because neither twin’s introspective access reveals any sensory property other than the one that is being represented by her/his visual experience.


Before I turn to my official array of objections,I must address a special one.[9]There is a possibility that the Disjunctive Realist theory of color and a particular psychosemantics for vision might interact in such a way as to make color qualia narrow.Suppose that what unites the Realist’s disjuncts for a given color is, in fact, that each of the disjoined physical properties causes the corresponding color sensation in a normal human perceiver under standard conditions.(Lycan (1996, Chapter 4) argues that this need not induce circularity in the Representational theory, so long as the disjunctive property could in principle be identified independently of human perception and the color is not metaphysically explicated in terms of perceivers.)Suppose also that a simple causal psychosemantics is correct: A visual brain state represents a given color if it is normally caused by visual contact with an object of that color.Then, depending on the exact formulation of the psychosemantics, it may follow that molecular duplicates always have the same color qualia.Red things always cause, via their respective physical substrata, sensations of red in normal perceivers under standard conditions; what makes a visual neural state a sensation of red is its being normally caused by a red thing.So there is no possibility of varying the environmental surround in such a way as to have the duplicates’ corresponding sensations caused by things of distinct natural color-kinds, even if they are caused by different physical substrata, i.e., by different disjuncts underlying redness.

If these things should turn out to be so, then I would accede to the thesis that qualia are narrow, but maintain that they are, so to speak, only accidentally so.That is, a quale would be narrow, not because of the nature of its containing mental state, but only because of the kind of property the state’s intentional object happens to be.(And color qualia would still be “wide” in the sense of being externally referential, though the example would show that that does not suffice for wideness in the core sense of nonsupervenience.)


3.Arguments against phenomenal externalism.


(1)Block (1996) offers an argument based on the Inverted Earth victim’s memories.But I shall not take up the Memory argument here, since I have already dealt with it in Lycan (1996), and Michael Tye (1998) has responded to it more extensively.


(2)Swampman (Tye (1998); cf. Rey (1998, sec. 2.4)):The Representational theory of qualia, interpreted widely, requires some psychosemantics.Representationalists typically appeal to teleological psychosemantics (at least Lycan (1987) and Dretske (1995) do). But teleology is to be explicated in etiological terms.Therefore an organism’s teleological properties require a history of some kind.

Now, suppose a creature is freakishly caused to pop into existence in a swamp, or in outer space, and it happens to be a molecular duplicate of some actual or nonactual human being at a given moment.The creature will have all the same internal properties and all the same behavioral dispositions as the human twin; and so we should expect that the creature has the same qualia as the human.But according to the Representational theory, qualia require representation.If externalist representation itself requires teleology and teleology in turn requires having a history, then externalist representationalism is refuted by the historyless swamp creature; the externalist Representational theory counterintutively rules that Swampman and other accidental spontaneous replicas of normal sentient creatures lack qualitative experiences entirely.


Reply:First, as Millikan (1984) would have it, perhaps Swampman does lack qualitative experiences at first.(To rebut this suggestion by citing his molecular similarity to his original would simply beg the question of externalism in favor of internalism.)But, second, we must take care to distinguish teleological theories of representation from etiological theories of teleology.One can easily hold either of those without the other, and actual theorists do so.I myself believe pretty firmly in a teleological requirement on representation, but I am not committed to any particular theory of teleology.

Now, the Swampman type of example was originally mobilized against Larry Wright’s etiological theory of teleology.[10]If it succeeds against etiological theories generally, then we should abandon such theories.But that does nothing to impugn or even embarrass teleological theories of representation, which per se are not committed to any particular account of teleology in turn.Nor, I assume, could Swampman be taken to count directly against teleological theories of representation, since some accounts of teleology are forward-looking[11] and would award full teleofunctional properties to him.


(3)If qualia are wide, then there must be a relevant sort of twin earth, on which the relevant twin WGL’s sensory state has a different representational content from mine even though he is appeared to in the same qualitative way as I am now.Yet if I were spontaneously transported to that twin earth and took the place of my twin, I would notice nothing introspectively, despite a change in representational content; so my qualia must remain unchanged and so are narrow.


Reply:The same goes for propositional attitudes, i.e., the Earthling transported to Twin Earth would notice nothing introspectively, despitethe change in her/his belief and desire contents.Yet the attitude contents are still wide.Wideness does not entail introspective change under transportation.

