1. Indirect realism is a serious theory of perception. Famous supports of indirect realism include John Locke and Bertrand Russell.  

2. The theory imagines perception as involving three elements. 1. The subject of perception, perhaps you looking. 2. The object of perception, say, what you are seeing. 3. Sense data, the mental representations you experience of the object of perception. The object of perception is mind-independent. Sense data are mind-dependent. The object of perception is said to be the cause of our mind’s sense data. Sense data are typically thought of as representations in and from the mind, which don’t necessarily resemble the object of perception. An analogy can be drawn between contour lines on a map which represent hills and mountains, but don’t resemble them, and sense data which work in the same way in relation to the object of perception. 

3. In response to the ontological question, indirect realists argue: “there are physical (mind-independent) things, and mental mind-dependent things.” Since there are mind-independent objects in this theory of perception it is a form of realism. However, indirect realism has been described as a two-world theory of perception, since it has a world of mind-independent objects and mind-dependent sense data. 

4. In response to the epistemological question, indirect realists argue: “the way things typically appear to us is not the same as the way they are – reality. We perceive sense data (or ideas) in our minds’ (appearances), which are caused by our interaction with the mind independent world, and which represent physical objects out there (reality).” Since our awareness of the mind-independent world is mediated, this form of realism understands our access to reality as indirect. 

5. Our knowledge of sense data and mind-independent objects differs in kind (type). Sense data can be known immediately, by introspection (the mind’s ability to examine itself). We have a knowledge by acquaintance with our sense data. As such this knowledge is often thought to be infallible. Any given sense data might not infallibly resemble the nature of reality, but we (supposedly) cannot be mistaken about the nature of our sense data. Mind-independent objects are known instead by inference. I infer (draw a conclusion/s) from sense data to the existence and nature of the mind-independent. Since my knowledge of the mind-independent is an inference from representations, this knowledge must be fallible (though it maybe reasonable).  

6. Indirect realists regard the immediate objects of perception (sense data) as distinct from mind-independent objects because of the apparent causal chain involved in acts of perception. For example, when I see a banana: (a) light strikes it; (b) is reflected off it; (c) the light crosses the space between my eye and the banana; (d) then this light enters my pupil; (e) is passes through my lens; (f) is focused on my retina; (g) there is transformed into an electrical signal which is pass down the optic nerve and (h) up to my brain where (i) I become aware of sense data. As a result, this theory insists perception is indirect. How can perception be direct if a causal chain is involved? The very concept of a chain implies links which are not all directly connected to each other. 

7. The sense data of indirect realism are sometimes described as composing a veil of perception. That is to day that sense data fall between the “mind’s eye” and the mind-independent world. It is never possible for the “mind’s eye” to see around the veil of perception to direct perceive objects of perception.  

8. Contrasts between mind-independent objects and sense data:

The argument from illusion 

1. Naïve direct realism answers the ontological question by saying that mind-independent objects exist and they have the properties we perceived them to have since we directly perceived them. They believe that: ‘how I am aware of things, is how things are.’ 

2. In response to the epistemological question, naïve direct realists insist that we are able to know mind-independent objects. In other words we are directly aware of mind-independent objects without the mediation of any third party. 

3. The argument from illusion is an argument against naïve direct realism which calls into question the directness of the naïve direct realist position. The argument claims that our awareness of mind-impendent objects is indirect (mediated) by a third party, named sense data. 

4. Sense data (plural, sense datum, singular) are collections of individual units of information from the senses, which can be thought of as mind-dependent ideas, which in turn might be thought of as mental pictures. Sense data might include colours, shapes, sizes, smells, sounds, etc. 

5. These mind-dependent sense data, the argument from illusion maintains, must stand between the subject of perception and the mind-independent object of perception, mediating perception – making it indirect, not direct. We will return to the detail of this picture of perception in the second half of this lesson. Now let’s turn our attention to the argument from illusion and start by considering three key concepts you will need to know to understand the argument.  

