Often enough, the shapes and colours of things are present to vision, just as the pitch and loudness of sounds are present to audition. We see and hear not just individual things but also the ways those things are. These may be banalities, but it is only a small step from them to a controversial – indeed, I think, mistaken – conclusion. This is the conclusion that the universals that are instantiated by particular perceived things are themselves present to perception.

In seeing that there is an inference here, we can also see how that inference can be resisted. My purpose here is to show that the conclusion does not follow from the premise. I begin by spelling out in greater detail what the inference consists in, drawing from contemporary philosophers of perception who have made it. In the second section I draw a distinction between what I call minimal property perception and aspect experience: the idea is that the inference conflates the two, as though all cases of minimal property perception were full-blown aspect experiences. Once the distinction has been drawn, we can see that mere property perceptions underdetermine universals, so that the inference holds only for the special case of aspect experience. In section three we turn to Wittgenstein’s late remarks on aspect seeing: I draw a number of points from his discussion which help us to a better understanding of what really occurs in aspect experience. Those late remarks also form a contrast with Wittgenstein’s early account of perception given in the Tractatus. In that early work, the inference is assumed: accordingly, the later work helps us to see what it would be like to resist it. Section four spells out the contrast.

Property perception and universals

Jennifer Church, at the beginning of her book on perception, writes (Church 2013: 11)

it does not seem possible to perceive objects without perceiving at least some of their properties, or to perceive properties without perceiving at least some instantiation of those properties.

Church takes this to be rather obvious. One can see the bottle without seeing its blue colour (for example, in poor light), but even in that instance one can see other of its properties (its shape, perhaps). The only possible circumstance under which one does not see any of its properties is that under which one does not see it at all. Contrariwise, one cannot see blue without seeing something blue. If I see blue, it must be because I see a blue bottle, or a blue sky.

I think Church’s observation, taken by itself, is unobjectionable, and I will assume it in what follows. We do indeed perceive properties: in fact, we perceive properties as often as we perceive at all. But there are different ways of interpreting the observation and building on it. In particular, I want to focus on one way which may seem natural or even inevitable, and argue that this way depends on an inference that may at first escape notice but is in fact questionable.

Call properties conceived of as shared by different things universals. These are general ways for things to be, such that the bottle may be the same way as the sky. Schematically put, the inference I have in mind is an inference to the conclusion that universals are either among the objects of perceptual experiences, or that they otherwise enter into those experiences’ contents. I will expand on this presently, but we will do well to see the inference at work first. I will mention three recent authors in whose work the inference figures: Arindam Chakrabarti, Susanna Siegel, and John McDowell.

For Chakrabarti, for example, the inference merely takes seriously ordinary claims about property perception. That we perceive properties as well as objects – colours as well as things, say – is part of our common sense view of perception. If, he argues (2006: 309)

we perceive the particular material objects that common sense undeniably feels that we do, then, in so far as our perception of these ordinary objects is irreducibly predicative in structure, we need to perceive the properties meant by those predicates as well.

Chakrabarti takes this to mean that we perceive universals or generalities – these being the entities that lend predicates their meaning. Thus, defending the view that the objects of perception include universals is merely a matter of defending common sense against philosophical objections. It may be objected, for example, that universals do not exist and so cannot be perceived; alternatively, it may be argued that even if universals do exist they lack causal power of a sort that is required of objects of perception. But if objections of this sort are defeated, and common sense vindicated, there will, Chakrabarti thinks, be no reason to suppose that universals are not perceived.

Frequently, claims about the perception of universals are elaborated as explanatory claims about the ‘phenomenology’ of experience. The phenomenology of my visual experience of one bottle is explained by the presence of a property; the difference between the phenomenology of experiences of two bottles of different colours is explained by the presence of two different properties.

Siegel (2010: 6-7), for example, argues that

Visual phenomenology can be an input to recognition and categorisation, but it is not always itself

devoid of information (and misinformation) about what categories things belong to and what

relations of causal dependence they stand in.