Throughout this paper, I am assuming the standard Putnamian picture of wide propositional-attitude contents.One may reject it, as does Searle (1983).My thesis is only that if one does accept the Putnamian picture of the attitudes, one has no good reason not to accept it for qualia in my sense as well.


(4)Wilebaldo Lara (1995) has complained that if qualia are wide, then then it is hard to see how the qualities of experiences can affect behavior (which they obviously do); what actually causes behavior must be in the head.


Reply:This is just another instance of the problem of wide intentional causation, much discussed for the cases of belief and desire (e.g., Heil and Mele (1993)).It is a vexing problem, but it is not particularly about qualia.(I would argue that it is not even particularly about the mind; wide, though not intentional, causation occurs in biology, in geology, and even in astronomy.)


(5)In the propositional-attitude literature, both the transportation type of argument and the wide-causation problem have been taken as the bases of arguments for “narrow content,” as has the problem of self-knowledge given externalism (Cassam (1994); and see Georgalis (1994)).And, come to think of it, any general argument for narrow content will presumably apply to sensory representation as well as to propositional attitudes.If there is narrow content at all, then, probably sensory states have it too, and so qualia can be taken to be the narrow contents of such states.


Reply:The literature features narrow content or “content” of at least three different kinds.Pursuing my comparison to the Putnam-Fodor line on propositional attitudes, I shall focus just on a kind of narrow content that is supposed to underlie an ordinary wide attitude.The idea is that although I and Twin WGL on Twin Earth have different beliefs because mine is about water and his corresponding one is about the different but superficially similar substance XYZ, impacted within each of those beliefs is a narrow content that we share.I am myself very doubtful of that idea; but suppose it is correct.Suppose further that it extends to sensory contents.Now, even if there are such things as narrow contents underlying sensory states, these still would not correspond to qualia in the sense I have defined.For all that has been shown, the yellowness of a patch in my visual field is still a wide property, even if some other, narrow property underlies it in the same way that narrow contents are supposed to underlie beliefs and desires.On the Putnamian model, it is wide properties that are the ordinary contents of mental states, expressed by and ascribed using ordinary public English words; the narrow contents are not, or not easily, so expressed and ascribed.And qualia in my sense are the sensory contents that do go with public English words such as “yellow.” 


(6)There is no such thing as representation without a mode of presentation.If a quale is a representatum, then it is represented under a mode of presentation, and modes of presentation may be narrow even when the representational content itself is wide.Indeed, many philosophers of mind, myself included, take modes of presentation to be internal causal or functional roles played by the representations in question.Georges Rey (1998) very reasonably wonders aloud (p. 455, footnote 29) why, since I too believe in modes of presentation under which phenomenal states are given in introspection, I did not consider those modes of presentation as candidates for qualitative content.Are they not narrow qualia?[12]


Reply:Remember, the qualia themselves are properties like phenomenal yellowness and redness and greenness, which according to the Representational theory are representata.The mode or guise under which redness and greenness are represented in vision are something else again.

It can plausibly be argued that such modes and guises are qualitative or phenomenal properties of some sort, perhaps higher-order properties.I agree that, in particular, they are appealing candidates for constituting the higher-order property of “what it’s like for the subject to experience” the relevant first-order quale.(Notice that qualia in my sense and the higher-order “what it’s like” are different properties; they must be, since the latter is a property of the former.[13] )But, for the record, there is a problem that arises if we make modes or guises into higher-order qualitative properties:If we suppose that those modes can come apart from the qualia they present to the point of switching or inversion, we get the kind of weirdness scouted in Chapter 6 of Lycan (1996). 


(7)Rey (1998), “Wide Semantic vs. Qualitative States” (pp. 442-43):


[Q]ualitative experience...simply doesn’t seem to display any of the standard relational links that are usually intuitively obvious when pointed out in the case of genuinely wide states.Being a parent is obviously wide, requiring children; being a knife wound requires a knife; and being a thought about Rembrandt seems to require Rembrandt.But what does a painful, or a phenomenal red[,] experience require?Not clearly any specific external stimulus; there are phantom limb sensations and endogenously hallucinated red experiences.