6. An illusion we can be define as an instance of failed perception, where our awareness does not match reality. Illusions can be contrasted with veridical perception, which can be defined as true perception, where awareness matches reality. Finally, the phrase ‘subjectively indistinguishable’ is supposed to capture the experience of not being able to tell apart from the inside, two mental events. For example, if we were to have a particularly lucid dream we might, in the dream, find it subjectively indistinguishable from wakeful experience. Other people, observing us from the outside, would be able to tell we are dreaming, but for us, in our mind, the dream experience cannot be distinguished from a wakeful experience. 

Stepped argument:

Responding to the argument from illusions 

1. There are forms of direct realism – as you will shortly see in more detail – which are not naïve, but rather philosophical. J.L. Austin defended a form of philosophical direct realism and responded to the argument from illusion in this way.  

2. Perhaps in an illusion we do see reality directly. However, what we directly perceive we form inaccurate judgements about. One way to illustrate this is to consider yourself perceiving a camouflaged animal in its habitat. Consider the image to the left, and at a glance, judge what animal is present and how many. Normally we don’t misinterpret the information we are (arguably directly) given by our senses. Even the relatively young, learn rather quickly to not misjudge a case of refraction. However, whenever we do make such a mistake we can say that this is not due – contrary to the indirect realist account – to being directly aware of reality-mismatching-sense-data and only indirectly aware of the mind-independent reality. 

3. Austin gives us a good example to help explain why inaccurate judgements don’t have to mean indirect perception. He asks you to imagine this situation. Mrs. Walker is walking through the countryside and someone has altered the appearance of a church to make it look like a barn. (I don’t know what they’ve done – put hay on the roof, perhaps boarded up the windows, etc.) Our walker now thinks: ‘that’s a barn!’. Of course, in some sense she is wrong, but in what sense? Austin askes us to consider two possible ways of viewing this situation.  

a) Mrs. Walker has directly seen a barn sense data, which misrepresents a church in the mind-independent reality only known indirectly. 

OR

b) Mrs. Walker has directly seen a mind-independent barn-made-to-look-like-a-church, and formed the misjudgement that she is seeing a barn. 

4. Austin insists that option b) is the better one. There is no need to regard our misjudgement as a mind-dependent thing itself. There is no need, in other words, to reify the appearance of the church/barn. (To reify is to make into a thing that which is not a thing. It is typically regarded as a bad thing to do. The noun form of ‘reify’ is ‘reification’.) We should instead treat the barn as an appearance and regard an appearance as a relational property/properties. 

5. Relational properties are those which exist in the interaction of the mind-independent object and the subject of perception. For example, a rectangular table has the property of being rectangular, but can appear a trapezium. What is – what has being – is one table with rectangularity as a property, but the trapezium property appears. The appearance is not a being, but rather a set of circumstances relating in a certain way. Supporters of philosophical direct realism will argue that their picture of perception has just as much explanatory and predictive power as indirect realism, but is much simpler and should therefore be preferred. In making this case philosophical direct realists are appealing to Ockham’s razor principle. 

The argument from perceptual variation 

1. Working in a similar manner to the argument from illusion there is the argument from perceptual variation. Before it is possible to understand this argument, we will need to first cover the concept of perceptual variation. When I look at a table from a bird’s eye view, directly above the center of the table, it will appear to be a perfect rectangle of a certain size. Yet, when I look at the table from a precisely parallel height to the front of the table, it will likely appear to be a thinner rectangle. From other positions around the table we can make the table appear to be a trapezium or a parallelogram. These are all examples of perceptual variations – changes in our perception relating to position, or we might add, lighting conditions and other factors. For example, under sodium street lighting objects appear a different colour to under natural light or even the light of an LED light. Lots of other example cases could be produced of perceptual variability as it seems to be “in play” all of the time. To those who say, perceptual variability is limited in extent, like illusions – it happens sometimes but not all of the time, we do have the “standard” perception – Russell argues that there is no very good reason to privilege one particular perception over another. All perception is, in Russell’s view, a perceptual variation of one sort or another. 

