Perceptual Content and Fregean Myth Ruth Garrett Millikan Mind, New Series, Vol. 100, No. 4, Mind and Content. (Oct., 1991), pp. 439-459. Stable URL: http://links.jstor.org/sici?sici=0026-4423%28199110%292%3A100%3A4%3C439%3APCAFM%3E2.0.CO%3B2-A Mind is currently published by Oxford University Press.
Your use of the JSTOR archive indicates your acceptance of JSTOR's Terms and Conditions of Use, available at http://www.jstor.org/about/terms.html. JSTOR's Terms and Conditions of Use provides, in part, that unless you have obtained prior permission, you may not download an entire issue of a journal or multiple copies of articles, and you may use content in the JSTOR archive only for your personal, non-commercial use. Please contact the publisher regarding any further use of this work. Publisher contact information may be obtained at http://www.jstor.org/journals/oup.html. Each copy of any part of a JSTOR transmission must contain the same copyright notice that appears on the screen or printed page of such transmission.
JSTOR is an independent not-for-profit organization dedicated to and preserving a digital archive of scholarly journals. For more information regarding JSTOR, please contact
[email protected].
http://www.jstor.org Wed Jun 6 07:04:41 2007
Perceptual Content and Fregean Myth RUTH GARRETT MILLIKAN
1.Introduction When my daughter Natasha was very small, she often arrived in my bed in the night. One morning she woke there, eyes big with excitement: "Mommy, I saw your dream!". The dream was mine, it seems, because it was in my bed. A charming mingling of the intentional contents of a representation with attributes of the vehicle of representation-something that we philosophers try hard to avoid. The possible forms of this confusion are numerous, however, and some are vanishingly subtle. Some, I believe, persist in the work of contemporary philosophers, philosophers explicitly aware of the danger. I will warm up by discussing a number of instances of this confusion in contemporary work on perception. But this will lead up, in the end, to a very abstract message about thought. I will claim that something like a mingling of vehicle with content motivates Frege's conception of Sinn or mode of presentation. Putting things very crudely, Frege has done something like confusing sameness in the vehicle of representation with a representation of sameness. For a starting intuition, compare Kant's suggestion in the Parallogisms that Hume had confused a succession of representations with a representation of succession. To clarify this claim about Frege, I will also have to take on the question what the act is that constitutes taking one thing to be the same as itself or another.
2 . The error to be excised Let me begin by retrieving the unconscious roots of the neurosis to be excised. These roots are most easily traced through naive theories of p e r ~ e p t i o n .In~ folI This paper is based on the Gareth Evans Memorial Lecture delivered at the University of Oxford April 23, 1991. I am indebted to Justin Broackes, John Campbell, Michael Lockwood, Elizabeth Fricker, and Timothy Williamson for especially helpful discussions of an earlier draft of this paper, and to James Hopkins and Mark Sainsbury for excellent written comments. My greatest debt is to Christopher Peacocke for his help and his patience with my views. I am grateful to Christopher Peacocke's challenging work on perceptual content (Peacocke 1986, 1987, 1989a, 1989b), in which he introduces "manners" of perception (1986, 1989a) and contrasts these with Fregean modes of presentation, for leading me to investigate these possible roots of Frege's Sinne. Although we disagree on some quite fundamental points, without Professor Peacocke's help I should never have thought of looking
. .
Mind, Vol. C 4 0026-4423191 $3.00
October 1991
O Oxford University Press 1991
440 Ruth Garrett Millikan
lowing these roots through the early mind that tries to understand perception, I will be tracing a path well mapped already by others. It all begins with an impulse to reify the claims of perception. I cannot say "the impulse to reify what is perceived", because "perceive" is a success verb; whatever I perceive is surely already real and needs no resurrection. Rather, what gets resurrected is something to serve as correspondent to another thing itself unquestionably real but lacking a laymen's name. This is the act or state that stands to perception as belief stands to knowledge. Call this act or state, without prejudice as to its nature or mode (nominal, verbal, adverbial; visual, auditory, tactual etc.), a "visaging". The impulse, then, is to resurrect to correspond to a visaging something that exemplifies the properties that the world would have to have for the visaging to be veridical. Thus primitive peoples take dreams to be true, though true of another realm. That is the first ingredient in the confusion. The second is the impulse to take visaging as like having pictures in the mind and, simultaneously, to take pictures to be literally like what they picture. That ordinary pictures should at least seem to be literal likenesses of what they picture is understandable. Likewise, that perception involves some form of representation is an obvious idea. Perceptions, like representations, can be mistaken. Like pictures drawn with the purpose of showing how things are, visagings can misrepresent. From these three ingredients, out falls the irresistible theory that visagings involve items appearing before the mind that have the properties that they represent. The properties claimed by visagings to characterize the world exist in "objective reality" (Descartes), or they, or doubles of them, are true of sense data, or percepts, or phenomenal objects, or visual fields, etc. When and only when the world resembles the inner picture, then the visaging is veridical, showing how things really are. Most philosophers nowadays have undergone analysis so that they explicitly understand the unconscious motivations behind this error, and why they are not supposed to make it. Gareth Evans calls this error "the sense datum fallacy". He continues, "[ilt might better be called 'the homunculus fallacy' ... when one attempts to explain what is involved in a subject's being related to objects in the external world by appealing to the existence of an inner situation which recapitulates the essential features of the original situation to be explained.. . by introducing a relation between the subject and inner objects of essentially the same kind as the relation existing between the subject and outer objects" (1985a, p. 397). He thus suggests that the main error that results from this tempting move is the invoking of a regress: how will the inner eye then perceive the inner picture? In the same way that the outer eye does? I would like to suggest another way of looking at the sense datum fallacy. For it is surely possible to give a perfectly coherent answer to the question how the inner at Frege in this light. My ungrateful choice of a couple of Peacocke's claims and arguments to use as negative examples in the text that follows reflects that these happened, so, to be on my de'sk at the time of writing, not that they are singular in any other way.