Reply:It may be that being a thought about Rembrandt requires the existence of Rembrandt, depending on your view of mental proper names.But more generally, being a thought about X-kind does not require the actual current presence of an X.In what seems exactly the same way, a phenomenal red experience “seems to require” at least one red object??and usually is caused by a red object in the immediate vicinity??even though of course red experiences can be hallucinations.A psychosemantics of color vision might well appeal to the property of being-normally-caused-by-light-waves-reflected-by-a-red-object.The phantom limb example is even better:Tinglings and pains in one’s foot “seem to require” having a foot with some damage or disorder in it, but not always; the very same tinglings and pains can occur in the absence of any foot (even though they continue to represent disorder in the foot).Unless I have misunderstood him, Rey is ‘way off here.


(8)On p. 442, Rey makes a point additional to (7): that “phenomenal similarities and differences don’t track wide semantic similarities and differences” (italics original).Intentional contents are environmentally more malleable than are qualitative contents, or at least our intuitions about intentional contents are more liberal in this regard than are our intuitions about qualitative contents.In a transportation example, most of us are willing to believe that a subject’s beliefs and other attitudes gradually shift their contents in the direction of the new twin planet.E.g., if I am transported to Putnam’s original Twin Earth, the beliefs that I express using the sign-design “water” will gradually stop being about water, i.e., H2O, and be more and more about XYZ.After a month, or a year, or at any rate a decade on Twin Earth (depending on the frequency of my perceptual contact with the waterish stuff XYZ), my beliefs will be about XYZ and mostly true, rather than about H2O and often false.But it is implausible to think that on Inverted Earth, qualitative contents would shift in this obliging way.It is very unlikely that after a month, or a year, or maybe even a decade, red things would start looking green to me, even though I quickly learned to call them “green” and perhaps even came to believe that they are green.[14]Color qualia do not change as readily under environmental pressure as color beliefs do. 


Reply:It may be true that belief contents are fairly malleable.(I do not respect Block’s own reason for thinking that, which presupposes that belief content shallowly and slavishly tracks local word meanings; though I do generally endorse Sellars’ view of the priority of public language to thought, I cannot go that far.[15])But belief contents are not the appropriate point of comparison.The Representational theory of qualia does not identify qualia with belief contents, even with the contents of specifically perceptual beliefs.It identifies them with the contents of sui generis perceptual representations.

Perceptual contents differ from cognitive attitude contents in each of a number of respects.It is probable that one such way is grounded in the comparative modularity of sense modalities (Fodor (1983)):Vision, for example, is informationally encapsulated and (at the level of sensory qualities, pace Hanson (1958), Gombrich (1960) et al.) fairly well insulated from the subject’s beliefs; this is the lesson of optical illusions.Belief and memory contents might shift under transportation between Twin planets, while visual contents remain determined by nature and by the structure and teleology of the visual system, not by common-sense propositional attitudes.That would explain Block’s view (which I share) that the look of a tomato would remain the same even if belief and memory contents changed.And the explanation strikes me as very plausible; and it also explains why we intuitively think that qualia would not shift as easily as attitude contents might.


(9)A problem about modelling a shift of qualia:Even if perceptual contents do not shift in the way attitude contents (arguably) do, perhaps they would eventually shift??say after one or more centuries, if a subject could stay alive that long.After all, they are wide according to me, and environments do change.But how would a distinctive quale even imaginably undergo such a shift?For example, suppose that a quale is supposed to shift from blue to yellow.A shift from blue to yellow might reasonably be supposed to be a smooth and gradual shift along the spectrum that passes through green.But it is hardly plausible that one would experience such a shift, or a period of unmistakable greenness in particular.


Reply:We really have no plausible model for a shift of everyday attitude content either.How would a type of belief state smoothly go from being about blue to being about yellow?Presumably not by being about green in between.[16]So our presumed quale shift is no worse off than the attitudinal shift in this regard; if the present argument works for the former case, it also works for the latter, contrary to hypothesis.


(10)Rey (1998), “Non-Transparency” (pp. 440-42):Even if the Moore-Harman transparency thesis holds for perceptual states, moods and bodily sensations do not have the same transparent character.“Many have noted that states like that of elation, depression, anxiety, pleasure, orgasm seem to be just overall states of oneself, and not ‘features of presented objects’” (p. 441, italics original). 