Responding to the argument from perceptual variability 

1. The first response a philosophical direct realist will make to perceptual variability, is to deny the extent of the problem raised. We don’t always disagree and we do privilege one size and one colour over others. Also, we manage to co-operate in a world we seem to share a perception of. We may disagree for a time about the exact proportions of a table viewed without any means for measurement. (Note the disagreement is not massive – ‘it’s a trapezium’, ‘no, it’s a parallelogram’, both of which are rectilinear shapes, and nothing like a circle.) However, any such disagreement is quickly resolved by objectively measuring the proportions of the table. We can even explain, using the science of optics, why the table might appear slightly different shapes from one angel and then another. A little physiology, physics and psychology will explain the different shades of the table. Ultimately, we believe there is one shape and size to the table, and one wave length of light bouncing off the table. These can be objectively measured scientifically, so are mind-independent. Perceptual variability only indicates that we need better understanding of the science of our direct access, not a philosophy of sense data which makes perception indirect. 

2. Secondly, the philosophical direct realist will claim that the argument from perceptual variability makes an unwarranted inference. It is an unwarranted inference to move from (reason) the fact that we each perceive an object differently due to our different perspectives, to the conclusion that we aren’t directly aware of mind-independent objects in the external world, but instead (directly) sense data, which represent these objects. There is another way to think of the fact in our reason. The different appearances of objects due to perspective are not things themselves. To say that, would be to reify appearances. (And we agreed, last lesson, following Ockham’s razor, that this is not a good idea.) An appearance is just the mind-independent conditions within which a person directly perceives the world. We can think of appearances in terms of relational properties. Relational properties are those which exist in the interaction of the mind-independent object in its circumstances and the subject of perception. The perceptual variations of a table are only it’s relational properties, and such relational properties do not prevent us from seeing the mind-independent table directly. 

3. At this point you might think that access to the mind-independent world must be indirect since perception of relational properties seems to involve a causal chain. You’ll remember that we have covered in the past the biology of sight, which involves many “causal links”, from light entering the eye through the pupil, passing through the lens onto the retina and down the optic nerve, for example. How can perception be direct if a causal chain is involved? The very concept of a chain implies links, which are not all directly connected to each other. (This argument is one in its own right – the causal argument. It is not on the A level specification, so I’ve chosen not to lay it out in full formal stepped fashion.) 

4. However, a philosophical direct realist might respond by accepting that there is causal indirectness in perception, but that this doesn’t mean that there is cognitive (or noetic) indirectness in perception. What’s the difference? Causal indirectness involves a series of causal intermediaries between the external object (or event) and the perceiver; cognitive (noetic) indirectness involves a prior awareness of something other than the external object (or event) which represents the external object (or event). It is not logically necessary to insist that since there is causal indirectness in perception, there must be cognitive indirectness too. For example, when I look at the moon through a powerful telescope the telescope is a causal link in the chain of my perception, but it is not a cognitive one. I do not look at the telescope to look at the moon (cognitive indirectness), rather I look by the telescope to look at the moon (causal indirectness). 

The argument from hallucinations 

1. To understand the argument from hallucinations you need to distinguish a hallucination from an illusion. In illusions, we are aware of a mind-independent object which seems to have, but in fact does not have, certain properties. In the case of refraction, I see a mind-independent stick which seems to have the property of being crooked, but in fact is not. It is important to note that the mind-independent object does exist in these cases, it’s just that it doesn’t have all the properties which it appears to have. This results in a difference to step 3 in the argument from illusion, compared to the argument from hallucinations. 

2. In the case of a hallucination we seem to perceive something which does not mind-independently exist at all. For example, if I experience a mirage of a cool lake in the desert, I am not perceiving something, say a lake, which, in fact, is not cool at all. Rather I am perceiving something which does not exist, mind-independently, at all – the whole lake. 

Responding to hallucinations 

1. A philosophical direct realist might respond that we can, as a matter of fact, tell when we are hallucinating. Step 5 in the argument is wrong. PERCEPTION AND HALLUCINATIONS ARE NOT SUBJECTIVELY INDISTINGUISHABLE. Another sense organ can confirm that we are having an hallucination. For example, if I think I see a pink elephant flying past me with little wings, I might reach out and attempt to touch the elephant. If I cannot feel the pink elephant, then I know I am having a visual hallucination.  