Perceptual Content and Fregean Myth 441
eye works that does not invoke a regress. The purpose of introducing inner representations was to account for error. But the inner eye should not have the problems the outer eye does of sometimes misperceiving what is there before it. So there would be no need to suppose that it must use additional still-more-inner representations in order to see. The inner eye or mind can be taken to understand the representations before it merely by reacting to them appropriately, by being guided by them for purposes of thought and action. Thus the regress can be avoided. The broader trouble with the sense datum fallacy, I suggest, is that it produces a facade of understanding that overlooks the need to give any account at all of the way the inner understander works, any account of the mechanics of inner representation, of what kind of reacting is comprehending. Having projected the visaged properties to the inside of the mind, the assumption is that there can be no problem about how they manage to move the mind so as to constitute its grasp of them as what is represented. Their mere reclining in the mind constitutes the mind's visaging of them. Call this the "passive picture theory" of inner representation. What's wrong with it in the first instance is the passive part. And once you see that it must be the mind's reaction that constitutes understanding of an inner representation, you see that the picture part is also suspect. Why would a picture be needed to move the mind appropriately? At least, wouldn't something more abstractly isomorphic do as well?'
3. Internalizing and externalizing; demands for coherency and completeness; intermediaries The passive picture theory produces projection of properties claimed in or by the visaging onto the inner vehicle of the visaging. Call this move "content internalizing". It also produces "content externalizing", whereby properties of the vehicle of the visaging are taken, reciprocally, to show up in the visaging's content. The illusion is thus created that one both directly apprehends the nature of the vehicle of perception through the visaging and also the reverse, that one can argue from the nature of the vehicle of perception to what must be being visaged. One result is that it becomes problematic how genuine incoherences or contradictions could occur in the content of a visaging. Incoherences in content would have to correspond, per impossibile, to incoherences in the actual structure of the representation's vehicle. We can call this the demand for coherency in content. A sister result is that there could be no visaging that does not visage also all logically necessary or internal features of what is visaged. Taking what will later emerge as a On numerous occasions I have taken the position that thinking and perception likely both involve inner representation and that representation involves abstract mappings by which representations are projected onto representeds. But this claim does not entail that any particular concrete properties and relations are shared by representation and represented. Nor is it likely to be open to merely philosophical demonstration which abstract mathematical relations are shared.
442 Ruth Garrett Millikan
central example, there could be no visaging of properties without a simultaneous visaging of their internal relations. Contents lacking or failing to claim logically necessary or internal features associated with their contents would have to correspond, per impossibile, to vehicles lacking logically necessary or internal features of themselves. We can call this the demand for completeness in content. Internalizing and externalizing moves are enormously interesting, for in certain forms these moves can survive the contemporary turn that explicitly denies the phenomenally given, substituting neural representations for phenomenal ones. Indeed, there are forms in which these moves can survive even the turning of inner representations into mere cognitive dispositions and capacities, or into the states that account for these. I will argue in the case of Fregean Sinn that they can also survive turning from perception to cognition, a mode generally thought of as very unpicturelike. For example, the demand for coherency and the demand for completeness each finds subtle expression in Frege's views on conceptual content. Because the confusions that I wish to discuss cut in this way across theories that postulate experienced and those that postulate non-experienced inner representations or other non-phenomenal states, I propose to ignore such distinctions entirely. Sense data, percepts, sensations, neural states and (graspings of) Fregean senses, even when the last are interpreted as mere concepts or capacities or as states that account for these, are none of them exempt from internalizing and externalizing moves. I will speak indiscriminately, then, of the postulation of "intermediaries". Let me emphasize that: I am counting as intermediaries even capacities and the states in which they are grounded, when these are understood to account for the intentional contents of mental episodes. The error to be eradicated, then, certainly is not that of positing intermediaries. Postulation of intermediaries of some kind is essential to understanding perception and thought. The error is that of projecting, without argument, chosen properties of what is visaged or conceived onto these intermediaries, and vice versa. The error is equally that of taking this sharing of properties to constitute an explanation of mental representing. The passive picture theory causes the underlying nature of the vehicle of thought to disappear from (the theoretician's) view as an agent. The nature of the actual intermediaries for perception or thought, the actual mechanics of these, retires, leaving in its place a frictionless substitute that translates meaning directly into mental action and vice versa. Taking our central example, sameness in meaning translates into sameness in intermediaries and then into sameness in cognitive mechanics and conversely. The resulting illusion is that meanings move the mind directly or, the reciprocal, that meanings are merely an image of movements of the mind.
4. Internalizing and externalizing temporal relations Now for examples from perception. No one supposes, nowadays, that visaging colors or shapes requires that any similarly colored or shaped intermediaries
Perceptual Content and Fregean Myth 443
should appear either before the mind or in the brain.4 But have we assimilated the parallel truth about temporal visagings? In a truly fine essay, Daniel Dennett and Marcel Kinsbourne (forthcoming) have recently spoken to the multitude of confusions about this that persist in the psychological and philosophical literature. But here are two leftovers worth examining. In "Molyneux's Question", Gareth Evans (1985a) discusses the classic view that the blind cannot perceive space, this because the parts of an object can only be touched in succession, and because successive touchings could not yield a perception of the object's simultaneous spatial layout. Evans' counter is that one cannot argue "from the successiveness of sensation to the successiveness of perception", and that there is no reason why "the information contained in the sequence of stimulations" might not be "integrated into, or interpreted in terms of, a unitary representation of the perceiver's surroundings" (1985a, p. 368). Evans calls such representations "simultaneous perceptual representations of the world" (1985a, p. 369), thus expressing his basic agreement with the assumption behind the classic view, that a representation of simultaneity can only be accomplished by simultaneity among elements in the repre~entation.~ In similar vein, Evans answers with a confident but unargued "yes" the question "whether a man born deaf, and taught to apply the terms 'continuous' and 'pulsating' to stimulations made on his skin, would, on gaining his hearing and being presented with two tones, one continuous and the other pulsating, be able to apply the terms correctly" (1985a, p. 372). The assumption behind Evans' confidence seems to be that continuousness and pulsatingness in whatever medium must be represented by continuousness and pulsatingness, hence will always be recognized again. Yet first Evans, and then I, have just now represented pulsatingness and continuousness to you without using the pulsatingness or continuousness of anything in order to do so. Evans' assumption illustrates first "content internalizing", then "content externalizing".'
5. Internalizing and externalizing constancy A second example concerns the perception of change vs. constancy. Consider one of Christopher Peacocke's arguments (Peacocke 1983) for the existence of an intermediary called "sensation". This argument, ironically, is presented in support of the view that the properties of sensation are not derivable as mere correlates of the intentional contents of perception. The argument concerns the But recall, for example, this passage from Strawson's Indi~'iduals,C hapter 2: "Sounds.. . have no intrinsic spatial characteristics.. . [by contrast]. . . evidently the visual field is necessarily extended at any moment, and its parts must exhibit spatial relations to each other" (Strawson 1959, p. 65). That this is what Evans intends comes out clearly in his discussion of "simultaneous" 5's."serial" spatial concepts (Evans 1980, Part IV). But see also McDowell's footnote (Evans 1985a, p. 373) suggesting that Evans may later have rethought this issue.