Reply:We should grant that the qualia presented by bodily sensations such as orgasm, pains and temperature are narrow in the usual sense.For even if such sensations are intentional??as Lycan (1996) argues??their intentional objects are parts of the subject’s own body, and so the subject’s molecular duplicates will share them.(Though, n.b., there will be more exacting sense in which they are still wide, if we draw the boundary of where the “subject” leaves off closer to the central nervous system and properly within the skin.)I suspect that this narrowness of bodily-sensation qualia is an important source of the notion that qualia generally are narrow, for it is likely that philosophers extrapolate from the narrowness of bodily-sensory qualia and impute it without thinking to color- and other world-perceptual qualia; it is tempting to assimilate the “feels” of perceptual states to those of sensory states, so far as we consider perceptual states to have feels.But the “feel” terminology suits perceptual states less well than it does bodily-sensory states.Bodily sensations are themselves conceived as feelings, while perceptual states are not.Thus, it would be fallacious to infer that since bodily qualia supervene on molecular constitution, perceptual qualia do also. 

What about “whole-person” states such as elation, depression, and free-floating anxiety?I agree that their felt phenomenal characters are not “features of presented objects.”So they do not have qualia in our strict Lewisian sense (though they can lend phenomenal features to presented objects, and so can be closely accompanied by qualia).The Representational theory does not apply.

For what it’s worth, I believe the “whole-person” states are representational nonetheless.Elation represents one’s surroundings, or if extreme the whole world, as being beautiful and benign.[17]Depression represents proposed actions as unrewarding and the subject as generally unworthy.Anxiety represents there being some impending harm or disaster.But I do not contend that these representational contents exhaust the overall “feel” or subjective character of the states in question.(Here and elsewhere I distinguish an “overall subjective character” from qualia strictly construed; see Lycan (1998, pp. 484ff.).)Functional and other dispositional elements are crucial to those states as well, probably a good deal more important to their subjective characters as are their representational contents.[18]


(11)In discussion following a presentation of Lycan (1998), Kent Bach and David Chalmers raised a more serious concern:Natural representation admits of various indeterminacies.One is in complex Twin-Earth travel scenarios, where there seems to be no fact of the matter about which duplicate a mental symbol refers to (cf. Rey, pp. 443-44).Another might be in a long-term evolutionary process, when for some centuries it is vague whether or not some etiological explicans is satisfied.Now, on the Representational theory of qualia, qualia ought to inherit those indeterminacies and/or vaguenesses.But (version 1) qualia are not vague and indeterminate at all; either you have a red blob in your visual field or you do not.Or even if we agree with Dan Dennett that there is a lot less determinacy in the visual field than philosophers think, (version 2) qualia are not vague and indeterminate in the same way(s) that representation is on any going psychosemantics.(Are our biblical beliefs really about Moses, a real person?How about King Arthur?Which twin is my belief about, if I do not realize that there are two of them?)


Reply:Problems of this sort have beset the psychosemantics of belief and thought (since almost all that has been written in the way of psychosemantics has been about belief and thought).But the psychosemantics of sense modalities is going to be different, in each of three ways.First, modularity again; a sense modality like vision is informationally encapsulated and fairly hard-wired, processing highly restricted sorts of inputs in comparatively rigid ways; and each module has an identifiable function.Second, more specifically, the function of any sense modality is feature detection, and the proprietary features in question are comparatively few and determinate.Third, as perceptual psychology continues to advance, it is becoming clearer that the main task of a given modality is to map that modality’s distinctive quality space.(Cf. Churchland (1989, Ch. 7, sec. 7):Discrimination along each of Edwin Land’s three color axes (short wave, medium wave and long wave reflectances) yields a huge 3-space of possible color-visual states; taste has four dimensions generated by the four kinds of receptors found in our mouths, and any humanly possible taste will be represented as a point in the resulting 4-space, more concretely a quadruple of spiking frequencies.)

For each of these three reasons, the psychosemantics of a sense modality can be expected to be far more tightly constrained than that of belief or thought.The teleological job description of a given module will specify a proprietary range of features that the relevant organ is to detect.A module will be (veridically or not) detecting a given feature or not.The determinacy of receptor sets and of the resulting state spaces will not leave much room for indeterminacy as to what is represented.And that, I believe, is why qualia are determinate in ways that belief- and other marshmallow-state contents are not.[19]


(12)Horgan (2000):A brain in a vat whose stream of consciousness is engineered to be exactly like yours would have exactly the sequence of perceptual and other intentional states that you do.Since the brain’s external surround would be entirely unlike yours, the intentional contents it shares with you must be narrow contents. 