Merely in having a visual experience, we are provided with information about the categories things fall into: that is, what universals they instantiate. When Siegel goes on to consider which categories visual experience represents things as falling into, the evidence she adduces is phenomenological. It concerns how things seem (visually, in the main cases she discusses) to observers.

At one point (2006: 485) Siegel contrasts what she calls the ‘content view’ with the view that visual experiences are ‘raw feels’. The contrast is that if visual experiences were raw feels, they would not involve the representation of properties. For a visual experience to be of a ripe tomato is, on the content view, for it to represent the tomato as being red. Since they do involve the representation of such properties, they are already sufficient to count as accurate and inaccurate. Like a belief, a visual experience represents things as belonging to categories, such that the thing may or may not belong to the category.

The work of John McDowell provides us with a third example. According to McDowell, when things perceptually seem a particular way to us concept-wielding creatures, some of our concepts are at work in the perceptual experiences themselves. When I see the redness of the tomato, my concept red is part of the content of my perception. It is this feature of perception that allows me to form non-inferential knowledge about the properties that are instantiated by perceived things.

It is crucial to McDowell’s conception that the very same concepts are at work in perception as in judgement (2009: 35):

an actualisation of the capacity to have objects come into one’s view is itself already an actualisation

of the capacity to have occur in one’s life occurrences with the sort of content that judgements have.

It follows that the same generality attaches to the content of perceptual experiences as attaches to the contents of judgements. Just as the judgement that the ball is red attributes to the ball a way of being that is attributable also to other red things, so does my perception of the ball whenever the red colour of the ball is perceptible to me.

Chakrabarti sees universals as among the objects of perception. Siegel sees differences in phenomenology between visual experiences as marking differences in the universals that are represented in those experiences. McDowell sees our concepts as entering into the contents of our experiences. What do all these views have in common? I shall say that they all regard perceptual experiences as determining universals. That is to say: mention of a determinate universal is required to characterise correctly a given perceptual experience. This is the universal that for Chakrabarti is perceived, that for Siegel is represented in the experience, and that for McDowell is expressed by its conceptual content.

There are very important differences between all of these views. But each of them, in their own way, licences the inference mentioned above. If I perceive a certain way things are, then a universal is instantiated which an adequate characterisation of my perception must mention. The inference is from the perception of properties – in the innocent sense mentioned by Church – to the determination of universals.1

Perceiving properties and experiencing aspects

It may seem to some that the inference just described is wholly trivial. But reflection on specific cases of perceptual experience may help us to see why it need not be considered so. The inference can be resisted. Here we must appeal to a notion of minimal perceptual sensitivity to properties. This is the case in which properties count as perceived, but without being marked out for special attention or inspiring thought. My contention is that such sensitivity is the normal condition under which we perceive properties, and that it underdetermines universals. We can see this by contrasting ordinary property perception with other sorts of experience in which universals are indeed determined.

To clarify the question, return to the reflections, inspired by Church, with which we began. Suppose that there is a square object – say a piece of card – in my vicinity. Suppose further that the shape of the card counts as perceived by me, in a strong enough sense to satisfy Church’s assumption that some properties of perceived objects must themselves count as perceived. Consider also that mere perceptual sensitivity requires only that I see the card, not that I attend to it or enter into explicit consideration of its shape.

All squares are also regular diamonds (example drawn from Peacocke 1992: 74-5). Thus the card instantiates both the universals square and diamond. Under the circumstances just described, need my seeing the shape of the card be sufficient to determine which universal is in question? To resist the inference, all we need is to deny that it need be so sufficient.

Note that we can imagine the circumstances under which a determinate answer would be appropriate. A prominent such circumstance is that in which the visual environment is manipulated in such a way that the shape strike me as diamond-like, and it takes some effort on my part to see it as a square. This can sometimes be done simply by orientating the card in different ways; or more effectively by placing it in the vicinity of other diamond and square shapes. Such circumstances may be deliberately brought about in the course of psychological research.