'
444 Ruth Garrett Millikan
"switching of aspects" that occurs as one fixates on a Necker cube (or a duckrabbit). "The successive experiences have different representational contents", Peacocke says, yet "the successive experiences fall under the same type.. .-as Wittgenstein writes, 'I see that it has not changed"' (Peacocke 1983, p. 16). Peacocke's conclusion is that beneath the change in representational content lies a constancy in properties on the level of sensation. Now assuredly, "that it has not changed" is something that I see, but that what has not changed? My visaging has as part of its content that the world has not changed-that is where the constancy lies. Peacocke has internalized this content to yield an intermediary, a sensation, that has not changed. Compare a man looking through a perfectly ordinary window who erroneously believes he is watching a 3D movie. He quite automatically takes it that whenever he sees a change or a constancy, that is because the movie screen image has changed or been constant. Analogously, Peacocke's effective assumption seems to be that a perception of constancy can only be accomplished via an inner intermediary that is itself constant. This assumption, call it "constancy internalizing", which both philosophers and psychologists routinely fall into, has pervasive and far reaching effects. It produces the illusion of constancy at an intermediary level not just as shifts in aspect occur, but more devastating, as shifts in attention occur, and over episodes of perceptual learning. Shifts of attention are, of course, routinely coincident with perception of constancy in the object perceived, indeed, coincident with perception of constancy in the very properties upon which attention focuses and then withdraws. And so for episodes of perceptual learning. Learning to perceive, for example, learning to distinguish major triads, or learning to see the microbes in the field of the microscope, is simultaneous with the perception that what is perceived is not itself changing or undergoing reorganization over the interval. When these constancies are internalized, the illusion is produced that there is a background intermediary corresponding to the whole detailed scene before or around one in perception, an intermediary that changes only when caused to change by changes in the world outside or by shifts in the perceiver's external relations to that world. This intermediary is traditionally labeled "the sensory field", for example, "the visual field". The constancy of the hypothesized sensory field may then be externalized again. If the intermediary that supposedly stays the same is projected to become a constant content for the visaging, we arrive at a backdrop of continuing content from which there emerges a varying foreground as learning or attention switches occur-perhaps as connections are made into conception. Peacocke calls such contents, which in the case of vision determine (densely grouped alternative sets of)' complete spatial configurations of objects or surfaces around one, "scenarios" (Peacocke 1987).8 This feature allows for indefiniteness or indeterminacy due to lack of perfect visual acuity. I had much the same scenario in mind when I wrote Millikan (1984). There are passages there on perception that may be uninterpretable if one does not take this view-and
Perceptual Content and Fregean Myth 445
Combined with this sort of subsequent externalizing move, internalizing of constancy threatens to produce contradiction. How can the intermediary of perception remain constant so as to account for the perception of constancy, yet change so as to account for changes in content over changes in attention or over learning? When contradiction threatens, distinguish levels. Peacocke distinguishes two levels of properties for his intermediaries, "representational properties", and "sensational properties", the first of which concern content, the latter of which do not, although "experiences with a particular sensational property also have, in normal mature humans, a certain representational property" (1983, p. 25).9 As we will see later, Frege, in a related sort of bind, distinguishes two levels of content so that one can receive from his intermediaries projected differences not found on the other. What would it be to refuse to internalize constancy? Perhaps the perceptualcognitive systems manufacture perceptual intermediaries piece by piece, only as one needs them, each expressing only a fragment of the content that would be available for expression given other needs. The question whether this is how it works may turn on empirical evidence, perhaps on neurophysiological evidence, rather than a priori argument.I0
6. Importing determinacy If some aspect of content is merely internalized and then externalized again, this will not result in any change of content. But if an aspect of content is internalized and then filled out so as to make consistent the hypothesis of its inner reality before it is externalized again, the result may be an apparent change in content. This change may introduce an ambiguity concerning the scope of the visaging operator. For example, any property or relation that is internalized from a visaging to an intermediary must then be filled out and made determinate. If the intermediary really has the visaged property or relation it must have it in determinate form. Thus Berkeley's arguments against abstract ideas. Internalized contents cannot be abstract. But when they are first made to be concrete and then externalwith it another relative of Peacocke's view, namely, that perception involves some type of analogue intermediaries. What I claim here is that at least certain arguments for this don't go through. Drawing the distinction between these two kinds of sensational . ~ r o.~ e r t i is e snot always easy. S& Peacocke (1983, p. 2 4 4 ) . I O Would refusal to internalize constancy result in rejection of secondary qualities? For there to be secondary qualities, there must be sensations, or some other designated intermediaries, that certain external objects produce in us, and that remain the same both before and after the development of these concepts. An easy argument for the existence of such intermediaries is that we suspect, by analogy with other conceptual changes, that things appear not to change as one acquires these concepts. Are there other arguments to show that there exist, for example, sensations of red prior to conceptions of red? (Arguments that there exist prior pe~reptionsof red will not do, of course, since perceptions of red need something to be about.)
446 Ruth Garrett Millikan
ized again, the result is a change of scope for the visaging operator. Using a familiar example, if my visaging claims that there exists a large number that is the number of speckles on a certain hen, then there must exist a certain large number that the visaging claims to be the number of speckles on the hen. That this inferential move is in error becomes clearly evident when one applies it to the visagings of imagination, where its result is that I should not be able to imagine a speckled hen without imagining that it has a certain definite number of speckles. Call the move that first introduces determinacy at the intermediary level, then externalizes it as part of the visaging's content, thus moving the scope brackets over, "importing determina~y".~ I This move illustrates the demand for content completeness (see $3 above), the internal feature required for completeness in this case being determinacy.