Reply:Non sequitur.From the viewpoint of psychosemantics, brain-in-vat cases have always been seriously underdescribed.Until the scenario is much better fleshed out, we can’t say what the brain’s intentional contents would be.Simply to assert that they are the same as yours begs the question.

And cf. my reply to (3) above; introspective indistinguishability does not show narrowness.


(13)In a paradigm case of a kind-term having wide content, one can make a direct Twin-Earth argument (cf. Rey, pp. 439-40).E.g., what is called “water” on Twin Earth is not water but XYZ; or the things that look like tigers are not tigers and the tigers look like donkeys.Accordingly, there is a “fool’s” distinction: Just as iron pyrites is not gold but fool’s gold, so XYZ is only fool’s water and the tigerish beasts on Tiger Earth are only fool’s tigers.And the contents of our various Twins’ thoughts and other representations differ from ours despite the aliens’ perfect molecular similarity to us Earthlings.But it is hardly obvious that the green objects on Inverted Earth count as “fool’s red” for Block’s victim, or that there could be any “fool’s red” at all.[20]

Suppose we agree that an object is red iff the object would look red to a normal percipient under standard lighting conditions.(My Armstrongian Disjunctive Realism is committed to that biconditional, even though the theory does not explicate redness in terms of looking red.)It follows that if an object looks red to a normal percipient in standard conditions, then the object is red.From which it seems to follow in turn that there can be no such thing as “fool’s red.”


Reply:To show the wideness of looking-red, the problem is to come up with a case in which two molecular duplicates, say Georges and some twin Georges, experience different color looks, red and, as it might be, green.Let us grant that the twins must be in the same neuroanatomically described visual state Nr (actually I doubt that even neuroanatomy is narrow).So Georges and the twin are both in Nr but supposedly in Georges this constitutes an object’s looking red to him while in the twin it constitutes the corresponding object’s looking green to him.

For expository purposes let us start with a naïve and casual psychosemantics of “typical causes”: A visual state such as Nr will represent redness iff that state is typically caused by irradiation of the retina produced by a red object.Now consider a planet, Light Earth, which is like ours except that on its surface the ambient light is strange.The light is such that green objects almost always cause normal percipients to go into Nr, and Nr is almost always produced in normal percipients by green objects.According to the naïve psychosemantics, then, Nr in Light Earthlings will represent greenness, and green objects will look green to them.If the relevant twin of Georges is Light Georges on Light Earth, we have our example of molecular duplicates visually representing different colors, hence experiencing different color looks.Moreover, if (original) Georges were to be suddenly transported to Light Earth and look at a green object, he would go into Nr, which for him represents red; so the green object would look red to him and he would be fooled.Thus, for Earthlings, green objects on Light Earth would be fool’s-red.(The foregoing argument against the possibility of fool’s redness is blocked by the stipulation that the lighting on Light Earth is not standard.Nor may it be replied á la Block that the greenness present in Light Georges’ visual state is only a shallow matter of word meaning or belief content and really has a red quale underneath it, for that would simply beg the question.)

I myself reject the naïve psychosemantics, for reasons related to what I have said in reply to the previous objection.What happens if we replace it by a psychosemantics based on having feature detection as a function?I do not see that this change would undercut the example.If the light on Light Earth has always been weird in the same way, then we may coherently suppose that in Light Earthlings Nr has the function of indicating green objects, and the argument goes through as before; molecular twins visually represent different colors, and so far as has been shown, that constitutes their experiencing different color looks, period.

There is an obvious further objection to the Light Earth argument: that Light Earth’s strange light is standard or normal for Light Earth, however weird it may be from the Earthling point of view.So when the original Georges arrives suddenly on Light Earth, sees the green object and goes into Nr, he is a normal percipient in what is locally standard lighting, so (according to our previous biconditional formula) the green object would look green, not red, to him, and so it is not fool’s-red after all.