Another such experience is (as I have discovered in conversation) undergone by some subjects.2 Take a square card and place it in front of you, so that the bisector of the sides is perpendicular. Now slowly rotate it until the bisector of its axes is perpendicular to your line of vision. Some subjects claim to undergo an aspect-switch experience akin to that commonly undergone when staring at the Jastrow duck-rabbit diagram: what had been square-looking now becomes diamond-looking.

These are occasions on which the distinction between the two becomes pertinent in one’s reports of one’s perceptual experiences. That is, there are times at which one sees a square as a square, and times at which one sees it as a regular diamond. Note that neither case is affected by the fact that I know that squares are regular diamonds: in neither do I discover that the card is in fact diamond-shaped.

In the first case, what appears to occur is that I am primed to make a certain sort of assertion. It is predictable that, if the visual environment is set up in one way, I will be more likely to respond to it with the square-predicate than with the diamond-predicate. Compare the case in which I am asked to taste some wine and respond to it by holding out an open palm or a closed fist: it can be predicted that some wines will elicit the first response from most tasters, and others the second.

What about the second case? What occurs (for a subject who really undergoes an experience of aspect switch)? There may be a feeling of surprise. Note that the surprise does not consist in finding out that the card is diamond-shaped: this is something that ex hypothesi was already known. Rather, the aspect switch is a surprising sort of experience in itself: there is a feeling that the shape both has and has not changed. As Wittgenstein says (2009: § 111), “I see that it has not changed; and yet I see it differently.”

All of these are cases in which it is necessary to specify a determinate universal in our description of the visual experience. But they are also all cases in which something more than mere shape perception is involved. Some further engagement with our shape concepts, above that which is involved in mere shape perception, is implicated in each case. Thus we have seen insufficient reason to suppose that all cases of property perception are like this, which is what is required for the inference to go through.

In order to bring out the point more clearly, I would like now to discuss in some detail some cases of auditory perception. The point of my discussion will be this: in order to justify these cases as cases of hearing sounds as falling under determinate universals, we need to describe them as involving more than mere perceptual sensitivity.

A note may be heard as the tonic in some musical contexts, and not in others. Relatedly, an interval, say C-F#, may be heard as an augmented fourth or as a diminished fifth, depending on whether C or F# is heard as the tonic. A musical context may determine the latter, but if the interval is heard in isolation, without any musical context it may be heard as either, or both (examples from Peacocke 1992: 81-3). As Peacocke emphasises, these differences in hearing-as do not require any knowledge on the part of the hearer of the concepts used in the description of them or indeed any music-theoretic knowledge at all.

A single tone heard in isolation may be called (say by someone with perfect pitch) an A-sharp or a B-flat, but it is surely far from clear that it make sense to say that it can be heard determinately as either. The hearer with perfect pitch is simply making a correct judgement about the heard tone. The case of experiences of tonality within particular musical contexts may seem significantly different. Take, for example, musical passages which are ambiguous as to key. The experience of listening to the music may vary depending on how such a passage is heard, one listener finding it recognisably in one key (whether or not the listener could name that key), another finding it atonal. We may dramatise the example by supposing that the expressive qualities listeners find in the music depend on how they experience it in regard to key.

In this case, whether a particular note is heard in the tonic will depend on the way in which the listener experiences the passage.. For example, suppose the passage ends on the note A-sharp/B-flat. The listener who has experienced the passage as tonal may think that this resolves it. The listener who has experienced it as atonal may find it unresolved. It seems natural to say that the first listener has heard the note as the tonic, and the second has not.

I have just used the expressions ‘heard as’ and ‘experienced as’ in describing the difference between the two listeners’ experiences. There is more than one thing that these expressions might be used to indicate. For example, the first listener’s experience of listening to the passage might be more pleasant than the second listener’s, or the first may find the piece harmonious and the second unsettling. The appearance of a particular note might be found fitting or unremarkable by one listener, and jarring by another. The first listener may find the ending of the piece satisfying. The second may be left with an uncomfortable feeling of something unfinished. Perhaps neither listener is able to appeal to any such concept as ‘tonality’ to express their feelings or explain their reactions. These are real differences in the character of their experiences.