7. Importing determinate relata A significant form of determinacy importation imports determinate relata. Any internal relation between properties (such as larger than, affth higher than) that is internalized from a visaging to an intermediary must be provided with appropriate relata: real relations don't exist without relata. If the relation is internalized as it were, "whole", that is, if it is taken that the intermediary embodies that very visaged relation, then the intermediary must possess determinate relata for the relation to relate, say, determinate sizes, pitches. But this is true for more subtle forms of internalizing too where it is not the very content itself that is taken to be internalized but rather an analogue. More precisely, the mathematical form of the transformation space around the content is internalized; a simple set of transformations (in the mathematician's sense) of the vehicle is taken to correspond oneto-one to a simple set of transformations of the represented, thus yielding in one stroke a whole representational system. In either case, determinate relata must be introduced at the level of the intermediary. Externalizing, it then appears to follow that the original visaging was of determinate relata. That is, from the fact that my visaging claims that there exist relata related by a certain relation, it is concluded that there exist relata that the visaging claims to be related by that certain relation. An easy example of the importing of determinate relata is found in Evans' "Molyneux's Question" (1985a). Evans has B, who gives the question an affirmative answer, use the "very familiar" argument that there could not be an experience of something rotating "in the visual field" without there being "four sides" to the visual field, "a, b, c, d, which can be identified from occasion to occasion" (1985a, p. 386). That is, the experience of rotation requires determinate directions for the rotation to occur from and to. I I Using "V:" as an "it is visaged that" operator, the move is from V: (3n)(n is large & a hen has n speckles) to (3n)(n is large and V:a hen has n speckles). There is nothing wrong with exporting the existence of a number, but the result here is also to import determinacy to within the scope of the visaging operator.
Perceptual Content and Fregean Myth 447
The importing of determinate relata is implicit in Peacocke's claim that a "matching profile" can be described, for example, for the visual experience of the direction from oneself in which the end of a television aerial lies (Peacocke 1986, 1989a). This matching profile is the area within which the aerial might lie given that the experience one is having of its direction is veridical. It corresponds to a solid angle, with oneself at origin, and it is determined by seeing how far to one side of its actual location the aerial can be moved without one's noticing the difference. But that one can perceive a discrepancy within a given small magnitude between A and B would be evidence that one is discriminating the absolute directions of A and B within that small magnitude only if visaging a discrepancy required that one visage a direction or range of directions for A and a different direction or range for B. And this would be necessary only if visaging a discrepancy required that the intermediaries for the visagings of A and of B be discrepant, thus having different absolute values. To appreciate that something has gone wrong here, compare pitches. If I can tell there's a difference when two pitches are as little as 2hz apart, does it follow that I can visage pitches (read the following transparently) within 2hz? I don't have absolute pitch. So either I don't hear absolutely within any such narrow range, or else I can visage exactly the same content twice without knowing it-a possibility to which I will turn a bit later. Certainly it is not clear that my visagings of pitch are in fact so accurate. Indeed, notice that there is no reason to think that there is even any natural information present in me representing the absolute values of the pitches I hear, for the phenomenon of adaptation is very deepseated in the structure of the nervous system. Quoting Oliver Selfridge (unpublished), "the range of stimuli that can be distinguished is greatly increased by the power of adaptation [of the nervous system], although the ability to signal absolute intensities is lost". In similar vein, Peacocke remarks on "what you can learn about the size of the room by seeing it" that you cannot necessarily learn by measuring it (1989a, p. 299). But my absolute sense of distance is not too good. What I can learn is mostly relative, it seems to me, and will help me only if I know independently something about the sizes of other relata involved. Suppose that 1 wrongly perceive two items on opposite sides of the room as different in length. In fact they are just the same length. Does it really follow that one or the other of my distance perceptions was wrong? How then is it determined which one was wrong? (Or can one, perhaps, wrongly perceive that the contents of two perceptions are different?)
8. Importing internal relations; the demand for same-different transparency Another scope distinction ambiguity resulting from internalizing and externalizing moves imports internal relations. Any relata that are projected from a visaging
448 Ruth Garrett Millikan
to an intermediary must be provided with all necessary internal relations. If an intermediary really embodies the relata (or analogues of them) it must also embody these relations. Externalizing, it follows that the visaging was also of these relations. Thus, from the fact that there exists an internal relation R that items A and B bear to one another and the fact that I visage A and B, it is concluded that I visage R. For example, I could not truly visage middle C and then orchestra A without visaging one as higher in pitch than the other, or visage a square and a triangle without visaging one as having more sides than the other. The demand here is for content completeness. Sameness can, for certain purposes, be treated as a relation.12 So treated it is of special interest because although there is only one kind of sameness relation on the level of intermediaries, two separable relations correspond to this on the level of content. A visaging can involve (1) separate visagings of what is in fact the same or (2) separate visagings taken as of the same. Call the first of these "visaging of sames", the second "visaging of sameness". Either can occur without the other or they can occur together. Similar remarks go for difference. We must distinguish "visaging differents" from "visaging difference". Because there are two kinds of visaging for sameness and two for difference, there are two kinds of internalizing and two kinds of externalizing moves for each. This makes possible the following relation importation moves. One can (1) import sameness: first internalize sames, yielding sames in intermediaries hence sameness, then externalize this sameness. The result is that sames are always visaged a s the same. Similarly, one can (2) import sames: first internalize sameness, yielding sameness in intermediaries hence sames, then externalize these sames, yielding that what is visaged as the same always is the same; (3) import difference, yielding that what is different is always visaged as different; (4) import differents, yielding that what is visaged as different always is different. Each of these moves yields to a demand for content completeness. No one implies any of the others although, on the assumption that what is visaged as the same is not also visaged as different," (1) & (3) put together imply (2) & (4). Totaled up, these moves produce what might be called total "same-different transparency ". I mention these importing moves here for two reasons, first, to round out the discussion of what importing consists in, and second, because it will be important in the discussion that immediately follows not to confuse sames with sameness nor differents with difference. Later I will suggest that a demand for same-different transparency is what Frege's Sinne are designed to meet, when these are given the role of bearers of information as over against mere determiners of Bedeutung. But before discussing Frege it will help quite a lot, I think, to leave sames and differents aside for a while and view some simpler examples of internalizing and externalizing and importing moves involving just sameness and difference. l2 l3
It is not in fact a relation, as I argue in Millikan (1984, Chapter 12).
This assumption is not trivial. See, for example, Crane (1988).