But this is a problem for the present version of Disjunctive Realism about color, not for the wideness of qualia.The problem stems from the biconditional formula, that was licensed by the Disjunctive Realist theory, which would have to be restricted in light of the apparent environment-relativity of “normal,” “standard” and such terms.My own view is that we are entitled to take those terms as defined geocentrically, by reference to Earth rather than Light Earth, because they are words of English rather than of Light English.But, granted, this is a large subject.


Conclusion:Not everyone will accept all of the foregoing replies.But I hope no one will continue to find it obvious that qualia are narrow.[21]


Adams, E.M. (1975), Philosophy and the Modern Mind, Chapel Hill, NC: University of North Carolina Press.

Anscombe (1965), “The Intentionality of Sensation: A Grammatical Feature,” in R.J. Butler (ed.), Analytical Philosophy: Second Series, Oxford: Basil Blackwell.

Armstrong, D.M., and N. Malcolm (1984), Consciousness and Causality, Oxford: Basil Blackwell.

Bigelow, J., and R. Pargetter (1987), “Functions,” Journal of Philosophy 84: 181-96.

Block, N.J. (1990), “Inverted Earth,” in Tomberlin (1990). 

Block, N.J. (1994), “Qualia,” in S. Guttenplan (ed.), A Companion to the Philosophy of Mind, Oxford: Basil Blackwell.

Block, N.J. (1996), “Mental Paint and Mental Latex,” in Villanueva (1996). 

Boorse, C. (1976), “Wright on Functions,” Philosophical Review 85: 70-86.

Butchvarov, P. (1980), “Adverbial Theories of Consciousness,” in P. French, T.E. Uehling and H. Wettstein (eds.), Midwest Studies in Philosophy, V: Studies in Metaphysics, Minneapolis: University of Minnesota Press. 

Carruthers (2000), Phenomenal Consciousness, Cambridge: Cambridge University Press. 

Cassam, Q. (ed.) (1994), Self-Knowledge, Oxford: Oxford University Press. 

Chisholm, R. (1957), Perceiving, Ithaca, NY: Cornell University Press.

Churchland, P.M. (1989), A Neurocomputational Perspective, Cambridge, MA: Bradford Books / MIT Press.

Davidson, D. (1987), “Knowing One’s Own Mind,” Proceedings and Addresses of the American Philosophical Association 60: 441-58.

Dretske, F. (1995), Naturalizing the Mind, Cambridge, MA: Bradford Books / MIT Press.

Dretske, F. (1996), “Phenomenal Externalism,” in Villanueva (1996). 

Field, H. (1973), “Theory Change and the Indeterminacy of Reference,” Journal of Philosophy 70: 462-81.

Fodor (1980), “Methodological Solipsism Considered as a Research Strategy in Cognitive Psychology,” Behavioral and Brain Sciences 3: 63-73.

Fodor (1983), The Modularity of Mind, Cambridge, MA: Bradford Books / MIT Press.

Fodor J.A. (1987), PsychosemanticsCambridge, MA: Bradford Books / MIT Press.

Georgalis, N. (1994), “Asymmetry of Access to Intentional States,” Erkenntnis 40: 185-211.

Gombrich, E.H. (1960), Art and Illusion, New York: Pantheon Books.

Hanson, N.R. (1958), Patterns of Discovery, Cambridge: Cambridge University Press.

Harman, G. (1990), “The Intrinsic Quality of Experience,” in Tomberlin (1990). 

Heil, J., and A. Mele (eds.) (1993), Mental Causation, New York: Clarendon Press.

Hintikka, K.J.J. (1969), “On the Logic of Perception,” in N.S. Care and R.H. Grimm (eds.), Perception and Personal Identity, Cleveland, OH: Case Western Reserve University Press.

Horgan, T. (2000), “Narrow Content and the Phenomenology of Intentionality,” Presidential Address to the Society for Philosophy and Psychology, New York City (June, 2000).

Jackson, F. (1977), Perception, Cambridge: Cambridge University Press.

Kaplan, D. (1979), “On the Logic of Demonstratives,” in P.A. French, T.E. Uehling and H.K. Wettstein (eds.), Contemporary Perspectives in the Philosophy of Language, Minneapolis: University of Minnesota Press.

Kraut, R. (1982), “Sensory States and Sensory Objects,” Noûs 16: 277-95. 