Alternatively, the first listener may recognise the tonality of the piece and the second may not, whether using the concept of ‘tonality’ or some other musical or aural concept. Such a cognitive difference need not be correlated with feelings of pleasantness or unpleasantness, or any other such reaction, but could be entirely analytic and intellectual in character. It involves a judgement which goes beyond the content of the aural experience itself.

In either case – that of aesthetic reaction or of music-theoretic judgement – the listener is doing more than merely hearing the pitch of a voice, instrument or recorded sound. They are beginning to react musically to what they hear.

We can mark this distinction by contrasting aspect experiences with mere property perceptions. Aspect experiences include those cases in which we undergo an aspect switch; those in which, when suitably primed, we blurt out a particular predicate; and those in which we react to a note as we would to the tonic or to the concluding note of a piece. One advantage of marking the distinction in these terms is that it aligns our discussion with Wittgenstein’s searching examination of the issue in his late writings, which I examine below. Mere property perceptions are those in which properties count as perceived. Aspect experiences are those more special cases in which things strike us as being a particular way.

We can now sum up the discussion above as follows: aspect experiences are sufficient to determine the application of universals; mere property perceptions are not. This means at least that, where a given object counts as instantiating more than one universal, one cannot capture the aspect experience without mentioning the right universal.

Wittgenstein’s discussion of aspects

At this point we can turn to Wittgenstein’s discussions of aspect experience in his late writings on the philosophy of psychology: especially the fragment frequently referred to as part two of the Philosophical Investigations, and the typescripts entitled Remarks and Last Writings on the Philosophy of Psychology. Or rather, we can turn to what I take to be an important thread in Wittgenstein’s thought running through those texts – I make no claim that it is the only or even the main thrust of the discussions. I see Wittgenstein as making three major points in the course of these discussions which are of importance for my purposes.

In the first place: property perception is not in general aspect perception. That is, what happens in ordinary cases when I am perceptually sensitive to ways things are is not what happens in the rarer case when I am struck by something being thus-and-so. Thus Wittgenstein’s key examples are not drawn from ordinary cases of property perception, but from more specific sorts of experience. The most prominent case is perhaps that of the aspect switch, as in the Jastrow diagram which switches from its duck to its rabbit aspect. Another is the example of seeing a rabbit run past and exclaiming “rabbit!” (Wittgenstein 2009: § 138). Another is his remark (2009: § 138) that “[O]ne doesn’t ‘take‘ what one knows as the cutlery at a meal for cutlery”, thus introducing the issue of what the difference is between merely seeing a fork and taking it to be one.

These are cases in which one evinces awareness not only of an object in one’s environment, but of that object as falling under some generality. One does not merely see the rabbit, but shows that one is aware that it counts as a rabbit. If one had instead exclaimed “furry!” one would have thereby evinced awareness of the same object as instantiating a different generality. Similarly, in the rare instance on which one has occasion to take a fork for a fork, one is showing awareness of the kind of object it is, in a way one need not be doing when one sits down at a dinner table with the fork in clear view.

One way in which aspect experiences differ from more mundane cases of perceptual experience, then, is that in aspect experience one is aware not only of things in the environment, but also of the generalities under which they fall. On this interpretation, Wittgenstein’s point in distinguishing the rare case of taking cutlery for cutlery from ordinary cases in which cutlery enters our lives, is that in ordinary cases of perception one is not confronted with things as instantiations of generalities but merely as particulars.

Second: the questions raised by the distinction between property and aspect perception are conceptual questions – they have to do with what we mean by ‘seeing’ and are philosophical rather than psychological. A note that Wittgenstein sounds again and again, throughout his late writings on the philosophy of psychology, is that the issue raised by aspect experiences is a conceptual one. One could mention his remark (2009: § 160) that “[T]he concept of ‘seeing’ makes a tangled impression. Well it is tangled.” But perhaps most significantly of all, we have (2009: § 181-3):

“But this isn’t seeing!” – “But this is seeing!” – It must be possible to give both remarks a conceptual justification.