Perceptual Content and Fregean Myth 449
9. Moves involving sameness and difference To externalize sameness is to take it that intermediaries that are (tokens of) the same necessarily project this sameness into the content visaged. Similarly, if we externalize difference, intermediaries that are different must present their contents as being different. Pair these two moves and then introduce determinacy: any two intermediaries must be, determinately, either same or different. It follows that visagings of a pair of contents must always visage these contents either as same or as different. For example, if my visaging is of two colored items, it must either be a visaging of them as the same in color or else as different in color. The result is a scope distinction ambiguity that imports determinate sameness-ordifference. From it is not the case that I visage A and B to be different we get that I visage that it is not the case that A and B are different, hence that I visage A and B to be the same. Consider the ambiguity of "they all look the same in the dark" (roguishly said of women). Harder to spot: is it true that colors tend to look more and more alike as it grows darker? Our ordinary conceptions "seems", and "looks", and "appears" vacillate on such questions. They have ingested part of the passive picture theory of perception. If one externalizes sameness, one is logically obliged also to internalize difference: things visaged as different are not visaged as the same hence, by modus tollens, cannot correspond to intermediaries that are the same, hence must correspond to intermediaries that are different. The converse does not hold, however. If one internalizes the difference relation, then when intermediaries are the same they cannot correspond to a visaging of difference, but this does not entail that they correspond to a definite visaging of sameness. Rather, internalizing lack of visaged difference, that is, the failure to visage a difference, is the stronger internalizing move, equivalent to externalizing sameness. In parallel fashion, if one externalizes difference one is logically obliged to internalize sameness but not vice versa, and externalizing difference is equivalent to the stronger move of internalizing lack of perceived sameness. Most important, there is no logical connection between externalizing sameness and externalizing difference, nor between internalizing sameness and internalizing difference. Nelson Goodman attempted to exploit these dissociations between internalizing and externalizing differences and samenesses in defining identity for qualia. Goodman began by calling attention to an apparent paradox concerning the nontransitivity of identity over appearances: one thing, A , can appear to be the same color as a second thing, B, and the second appear the same as a third, C, yet A appear to be a different color from C. The weakest premises from which the paradox results is conjunction of the following internalization moves. (1) Internalize constancy: if B is visaged not to change as it is compared first with A and then with C, it corresponds to an intermediary that remains the same over the comparisons. (2) Internalize the sameness relation: if A and B are visaged as being the same, their corresponding intermediaries are the same, and so for B and C. (3) Internalize the difference relation: if A and C are visaged as different, their
450 Ruth Garrett Millikan
corresponding intermediaries are different. Goodman calls his intermediaries or their relevant qualities "qualia", and he does not, of course, analyze the paradox in this way. But he tries to avoid it, in effect, by externalizing sameness but not difference. Qualia a and P are identical just in case every quale y that matches either a or p also matches the other (Goodman 1966, p. 290), where ". . . to say that two qualia are so similar that they match is merely to say that on direct comparison they appear to be the same" (1966, p. 272-3). Being very careful, it is not merely difference that is internalized here but lack of sameness, that is, sameness is externalized. The assumption is that qualia that are the same never fail to appear so "on direct comparison", so that not appearing the same on direct comparison-not matching--can be a criterion of qualia difference. If one is seriously attempting to understand the ontology of qualia, a weakness of Goodman's solution is the problematic status of "on direct comparison", which nearly always invokes a counterfactual. (Surely an infinitesimally small proportion of the actual pairs of qualia particulars in the world in fact get directly compared.) Applying Goodman's criterion not epistemologically but ontologically, as telling what constitutes the fact of the matter of qualia sameness, perhaps it is immaterial whether we can know that two qualia would "match if directly compared". But the very essence of the matter would surely be to define what would constitute the fact of these qualia's being directly compared, in particular, what would constitute its being the same pair of qualia as the original pair that is being compared. We would have to explain how the identity of each quale across different comparisons-the sameness of a as compared first with P then with Y, etc.-is defined. And we must do this in a manner that permits possible comparisons for every pair of qualia (including, it should be noticed, all the merely possible ones). Appeal to counterfactual perceptions of constancy might do part of the job, but surely not very much of it. How is the job to be done, for example, for qualia separated by years of time, or appearing to different persons?I4 It is easy to produce paradox by combining internalizing of constancies with internalizing of visaged samenesses and differences. Suppose, for example, that between two identically colored objects a colored band is inserted, one that is subtly graded in color from side to side. The effect may be that while it appears that nothing has been changing color, still what started out looking like two samples of the same color now look like samples of different colors.15 Or suppose while you are watching, someone draws arrow ends on l 4 See Goodman's remarkable discussion (1966, p. 132 ff.). Goodman supposes that, as with perception of the colors of ordinary objects under uniform lighting conditions, one sees qualia sames as same but cannot discriminate all differences. At the same time he proposes a purely phenomenal definition of qualia sameness. Thus although Goodman's intermediaries are introduced as having aspects not externalized to content, still Goodman affords these intermediaries no aspects not originally internalized from content. They are not independent vehicles operating beneath perception with any properties not derived from awareness. Goodman's qualia abide in ontological limbo between the realms of appearance and reality. l 5 Alternatively, perhaps nothing appears to have been changing color. Then the paradox results only if we also import determinate sameness-or-difference(see four paragraphs above).
Perceptual Content and Fregean Myth 451
each of two parallel lines, turning them into Miiller-Lyer arrows. While appearing not to grow or to shrink, the lines will begin by appearing the same length and end by appearing different lengths. Again: those trees in the distance looked the same size until I noticed the men standing beside them. Now they appear to be quite different sizes, yet things appear not to have changed. If we internalize constancy, sameness, and 'difference, such visagings would appear to be impossible. The demand for content coherency (see $3 above) seems to rule them out. An entirely explicit externalizing and internalizing of the sameness relation occurs in Peacocke's discussion of manners of perception in (1986). Using perception of distances as his example he writes "if p is the manner in which one distance is perceived and p' is the manner in which a second distance is perceived by the same subject at the same time, and p = p ' , then the distances are experienced as the same by the subject (they match in Goodman's sense)" (1986, p. 5). Next, " ... the same manner can enter the content of experiences in different sense modalities. You may hear a birdsong as coming from the same direction as that in which you see the top of a tree: we would omit part of how the experience represents the world as being were we to fail to mention this apparent identity" (1986, p. 6). That is, an apparent identity between directions necessarily indicates a sameness in the intermediaries through which these directions are presented, a sameness of manner, regardless of the difference between these intermediaries with regard to modality. Thus there are amodal manners of perception. Peacocke clearly intends these moves to be stipulative, defining what constitutes sameness of perceptual manner. But such stipulations do not come for free. That there is or could be any such sameness must be argued. For exactly the same reason as in the case of Goodman's qualia, the sameness relation for perceptual manners could not be wholly constituted via facts regarding merely "apparent identities". Either (1) the manners themselves or, if manners are being defined here as equivalence classes, (2) whatever those intermediary items are that make up these equivalence classes, must have prior not-merely-phenomena1 identity. If the former, we need an argument that the appearance of identity can only result from a real identity in these (independently defined) manners. If the latter, we need an argument that producing-the-appearance-of-identitywhen-compared is a necessarily reflexive relation (i.e. that sameness can be externalized here) and a transitive relation. And we need to know that each pair of intermediary items is invariant over contexts with regard to producing vs. failing to produce the appearance of identity. But most important, we need an argument that the equivalence classes so formed are of relevant theoretical interest. For example, is there a necessary similarity among the ways the various members of each class determine their identical references or extensions and a necessary dissimilarity among the ways same-reference pairs from different classes do this?