Lara, W. (1995), commentary on Dretske (1996), SOFIA, Cancun, Q.R., Mexico (June, 1995).

Lewis, D. (1983), “Individuation by Acquaintance and by Stipulation,” Philosophical Review 92: 3-32. 

Lycan, W.G. (1987a), Consciousness, Cambridge, MA: Bradford Books / MIT Press.

Lycan, W.G. (1987b), “Phenomenal Objects: A Backhanded Defense,” in J. Tomberlin (ed.), Philosophical Perspectives, 1: Metaphysics, 1987, Atascadero, CA: Ridgeview Publishing.

Lycan, W.G. (1994), Modality and Meaning, Dordrecht: Kluwer Academic Publishing.

Lycan, W.G. (1995), “Consciousness as Internal Monitoring, I,” in J. Tomberlin (ed.), Philosophical Perspectives, 9: AI, Connectionism and Philosophical Psychology, Atascadero, CA: Ridgeview Publishing.

Lycan, W.G. (1996), Consciousness and Experience, Cambridge, MA: Bradford Books / MIT Press.

Lycan, W.G. (1998), “In Defense of the Representational Theory of Qualia (Replies to Neander, Rey and Tye),” in Tomberlin (1998). 

Lycan, W.G. (2000), “Representational Theories of Consciousness,” in E.N. Zalta (ed.), Stanford Encyclopedia of Philosophy (Summer 2000 Edition), URL =

Malinas, G.A. (1975), “Sensations and Understanding,” Australasian Journal of Philosophy 53: 36-45.

Millikan, R.G. (1984), Language, Thought, and Other Biological Categories, Cambridge, MA: Bradford Books / MIT Press.

Putnam, H. (1975), “The Meaning of ‘Meaning’,” in K. Gunderson (ed.), Minnesota Studies in the Philosophy of Science, Vol. VII: Language, Mind and Knowledge, Minneapolis: University of Minnesota Press.

Rey, G. (1998), “A Narrow Representationalist Account of Qualitative Experience,” in Tomberlin (1998). 

Searle, J. (1983), Intentionality, Cambridge: Cambridge University Press.

Sellars, W. (1967), Science and Metaphysics, London: Routledge and Kegan Paul.

Stalnaker, R. (1996), “On a Defense of the Hegemony of Representation,” in Villanueva (1996). 

Stich, S.P. (1978a), “Autonomous Psychology and the Belief-Desire Thesis,” Monist 61: 573-91.

Stich, S.P. (1980), “Paying the Price for Methodological Solipsism,” Behavioral and Brain Sciences 3: 97-98.

Thomason, R. (1973), “Perception and Individuation,” in M. Munitz (ed.), Logic and Ontology, New York, NY: New York University Press.

Tomberlin, J.E. (ed.) (1990), Philosophical Perspectives, 4: Action Theory and Philosophy of Mind, Atascadero, CA: Ridgeview Publishing.

Tomberlin, J.E. (ed.) (1998), Philosophical Perspectives, 12: Language, Mind, and Ontology, Atascadero, CA: Ridgeview Publishing.

Tye, M. (1995), Ten Problems of Consciousness, Cambridge, MA: Bradford Books/MIT Press.

Tye, M. (1998), “Inverted Earth, Swampman, and Representationism,” in Tomberlin (1998). 

Villanueva, E. (ed.) (1996), Philosophical Issues, 7: Perception, Atascadero, CA: Ridgeview Publishing.

Wright, L. (1973), “Functions,” Philosophical Review 82: 139-68.


[1] See Chapter 6 of Lycan (1996).

[2] Lycan (1987a, 1987b, 1996), based on Anscombe (1965) and Hintikka (1969).See also Thomason (1973), Adams (1975), Kraut (1982), Lewis (1983), Harman (1990), Dretske (1995) and Tye (1995).

[3] This argument is reprised from Chapter 8 of Lycan (1987a).

[4] The metaphysics of nonexistence is of course another matter.Another reason that the Representational theory may seem weird is that one cannot go into the metaphysics of nonexistence without getting weird.But that is everyone’s problem, not the Representational theory’s in particular.For our purposes here, just remind yourself that of course, in some sense, there are things that do not exist, and plug in your own favorite account of the nonactual.(My own view of the nonactual is defended in Lycan (1994).)