But this is seeing! In what sense is it seeing?

“The phenomenon is at first surprising, but a psychological explanation of it will certainly be found.”

Our problem is not a causal but a conceptual one.

The various examples discussed by Wittgenstein may well mark out a type of experience of psychological interest. But Wittgenstein’s purpose in discussing them is rather to clarify our concepts of experience, including centrally our concepts of visual experience.

What specifically are the conceptual confusions which require clarification? Wittgenstein writes (1982: § 542):

Here we must be careful not to think in traditional psychological categories. Such as simply dividing experience into seeing and thinking; or doing anything like that.

Here Wittgenstein sounds another note which can be heard throughout the aspect seeing discussions. He is pointing to experiences of a very particular sort, experiences which are satisfactorily seen neither as perceptions nor as experiences of thinking. The experience of seeing the Jastrow diagram as a duck- or a rabbit-picture, and in particular of undergoing the switch from one to the other, is a famous example. The experience of the Necker schematic diagram switching in orientation is another.

At times, he seems to suggest that these experiences are cases in which seeing and thinking are combined, so that what they indicate is the intrusion of thought contents into perceptual experiences themselves. But the passages in which this suggestion is made are equivocal at best. Again and again the suggestion is put forward as tentative, or is put in the mouth of an interlocutor. Thus in part two of the Investigations he writes that “the flashing of an aspect on us seems half visual experience, half thought” (2009:§ 14) ,and that in the dawning of an aspect is “'[T]he echo of a thought in sight’ – one would like to say” (2009:§ 235) (my italics in both cases).

Far from offering the correct answer, the interlocutor here (if we take their words literally) is in fact making the very mistake that Wittgenstein warns against in the passage from Last Writings. This is the error of supposing that one can decompose experience into its perceptual and thought elements. As Wittgenstein says rather directly (2009: § 245), “Is being struck looking plus thinking? No. Many of our concepts cross here.”

At (2009: § 190), Wittgenstein addresses the question, “Is it [an aspect experience] a genuine visual experience?” and answers “[T]he question is: in what sense is it one?” Throughout the aspect discussions, Wittgenstein returns again and again to this question of just how and in what sense aspect experiences count as visual (or, as it might be, auditory).

Thus, for example, at (2009: § 141) the question is raised, but not settled, whether someone who recognises something immediately on seeing it has a different visual experience from someone who takes a moment to recognise it. On the face of things, there do not seem to be differences between the two cases as to which visible features of the environment are involved in the experience. The latter person can describe the scene just as well as the former, and the two experiences can eventuate in the same knowledge.

There may, nevertheless, be differences between the two cases. Wittgenstein mentions a few. Though we have no reason to think that either subject will describe the scene in way that captures it more accurately than the other, nevertheless they may describe it differently. Similarly, and this is not just an analogy but a further difference between the two cases, I imagine I might paint a portrait differently when I see the face as a familiar one – an old friend, say, or a half-forgotten celebrity (Wittgenstein: 2009: § 144).

These are differences, one might say, in our responses to visible features of the environment, and not differences in which features are visible to us, or which visible features we in fact see. Can they therefore act as criteria for whether visual experiences of the same sort are had by each subject? The question points to a fault-line in the concept of a visual experience: on the one hand, someone who merely sees and someone who recognises do not see different things; on the other, their experiences differ in what seem to be essentially visual ways.

Wittgenstein’s recommendation is not for an expansion of the domain of visible objects, similar to that effected in earlier episodes of the history of philosophy and psychology. He does not propose anything analogous the proposal that we see in three dimensions rather than two, or that we see physical objects rather than sense data or surfaces. Those philosophers and psychologists who made these proposals may very well have been in the right, but they were addressing fundamentally different questions.

To opt for such a solution would be to miss the point that we have here a tension in the concept of the visible, or of visual experience. It would be to attempt to solve the problem by distinguishing between different categories of visible things, rather than investigating the nature of our concepts of visual experience. These cases cannot be understood by considering vision on its own, or by formulating a theory of vision. This is part of the point of emphasising that the investigation is a conceptual one.