452 Ruth Garrett Millikan
Another way of externalizing the sameness relation is suggested when Evans' (1985a) gives a tentative "yes" answer to Molyneux's question.16 His reasoning is that if perceptions of shape by sight and by touch produce parallel behavioral orientations in the space surrounding one, hence constitute perceptions of space for the same reason, then they are understood to be perceptions of the same. Because "[tlhere is only one behavioral space" (Evans 1985a, p. 390) within which grasp of visual and felt shapes are manifest, there is no problem about identifying across these modalities. Again, relevant sameness in relevant intermediaries-the intermediaries here are dispositions or the states in which these are rooted-is externalized to yield a visaging of sameness.
10. Frege externalizes ordinary sameness and internalizes necessary sameness Now perhaps we are warmed up enough to discuss Frege's Sinne. Frege's Sinne (or more accurately but awkwardly, graspings of these) are his intermediaries for thought about the world. Sinne are what stand between mind and world, making errors in thought possible. More telling, Sinne are what move the mind: differences among Sinne account for differences in movement if the mind is rational. Although Sinne also actually constitute a level of content, these contents or graspings of them serve (both as their own vehicles and) as vehicles for thoughts about things in the world. But Frege's explicit intention certainly is not to project properties of things as thought of-we can generalize and also call these "visaged propertiesw-upon his intermediaries. Contrast Hume, who took thoughts to be copies of impressions, themselves clearly picture-like. Frege's Sinne are modeled on sentences and sentence parts, not pictures. The only internalizing1 externalizing games that can still be played given this model are with sameness and difference. First consider the relation that Sinne bear to Bedeutungen. The relation is that of intermediary to content: one visages a Bedeutung via a Sinn. Sameness has been externalized here. If Sinne are the same, then the corresponding Bedeutungen are necessarily visaged, or at least available to the rational mind, as same. That is why the rational mind cannot take contrary intentional attitudes toward Bedeutungen visaged via the same Sinn-"the same mode of presentation". Similarly, whenever Bedeutungen are not visaged as same-both when they are neither visaged as same nor as different and when they are visaged as definitely different-then the corresponding Sinne are always different. Difference, however, has not been externalized. Bedeutungen may be taken to be the same without the corresponding Sinne being the same (this is how an insider's thoughts "Cicero" and "Tully" are supposed to be related). Put differently, when Sinne are l 6 Molyneux's question was whether a man born blind could, upon suddenly regaining his sight, immediately recognize shapes by sight.
Perceptual Content and Fregean Myth 453
different, their Bedeutungen may be taken either as same, or neither as definitely same or definitely different. On the other hand, although difference is not externalized hence sameness is not internalized, the necessary visaging of Bedeutungen as the same-the impossibility of taking opposing intentional attitudes towards Bedeutungen, the lack of informativeness of identity judgments-is internalized as implying sameness in intermediaries. According to Frege, the only way it can happen that, simply qua rational being, one necessarily takes the same attitude towards the same Bedeutung is that it is presented under the same mode of presentation again. This entails that in some cases we have a priori knowledge of the sameness of Bedeutungen, namely, in those cases where we find that adopting conflicting attitudes towards two visageds would be impossible for us as rational beings. What we find psychologically impossible as rational beings corresponds to what is ontologically impossible. Expressing this, whimsically, as an importation move, necessary visaging of sameness has become visaging of necessary sameness (taking genuine sameness as ontologically necessary). I wish to argue that neither externalization of sameness nor internalization either of sameness or of necessary sameness is a warranted move for perception or thought. Nor can these moves be saved by fiat without argument-by making them merely definitional. But let me first describe two more moves that I believe Frege makes. He externalizes sames, and he yields to the demand for same-different transparency in content (58 above). In the present paper I will only clarify, not criticize these latter moves."
11. Frege externalizes sames Still talking here of Sinne under the aspect of intermediaries for Bede~~tungen, note that Frege externalizes sames. He takes it that if two Sinne are the same, then their Bedeutungen are the same. Since things visaged must in fact be either same or different, externalizing sames equals internalizing differents (by modus tollens; similarly, externalizing differents equals internalizing sames). Compatibly, for Frege, if Bedeutungen are different, all their corresponding Sinne are also different. Could there possibly be anything problematic about this? It does not, apparently, allow for equivocation in thought. Two tokens of (graspings of) the same Sinn always have the same content, and no Sinne have double content. But, you may say, surely sense is defined as what determines reference; same sense inzplies same, determinate, reference. And you can, of course, invoke this as a definition, rather than just as a theory about items already introduced in some other way. But having done this you must not go on without argument to import sameness for Sinne qua contents. You must not just assume, that
''
1 have discussed the last of these moves-the introduction of modes of presentation as a variety of content- at length in Millikan (forthcoming).