2. Those who want to support the argument from hallucinations will ask: ‘what about total hallucinations where all our senses are deceived?’ In this case we would not be able to tell that we are having a hallucination by checking the sensations of one sense organ with another. 

3. First of all, a philosophical direct realist will argue that such total hallucinations would cause a problem (of one sort or another) for any theory of perception. Secondly, the philosophical direct realist will argue that even total hallucinations must be distinguishable from veridical perception because if that were not the case we would not be able to distinguish the two concepts, which is something we do. Indeed, it is something which the argument from hallucinations has to do itself, to be able to start to convince us of its point at all. If hallucinations were utterly indistinguishable from veridical perception then we would never become aware of them, and there would be no distinction between them in our language and therefore no possibility to form the argument from hallucinations. 

4. A particularly sceptical thinker might ask: ‘how do we know that we have correctly identified two different experiences – one hallucinatory and one veridical – perhaps life is all one big hallucination?’ Further, perhaps it is part of the hallucination to think that there are distinct mental events hallucinations and veridical perceptions.  

5. This kind of sceptical (doubting) argument raises a logically possible point but there is little or no empirical (observed) evidence for it. Many things are logically possible, but we typically think that we only have to give consideration to those logical possibilities which have been shown empirically to have some chance of being actual. 

6. Ultimately, the philosophical direct realist will argue that it would be more reasonable to adopt a disjunctive theory of perception. A disjunctive theory claims that hallucinations might be subjectively indistinguishable from veridical perception, but that doesn’t mean that they veridical perception and hallucinations are objectively the same kind of mental events. Perhaps they are not. It is not uncommon to for someone to subjectively think/feel something about their own mental life, but to nevertheless be objectively in the wrong. We often make the mistake of believing that since we are so intimately acquainted with our inner mental life we can’t possibly be mistaken about it. However, this is clearly false. We often misunderstand the way we feel, or beliefs we hold about the world. Only a lengthy conversation with a counsellor might reveal to us how the contents of our own mind have been hidden to us. 

7. A disjunctive is an ‘either/or’ statement. We might want to argue that our mirage (hallucination) is either perception, or it is not perception but a hallucination instead, and therefore, imagination. Perception and imagination are two different mental states. We might concede that something like sense data are present in imagination, but not in perception. In this way – removing sense data from perception – the philosophical direct realist keeps perception direct and not indirect. 

The argument from time-lags 

1. A time-lag occurs when there is a delay between the time at which an event happens in the mind-independent world and the time at which an event is recorded through the senses. Two very clear examples of time-lags can be used to introduce the concept.  

2. Because of the distance between the Earth and the Sun, it takes light emitted from the sun eight minutes (approx.) to reach the Earth. Therefore, any event which we see on the surface of the Sun, say a solar flare, we see at an eight-minute delay to the mind-independent world. So, if an event on the surface of the sun should run for less than eight minutes, we would see that event when it was no longer there. Imagine a solar flare that burns up in a particular pattern for a minute. We see that event seven minutes after it has ceased to exist. If the event no longer exists in the mind-independent world then we cannot be seeing it directly; it’s not there to see. Instead, we must be seeing sense data of that event. Hence perception must be indirect, not direct. 

3. Something similar happens with thunder and lightning. Since the speed of light is far greater than the speed of sound we see the lightening happen before we hear the thunder happening. Yet, both events happened simultaneously in the mind-independent world, since thunder is simply the sound of lightening. Since what we are aware of in perception – split thunder and lightning – is not what exists in the mind-independent world – simultaneous thunder and lightning – we must only access the mind-independent world indirectly through sense data. 

4. Now, contrary to appearances, the sun and thunder and lightning examples above are not the exception in perception. All perception takes place at some delay, even the most rapid form, sight. Since the speed of light is finite (not infinite) it will always take some, at least, very small amount of time, for light to travel from the object that emits it or reflects it, to the subject of perception. Therefore, whatever we perceive – even through sight our most rapid sense – is always the mind-independent world as it was, but not as it is. So, we cannot be perceiving the mind-independent world directly, but instead indirectly.