A visual experience – if this category is to include all experiences which count as relevantly visual – cannot in general be understood in terms of the experiencing of visible things. That is, aspect experiences are cases of visual experience the nature of which cannot be described by referring to things which are seen, or to visible features of the environment.

We come now to the third point. We understand what is meant by ‘seeing’ in these different cases by considering what is meant by ‘blindness’. Suppose that I see an object, under good visual conditions, but fail to see its colour. I see the book but mistake the colour of its cover, say, taking it to be green when it is in fact blue. This failure indicates a deficiency in my eyesight. This is true, I suggest, of perceptible properties in general. One who can see a given visible property has to that extent better eyesight than one who cannot.

No deficiency in my eyesight is indicated, however, if I cannot see the duck aspect of the Jastrow diagram. I can see the entirety of the diagram, in all its details, clearly and plainly, without undergoing the aspect switch or being aware of either aspect. ‘Aspect blindness’ is only in an extended sense a sort of blindness Similarly, someone who cannot hear sounds at a certain pitch is deficient in hearing. But one who fails to hear a musical passage as an introduction or as an ending is not so deficient. Although one could label this incapacity as a sort of ‘deafness’ it is clear that this is a different sort of deafness from that suffered by one who is simply hard of hearing.

In his Remarks on the Philosophy of Psychology Wittgenstein writes (1980: § 189) :

And now the question arises whether for the same reasons it wouldn’t be totally misleading to speak of “form-blindness” or “meaning-blindness” (as though one were to talk of “will-blindness”, when someone behaves passively). For a blind man just is someone who does not have a sense. (The mental defective – e.g. – can’t be compared to the blind man).

The “meaning blind” are comparable to the “mentally defective”: neither are like the blind man who simply lacks a sense, so that the use of the term “blindness” is apt to mislead.

Edward Minar (2010: 197) interprets this passage as Wittgenstein’s way of emphasising the strangeness of the situation of the meaning blind, how far away they are from everyday human life as we experience it. Minar suggests that to this extent the meaning blind are for Wittgenstein mentally defective. But I take it that Wittgenstein’s purpose in this passage is not to characterise meaning blindness as a mental defect but rather to make a negative comparison: to emphasise the distance of both conditions from literal blindness. Neither incapacity – that suffered by the meaning or aspect blind nor that suffered by the mentally defective – is a sensory incapacity as such. In the aspect case in particular, it is an inability to take part in practices (for example musical practices) which our culture finds important.

It is perhaps for this reason that Wittgenstein (2009: §260) compares “aspect-blindness” with “the lack of a ‘musical ear'”. Someone who lacks a musical ear will typically be able to hear the same sounds as a musician, but will not make musical sense of what they hear. They may, for example, be insensitive to the harmonic relations among the sounds. Such insensitivity is not indicative of any deficiency in the auditory faculty, so that the idiom involves an extended usage of the word “ear”.

I suggest that this is Wittgenstein’s conclusion and ought to be ours: aspect experiences are those which suffice for the application of determinate concepts. But this is not because universals are perceptual relata, or because they enter into the content of perceptual experiences themselves, but rather because of the role aspect experiences play in the rest of our lives.

Wittgenstein early and late

Wittgenstein’s late views on aspect experience, as I have been reading them, contrast with his brief early treatment of the topic in the Tractatus Logico-Philosophicus (Wittgenstein 1961). In this section I will give a brief sketch of that early treatment,3 and then draw out two points of contrast. Part of the point of doing so is to show that the Tractatus view aligns with the views of Chakrabarti, Siegel and McDowell. In seeing how Wittgenstein himself left his earlier views behind, we can see how the conclusions of these more recent authors might be resisted.

In the Tractatus, starting at remark 5.541, Wittgenstein gives an account of propositional attitudes – judging, believing, and so on – according to which such attitudes can be understood entirely in terms of their contents. The subjective conditions of the attitudes disappear from view in favour of the way they represent the world as being. When combined with the metaphysics of the Tractatus, which sees the world as constituted by facts, this leaves the way open to a very simple account of what occurs in true judgement. Where I judge truly, my judgement has a content which mirrors the worldly fact which it is about.