454 Ruth Garrett Millikan
is, that sameness of sense, defined now via reference to sameness of Bedeutung, (or indeed that the having of sense at all) is a priori known.18 It could be that we must choose either that thought can be equivocal or that transparency of identity fails for Sinn.19
12. Frege yields to the demand for same-different transparency When treating Sinne as intermediaries for Bedeutungen, Frege externalizes both sameness and sames but neither difference nor differents. Thus he imports neither sameness nor sames, and imports neither difference nor differents (see 08 above). To produce same-different transparency he would need, minimally, also to externalize difference and differents (hence to internalize sameness and sames). Frege was surely right not to make these moves, for of course one need not in fact always know whether two items one thinks of are the same or different, and one can in fact make mistakes either way. If one were to make these moves, the result, put intuitively, would be that thought formed an ideal language, nonredundant and unambiguous, one thought one content, one content one thought. Hence the fact of sameness or difference in content could be read off the sameness or difference of the thoughts and vice versa. Then for the rational thinker no misidentifications of thought content should ever occur. And since contradictions would show up right on the surface of thought, inconsistencies should not occur either. Indeed, on this conception of the relation between thought and its content, the medium, being perfectly transparent, would disappear. There would seem to be no vehicle moving the mind but the very content itself. And just so it is that Frege describes Sinne-when he describes them, that is, as themselves contents rather than mere intermediaries for Bedeutungen. For Sinne as thought contents there is no distinction between the grasping of content and the thought vehicle. Identity and difference in content are identity and difference in intermediary. Sinne are contents that directly move the mind. Nor, on this level, is failure correctly to recognize content sames and differents a possibility. What then makes Frege think there are any such things as Sinne? Better, focusing the question more sharply, why does Frege introduce a second level of content onto which differents and difference in intermediaries can be projected? What happened in Frege's mind is clearly documented. First, he saw that in the case of differing definite descriptions that refer to the same thing there was a sense in which they did, but another in which they didn't, have the same content. They referred to the same thing, but they got there by different routes and from different starting points, from initial thoughts20 that had different Bedeutungen. For a discussion of these issues, see Millikan (forthcoming). Of great interest here is Campbell's discussion (1987) and prior to that, of course, Evans (1982). 2 0 Thoughts of, not thoughts that. I am not employing Frege's technical usage of "thought" here or elsewhere in this essay. Is
l9
Perceptual Content and Fregean Myth 455 But this does not give us a distinction among levels of content for the starting points, or not without regress. It does not give us a difference in content between the thoughts Tully's father and Cicero's father, for example. Why then did Frege generalize? Why did he project two levels of content upon simple thoughts? Frege's second move is continually rehearsed in the literature. Cicero is Tully is an informative thought whereas Cicero is Cicero is not, so these thoughts must have different contents. But, quite transparently, that begs exactly the question at issue. Of course the thoughts Cicero and Tully are different (or at least for some people they might be) or they couldn't move the mind differently. Their mechanical action on the mind is not the same, so clearly they have different intermediaries. The question is whether their contents must be different in order for this to be so. Might they not differ, as it were, merely in n o t a t i ~ n ?One ~ ' has to assume same-different transparency, in particular, one has already to have externalized differents, for this Fregean argument to go through. One has to already believe that content and only content can move the mind directly. Alternatively, one simply means by "content", mechanical vehicle or, more abstractly, production or occurrence of a disposition for the mind to move in a certain way. But then one needs an argument that this sort of "content" is content-that it catches something with a semantic dimension, something the sameness of which will imply, say, same way of helping to determine truth value. That this is the case is not given.22
13. The act of recognizing sameness I wish now to argue that we have no reason to suppose that the duplication in perception or thought of any kind of intermediary would make available or necessary an act of recognizing sameness in content. That is, no matter what kind of description is given of "modes of presentation", say, description in terms of thought-vehicle types, or graspings of abstract objects, or ability types, or Kaplan-style character types, or concept possession conditions (Peacocke), or ways that the thinker knows which object it is he thinks about (Evans), still, the iteration of a thought via the same "mode of presentation" cannot, simply as such, necessitate (for a rational or well-oiled mind) an act of identifying, of grasping the sameness of the contents involved. Reciprocally, there can be no direct argument from the fact that an identity is or is not epistemically given to a conclusion about identity or difference for corresponding intermediaries. There is no argument from the necessary perception of sameness (by a rational or well-oiled mind), say, from the impossibility of taking opposing attitudes toward contents, to a conclusion about sameness in the mode of presentation. That the act of identifying is not accomplished by the mere occurrence in the mind of two like thought tokens is clear in the case of tokens widely separated in 2 ' For an analysis of the function of public language sentences asserting identity and existence that accords with this remark, see Millikan (1984, Chapter 12). 2 2 There is much more on this in Millikan (forthcoming).
456 Ruth Garrett Millikan
time. Thinking of a thing one week that it is red and thinking of it the next that it is round, no matter how similar these manners of thinking of it are, does not constitute grasp that the same thing is red and round. Nor do identical thoughts just lying beside one another within the same act of consciousness or judgment constitute such an act. Suppose that two out of the twenty marbles spread out before you appear to be the same shade of blue. Surely appreciating this sameness requires an act that could not be effected just by certain intermediaries being the same. Conversely, a grasp of sameness does not require the simultaneous presence of identical intermediaries representing these sames. If it did, then one couldn't grasp that the same was both red and round had the redness been perceived a week ago. Moreover, then my grasp that it has been the very same person, namely, my daughter Natasha, that I have seen on each of innumerable occasions over the years since she used to think that dreams inhabited beds-this grasp would have to involve my having innumerable identical thoughts of her in my mind simultaneously. Nor could I grasp that this person before me is Natasha, or that little Ta is big Nat-not if these thoughts have different modes of presentation. But if intermediaries' being the same does not equal an act of taking contents to be the same, just how is it supposed to make such an act necessarily available? True, supposing that identical intermediaries always possess identical contents (that is, supposing we externalize sames), then sameness in intermediaries will be an indication of sameness in content, perhaps it will contain the fact of this sameness as natural information, and so forth. But compare: two bee dances danced side by side may jointly be an indication, or between them contain the natural information, that two sites of nectar are forty yards apart. It does not follow that the bees can read this information off the pair of dances. Not everything that falls out of a representational system is necessarily read or readable even by its primary interpreters. If we have rejected the passive picture theory of inner representation we should also be able to see that the mere being the same of two thoughts or percepts does not accomplish anything all by itself even when the fact of this sameness indicates a sameness in content, or when this falls out of the inner representational system. It must be read somehow if it is to represent a sameness. This sameness must appropriately interact with or move the thinking system in some way if it is to represent itself. The question arises then how it must move the thinking system in order to represent itself. In the same way, presumably, as the system is moved when nonidentical intermediaries are understood as representing the same. In the same way, for example, as when this person is understood to be Natasha, or Ta understood to be Nat. From this perspective, identity among intermediaries appears incidental to the thinker's act of recognizing content sameness. Nor should we fall into this nearby error: the way that the system must move or be moved in order to be grasping a sameness is just in-the-same-way-again. Given the same context, having the same effects is secured after all just by being the same, hence can add nothing to being the same. Consider: does the frog that reacts the same way each time its retinal bug-detector fires thereby cognize a sameness