Responding to time-lags 

(1) The ‘what’ and the ‘how’ of seeing 

1. The first response which a philosophical direct realist might make to the argument from time-lags is to claim that the argument is confused about the ‘what’ and the ‘how’ of seeing. Consider the following two pairs of questions.  

A) ‘Can you see the lake?’ 

‘Can you see the light reflecting from the lake? 

B) ‘Can you see the paper? 

‘Can you see the light reflecting from the paper?’ 

2. In pair A) we can make a distinction between seeing the lake and seeing light reflecting from the lake.  

3. In pair B) we can’t make a distinction between seeing the paper and seeing the light reflecting from the paper. Seeing the paper is seeing the light reflecting from the paper. Seeing the light is so tightly bound up with seeing the paper that we say we see the paper by seeing the light. 

4. The confusion the time-lag argument makes is to forget that, except in a limited set of circumstances – those which are not the norm – we don’t distinguish the ‘how’ of seeing light, from the ‘what’ of seeing the object of perception. What we see is, ordinarily, by how we see. So much so, that when we see some paper by light we don’t see the light but only the paper. This returns us to the distinction we made a few lessons ago between causal indirectness and cognitive indirectness. How we see involves a certain causal indirectness (light waves among other things), but we don’t think that this causal indirectness entails cognitive indirectness. What we see – the mind-independent object – is seen by the light waves, which reach us at some delay. We don’t look at the light waves, but rather by them. 

5. In the case of a dead star we see by the light from the star and this is seeing the star. It is not, as indirect realism claims, that we see a sense data which represents a star which we do not see because it is no longer present. 

(2)We see into the past 

1. J. L. Austin would argue that the time-lag argument presents us with two options.  

1) We don’t directly perceive mind-independent objects in the external world (the indirect realist’s position), or, 

2) We directly perceive into the past of the mind-independent external world (the philosophical direct realist’s position). 

2. For Austin, presenting time-lag arguments does not settle which option is best. We are still logically entitled to take option 2. 

3. Option 2 is what modern science says when it describes stars which we see today in distant galaxies, but which no longer exist in such galaxies. For example, we see the Carb Nebula as it was six thousand years ago. A naïve direct realist might have to give up the belief that we see the mind-independent world instantaneously (in real-time), but that doesn’t mean that the philosophical direct realist has to give up the view that we have direct access to the mind-independent world. 

4. So, for Austin we don’t need a better philosophy with sense data; we only need a better understanding of physics: light/sound take time to travel. Once you understand this physics, and some biology about how sight/hearing happen, you have all you need to know to understand that we directly access a mind-independently complex world.   

(3)The problem with giving up ‘openness’ 

1. This response attacks indirect realism as much as it defends direct realism.  

2. If indirect realism were true we would have to give up “openness” to the mind-independent world. (We’d be committed to the view that we can only directly “see” sense data not the mind-independent.) We would have to maintain that the immediate content of perception was not mind-independent physical objects but sense data instead. 

3. If we tried to live like this it would be very counter-intuitive, if not impossible. We’d have to edit out of our speech all language about physical things. This is because words, according to indirect realism, label sense data and not mind-independent things.  

4. The image to the left would have to be described as: a tall green patch is left of a wavy yellow patch. 

5. BUT if someone asked what shape the green thing is we’d have to say ‘plant-shaped’ wouldn’t we? But that’s exactly what we are not allowed to do. A plant being a mind-independent thing which we don’t have direct access to in perception. 

6. PERHAPS we could break the plant down into some not-physical-thing-shapes, but we’d still be left with the problem of explaining what ‘left of’ means in a non-physical sense. ‘Left-off’ seems to be an inherently mind-independent, spatial concept, which could not be a sense data, since sense data are mind-dependent and therefore non-spatial. 

7. It is exceptionally difficult to explain perception in a way that doesn’t involve physical concepts. This doesn’t prove that perception definitely is direct, but it means we need really good reasons to give up the openness of direct perception.