At remark 5.5423, he takes up the example of the Necker cube. Before and after the aspect switch, he says, we see different facts. It seems clear, given that Wittgenstein makes this point in the context of a discussion of the propositional attitudes, that he means to assimilate the case of aspect experience to that of judgement. Just as in the case of judgement, such an experience is all in the way it represents the world as being. To specify the fact with which I, in a veridical case of such an experience, am confronted, is already fully to characterise the experience.

Wittgenstein introduces the cube with these words: “To perceive a complex means to perceive that its constituents are related to one another in such and such a way.” He goes on to say that this explains the fact that there are two ways to see the cube. One point to notice here is that Wittgenstein thinks that the seeing of complexes is the seeing of facts. To see a complex is to see that such and such a fact holds. This makes sense given the overall doctrine of the Tractatus, according to which complexes are constituted by the holding of relations between their parts: that is, the existence of a complex objects consists in the holding of a fact. Since the ordinary objects of perceptual experience – tables, cats and the like – clearly are complexes, to make this move is to assimilate perceptual experience generally to aspect experience.

In the Tractatus, then, Wittgenstein denies any fundamental distinction between perceptual experience generally and aspect experience. The same analysis, in terms of the apprehension of facts, will suffice for both. Further, he takes the content of an aspect experience – when veridical – to consist in the mirroring of worldly facts. The content of each aspect experience is some fact which holds in the world – a fact [Tatsache] of the sort of which the world is made, according to the doctrine of Tractatus 1.1.

On Wittgenstein’s early view, the contents of true judgements (and, by extension of veridical aspect experiences) mirror the worldly facts which they represent.4 Otherwise put, worldly facts can be read off the content of a true judgement or veridical experience, and vice versa. Insofar as universals are involved in aspect experience, the same universals are already determined by the worldly facts which the aspect experience mirrors. The first point of contrast with Wittgenstein’s later view on aspect perception (as we have been reading it) is that this mirroring view is abandoned in the later work. Aspect experience involves universals which are not already determined by worldly facts. As a result, on the later view one can count as being aware of the worldly facts without one’s awareness determining specific universals. Such bare awareness underdetermines our ways of thinking about the world.

The second point of contrast has to do, not so much with the determination or otherwise of generalities by aspect experiences, but rather with which generalities are involved in such experiences. What sort of fact is perceived on the Tractatus account? Wittgenstein’s example relates to the spatial arrangement of lines on the page. Both ways of seeing the cube can be explained by reference to my perception of that arrangement. Perhaps I can also see colour facts. But the Tractatus account appeals only to facts which are, one might say characteristically visual.

On the later approach, this restriction is lifted. There is no reason to expect aspect experiences to involve any particular limited range of concepts. Thus, towards the opening of the discussion in Philosophical Investigations (2009: § 116) , Wittgenstein provides another discussion of the ways in which a schematic cube can be seen. Where the Tractatus envisaged precisely two such ways – where the cube is seen as receding from the page or as protruding toward the viewer – in the Investigations Wittgenstein mentions many others. The cube can be seen as an open box, as a wire frame, as “three boards forming a solid angle.”

One possible interpretation is that Wittgenstein is arguing for a larger range of visible properties than we had hitherto conceived of. But part of my argument here has been that this would be a mistake. Wittgenstein is not pointing to the variety of the visible, but rather to the variety of experiences which can count as visual experiences. Some experiences of being struck by the ways things are, of being surprised by them, and of coming to recognise them, can count as visual experiences. But that is not because these experiences consist in achieving mere visual awareness of the ways things are.