[5] A surprising but harmless consequence of this view is that qualia are not themselves properties of the experiences that present them:Qualia are represented properties of represented objects, and so they are only intentionally present in experiences.As before, the relevant properties of the experiences are, representing this quality or that.

[6] Since Moore, Malinas (1975) is the first I know to make this point.

[7] Most notably at the SOFIA conference held in Cancun, Q.R., Mexico, in June of 1995.See Villanueva (1996).

[8] Of course there may be narrow representational properties.Rey (1998) defends a narrow representationalist theory of items he calls “qualia,” but those items, like Block’s, are not qualia in my sense (see Lycan (1998), p. 482).Horgan (2000) advocates a narrow representationalist theory also. 

[9] It was put to me a few years ago by an audience member at York University, and recently by Jesse Prinz.

[10] Boorse (1976), Wright (1973).Millikan (1984) entertains a similar example as an anticipated objection to her own etiological theory.The term “Swampman” is Donald Davidson’s (1987).

[11]E.g., that of Bigelow and Pargetter (1987). 

[12]Rey goes on to compare experiences’ qualitative representational contents to the narrow contents of indexical sentences like “I am ill” and “I am wounded.”There is no denying that such sentences have a kind of meaning that is independent of their referential truth-conditions; as Kaplan (1979) has emphasized, any competent English speaker knows their meanings without knowing who they refer to or what propositions they express.And I agree that such contents or schmontents??actually I think they are just what Kaplan calls “characters”??carry the kind of ineffability that Rey says they do (p. 450).But by that point in his paper Rey is talking about something other than a quale in my strict sense; see again Lycan (1998), p. 482.

[13] For two further reasons why qualia in my sense and “what it’s like” are distinct, see Lycan (2000).The distinction is obscured by some writers’ using the phrase “what it’s like” to mean merely qualia in my sense; Dretske (1995) and Tye (1995) do this.Carruthers (2000) is also good on this topic.

[14] As I have noted, Block disagrees, since he thinks as I do that English “looks” expressions such as “looks green” express wide intentional contents.Nonetheless I side with Rey on the present point.

[15]Nor as far as to accept Sellars’ claim that animals do not have intentional states.

[16]One suggestion would be to understand the semantic shift in terms of Hartry Field’s (1973) notion of “partial reference.”I do not find that notion at all plausible, but there is not space to digress on that here.

[17] In Robert Benchley’s meditation on dentistry, “The Tooth, the Whole Tooth, and Nothing But the Tooth,” he evinces the elation that follows release from the dentist’s chair (remember, this was dentistry circa 1920, not the comparatively painless dentistry that we enjoy today):

Heigh-ho!Here’s the elevator man!A charming fellow!You wonder if he knows that you have just had a tooth filled.You feel tempted to tell him and slap him on the back.You feel tempted to tell everyone out in the bright, cheery street.And what a wonderful street it is too!All full of nice, black snow and water.After all, Life is sweet!

(Inside Benchley, New York: Harper and Bros., 1921, p. 83.)

[18] Rey’s “Non-Transparency” argument has a second half, involving an example of a well compensated color-blindness victim.But I do not yet understand the example.It is underdescribed, and in any case I do not see how it is supposed to show that any quale is narrow. 

[19]Objection:There is still notorious indeterminacy even in the simple perceptual states of lower animals.Does the frog represent flies, food, small dark moving objects, or black blobs?I say, yes.(All of the above.)See the Layering Thesis defended in Chapter 7 of Lycan (1996).

[20] Malinas (1975) expresses doubts that here could be Twin-Earth-style examples for color properties.It is salutary that Malinas made that comparison so soon after the publication of Putnam’s original article.

[21] Thanks to audiences at the University of Arkansas, the North Carolina Philosophical Society (February, 1999), the Australasian Association of Philosophy Conference (July, 2000), Duke University, and Washington University in St. Louis, for their helpful discussion.And thanks especially to Joe Levine and to Güven Güzeldere for their formal commentaries, at Arkansas and at the NCPS respectively, as well as to Dave Chalmers for helpful correspondence. 

I am especially grateful to Washington University for appointing me as Clark Way Harrison Visitor in the Fall semester of 2000, during which period I finished this paper.The hospitality of faculty, graduate students and staff was bountiful and is much appreciated.