Perceptual Content and Fregean Myth 457
among the bugs it eats? Something like the opposite is true, I suggest. A creature's perception that it is encountering the same thing again shows up, characteristically, in its reacting differently this time, differently in accord with what was learned last time. That the baby recognizes me is exhibited not in its crying again-that is how it reacts to strangers-but in its smiling, or exhibiting other behaviors apparently based on its earlier experience with me. Generalizing this, merely applying the same concept again, or merely effecting connection again with Evans' "behavioral space", is not to manifest a grasp of anything's sameness. In what kind of way does one's thinking have to move, then, in order to grasp an identity? We can approach the matter this way. What is the function, the point, of grasping an identity? What does one do with a knowledge of identity, say, with an identity judgment? What changes take place when I recognize that Cicero is Tully? Well, if I knew before that Cicero was bald I now also know that Tully was bald. And how does that change anything, that I now know that Tully was bald? After all, I already knew that Cicero was bald and that was exactly the same thing to know. Why not be satisfied with knowing some things about Cicero under one idea, other things under others, even if I don't know that these ideas grasp the same? So long as I pack all the right information in one way or another, why does it matter what notation I use? It matters, of course, because if I don't recognize the identity of Cicero with Tully, then I cannot combine the various things that I know about Cicero so as to yield anything new. Every mediate inference pivots on amiddle term of some kind, on the ability to recognize that two thought tokens share a content. Similarly, only through recognizing the identity of an item currently perceived with an item perceived earlier can what was learned earlier be joined with what is perceived now to yield informed action. Recognizing incompatible attitudes, hence avoiding contradiction, also turns on a grasp of identity. In short, grasp of identity is the pivot on which must turn every exercise of thought that collects together information, effects its interaction, or applies it. Indeed, to identify the contents of two ideas (tokens) consists in no more and no less, I suggest, than that the thoughts-that (tokens) in which these are embedded interact or combine, or that there is preparation for them to combine, in this manner.23For example, for the contents of ideas had yesterday or last year to be identified by me with the contents of ideas had today requires that the former causally affect the uses of the latter in ways that produce new information or informed action.24Conversely, returning to Evans' speculations on Molyneux's question (end of $9 above), only in the act of combining spatial information obtained through one modality with information obtained through another to yield behavior or thought guided by both together does a person manifest grasp of the identity of spatial aspects over different sensory modalities. 2 3 More accurately, acts of identifying are performed by systems that have not mere dispositions but proper functions to perform such acts. See Millikan (1984, Chapter 15) for a fuller discussion of these acts. 2 4 Being more accurate, in ways that have production of such information or action as a proper function (Millikan 1984, Chapter 15).
458 Ruth Garrett Millikan
14. Why sameness need not externalize nor necessary sameness internalize Such a combining, as I have already suggested, does not obviously require a sameness in the intermediary tokens combined. Nor is it obvious that it would be required by such a sameness. Nor can the appearance of identity in content be used to define identity for corresponding intermediaries. As we saw from the discussion in $9 above of Goodman's and Peacocke's moves, externalization of sameness cannot be accomplished by fiat. On the other hand, it might be claimed that a reasonable hypothesis about the mechanics of human conception (possibly as different from human perception) is that our conceptual abilities include the compulsory disposition, under specifiable mental conditions, to perform acts of identifying over pairs of identical thought vehicles. By "vehicles" we would mean here things whose nature determines thought movement, "mental mechanics". And by "identical" thought vehicles, we would mean vehicles that determined the same dispositions to movement. Given these definitions and the above hypothesis, then a capacity (given a rational well-oiled mind) to withhold a disposition to identify under the specified conditions would indicate a difference in thought vehicles. But first, the hypothesis would have to be evidenced separately. Second and far more important, for this to yield the result that identical modes of presentation always make the sameness of their contents available to the rational mind, one would have to argue that vehicles of thought so defined are never just differences in mental orthographies but must correspond one-to-one to modes of presentation-to differences in mental semantics. They must correspond, that is, to different ways of contributing to the determination of truth values for t h o ~ g h t sOnly . ~ ~ then would one have an argument for externalizing sameness. Now suppose that these things could be argued. It still would not follow that necessary visaging of sameness could be internalized. For it would not follow that the only acts of identifying that were compulsory under the specified conditions were acts resting on sameness of vehicle. Recall, for example, the identifying of a heard with a seen direction. And recall that we can learn to perceive hitherto unrecognized identities directly or compulsorily via perceptual learning.
15. Conclusion I wish to state clearly that the official conclusion of this essay is not negative but agnostic. I have argued that certain arguments don't go through, not that their conclusions are false. I have argued that certain points cannot be taken by definition, 25
I claim that this argument cannot be given (Millikan forthcoming).
Perceptual Content and Fregean Myth 459
not that they cannot be taken at all. Readers must draw their own conclusions about whether there are any other arguments for various of these conclusions, or any other means of taking these points. But I would like also to suggest that there is a very strong bias in the contemporary tradition toward trying to construct theories of mental content that will support the Fregean moves with sames and sameness, and that there seems to be no particular reason to nourish this bias.
Department of Philosophy University of Connecticut Storrs CT 06268 USA
RUTH GARRETT MILLIKAN
REFERENCES Campbell, J. 1987188: "Is Sense Transparent?". Proceedings of the Aristotelian Society, 88, pp. 273-92. Crane, T. 1988: "The Waterfall Illusion". Analysis, 48, 3, pp. 142-7. Dennett, D. and Kinsbourne, M. (forthcoming): "Time and the Observer: the Where and When of Consciousness in the Brain". Brain and Behavioral Sciences. Evans, G. 1980: "Things without the Mind". In Philosophical Subjects: Essays Presented to P. F. Strawson, Zak van Straaten, ed., Oxford: Clarendon Press, pp. 76-1 16. Reprinted in Evans 1985b. -1982: The Varieties of Reference. Oxford: Clarendon Press. -1985a: "Molyneux's Question". In Evans 1985b, pp.364-99. -1985b: Collected Papers. Oxford: Clarendon Press. Goodman, N. 1966: The Structure of Appearance, Second Edition. Indianapolis, IN: Bobbs-Merrill. Millikan, R.G. 1984: Language, Thought, and Other Biological Categories. Cambridge, MA: Bradford Books/MIT Press. -(forthcoming): "White Queen Psychology". In her White Queen Psychology and Other Essays for Alice. Cambridge, MA: Bradford Books/MIT Press. Peacocke, C. 1983: Sense and Content. Oxford: Clarendon Press. -1986: "Analogue Content". Proceedings of the Aristotelian Society Supplementary Volume 60, pp. 1-1 7. -1987: "Depiction". The Philosophical Review, 96, pp. 383-41 1. -1989a: "Perceptual Content". In Themes From Kaplan, J. Almog, J . Perry and H. Wettstein, eds., Oxford: Oxford University Press, pp. 297-329. -1989b: Transcendental Arguments in the Theory of Content; An Inaugural Lecture delivered before the University of Oxford on 16 May 1989. Oxford: Clarendon Press. Selfridge, 0.(unpublished): Tracking and Trailing. Strawson, P.F. 1959: Individuals. London: Methuen & Co.