In principle, the range of universals that can be involved in aspect experience is as broad as the conceptual life of the experiencing subject. Correlatively, the range will differ from subject to subject dependently on their interests and background. Perhaps one could not see at all without awareness of Tractarian facts. But someone brought up in a world without wire frames cannot have the experience of seeing the cube as a wire frame, just as someone sufficiently ignorant of the ways of nature cannot have the experience of being struck by the presence of a hare rather than that of a rabbit. The point illustrates the extent to which aspect experience expresses the conceptual life of the subject as well as the visible layout of the environment.


We began by describing an inferential step which is made by many contemporary philosophers of perception. The step is from the premise that we perceive colours, shapes and the likes – the ways things are – to the conclusion that our perceptual experiences are of universals or have contents which implicate universals. Section two showed how the inference could be resisted: it is only a special case of property perception – which we might call an aspect experience – that determine specific universals. In section three we turned to Wittgenstein’s writings to help us understand the confusion: to assimilate property perceptions and aspect experiences is to miss the complexity of the concept of visual (and aural) experience. Finally, we saw how in these late writings Wittgenstein renounces his own early acceptance of the inferential step, and shows us how the scene appears once it is renounced.5


Chakabarti, Arindam. 2006. ‘On perceiving properties’. In A. Chakrabarti and P.F. Strawson eds. Universals, concepts and qualities: New essays on the meaning of predicates. Aldershot: Ashgate.

Church, Jennifer. 2013. Possibilities of perception. Oxford: Oxford University Press.

McDowell, John. 2009. Having the world in view: Essays on Kant, Hegel and Sellars. Harvard: Harvard University Press.

McGinn, Marie. 2006. Elucidating the Tractatus. Oxford: Oxford University Press.

Minar, Edward. 2010. ‘The philosophical significance of meaning blindness.’ In William Day and Victor J. Krebs eds. Seeing Wittgenstein anew. Cambridge: Cambridge University Press.

O’Sullivan, Michael. 2015. ‘The visual field in Russell and Wittgenstein’, Philosophical Investigations 38 (4): 316-332.

Peacocke, Christopher. 1992. A study of concepts. Cambridge, MA: MIT Press.

Pears, David. 1987. The false prison. Vol. 1. Oxford: Oxford University Press.

Siegel, Susanna. 2006. ‘Which properties are represented in perception?’. In Tamar Szábo Gendler and John Hawthorne eds. Perceptual experience. Oxford: Oxford University Press.

Siegel, Susanna. 2010. The contents of visual experience. Oxford: Oxford University Press.

Travis, Charles. 2013. ‘Reason’s reach’. Perception: Essays after Frege. Oxford: Oxford University Press.

Wittgenstein, Ludwig. 1958. The blue and brown books. Oxford: Blackwell.

Wittgenstein, Ludwig. 1961. Tractatus logico-philosophicus. Translated by David Pears and Brian McGuinness. London: Routledge.

Wittgenstein, Ludwig. 1980. Remarks on the philosophy of pyschology, vol. 1. Translated by G.E.M. Anscombe. Oxford: Blackwell.

Wittgenstein, Ludwig. 1982. Last writings on the philosophy of psychology. Translated by G.H. von Wright and H. Nyman. Oxford: Blackwell.

Wittgenstein, Ludwig. 2009. Philosophy of Psychology: A fragment. In Philosophical Investigations. Translated by P.M.S. Hacker and Joachim Schulte. Oxford: Blackwell.

1See also the writings of Charles Travis (see for example Travis 2013: 123ff), who argues that the objects of perception lack the sort of generality required for the occurrence of judgement or of truth- or accuracy-evaluable contents.

2 Though not by me.

3For further discussion see O’Sullivan 2015.

4Here I use the term ‘mirror’ to denote mere isomorphism of content, without implying anything about causal dependence. By doing so, I remain neutral in the dispute between ‘realist’ interpretations of the Tractatus (see for example Pears 1987: 9-10 and passim) on which the truth or falsity of propositions is determined by the dispositions of the objects named in the propositions, and alternative interpretations which deny that worldly facts have such explanatory priority. See McGinn (2006) for further discussion of the interpretive issue.

5Thanks to Michael Campbell, Charles Travis and Patrick White for numerous discussions at various times.