Is the imagination real like the instantiated things?

The concept of sensory data in Alfred Ayer

1. The problem of perception

From time to time we get confused about the strange things our senses hypocritically offer us. A rod that was just perfectly straight relinquishes this property when it is immersed in water in order to obtain the predicate `broken ', which we withdraw from it as soon as we feel the rod under water or pull it out of the water. A circular coin lies on the table in front of us, but we cannot avoid calling its outline, which we draw in the air with our finger at it, elliptical. The watering hole, the goal of our pathetic crawling through the desert, turns out to be a mirage and any further effort is completely superfluous.

It doesn't matter whether such oddities represent an ignorable minor matter or the starting point of a heated academic debate: The drowning woman finds herself exhausted and with an undamaged rescue tool on the bank; his lovingly handcrafted ceramic plates, and an overworked desert traveler cannot accuse her equally overworked companion of clandestinely stealing water.

Obviously, it doesn't take any intellectual giftedness or rash to get to the heart of the matter: one thing is just that reality, and the other is just Sham, in which Material things and actual qualities make up the real world, Deceptions or. mere appearances on the other hand the apparent world. The rod is in reality straight and seems just broken to be; the coin sees actually elliptical out, is but actually round; and the apparent Mirage exists basically not at all.

Of course we haven't said a word about it yet What it is at all that we perceive there every time. Because the objects of our perception cannot be material things, since we do not perceive them with their ´real´ properties. Still we take anything true. This anything is called `sensory date`, was introduced ad hoc and is constantly growing the most admirable flowers, sown in life-affirming spirits. The breeding ground for this is provided by questions that inevitably follow the assertion of sensory data, and the philosopher Alfred Ayer has listed a number of them: If sensory data are a function of material things, “how they perform this function, and what is theirs relation to the material things [?] [...] There is also dispute about the properties of sense-data, apart from their relationship to material things: whether, for example, they are each of them private to a single observer; whether they can appear to have qualities that they do not really have, or have qualities that they do not appear to have; whether they are in any sense `within` the percipient`s mind or brain."1

These questions could elicit different and quite imaginative answers, each of them at least vain enough to make an excellent name for itself: on the one hand there is the Naive realismwho freely claims that the senses tell us something about the things that exist outside and independently of us. A less naive descendant of this theory, the Perceptual realism 2, additionally introduces the aspect of causal relation a: "The field, by reflecting light, causes me to have the visual experience that is part of my seeing the very field." (Audi: 1998, p. 28). With a lot more sensitivity and suspicion, she pulls Show Theory (see Audi: 1998, p. 30) on field; to say that someone perceives something is to say that something is so and so to him appears. Since it is not claimed that things have really the respective properties can be explained with this theory also hallucinations. In addition, it rejects any concession to possible causal relations, but gets into difficulties as soon as the question of hallucinations comes up, because the hallucinated appearances are not appearances of something, but just about something non-existent. The adverbial theory of perception (see Audi: 1998, p. 37) distances itself from the popular breakdown of perception into a (perceived) object and the person who perceives it. Instead, it analyzes perception as one Kind and meadowTo perceive things which in us are an experience of a certain quality cause: "Seeing a parallelogram" (in illusional and hallucinatory cases) can be a term for ascribing a certain visual experience. "(Audi: 1998, p. 37). Unlike the theory of appearance, it even manages to explain hallucinations: namely, it denies that in these cases we have the sensory experience of one Object to have; Such experiences do not represent any kind of perception for them.

A rejection of all of these types of direct realism comes from the Sensory data theories: The only thing that is allowed to be called an object of our perception is Sensory data. Even so, the existence of material things is not flatly denied, but at most we take them indirectly, by means of the sensory data, true: This is why we sometimes call this type of theory indirect, representative realism designated. There is, however, a kind of sensory data theory that represents a compromise between the above-mentioned theory and the adverbial theory. You could use the latter as it is primarily based on the weight Experience and not on that experienced object lays as Sensory experience theory the former, on the other hand, as they want to give the concept of an object an appropriate place Theory of Perceiving Experience (see Audi: 1998, p. 40). Now, however, the need for sensory data can be maintained, even if we are advocates of sensory experience: when we perceive something, we make direct acquaintance with sensory data.

One direct unrealism represents the so-called Phenomenalism dar: We don't need to look desperately for impressive empirical evidence for the existence of material things. It is quite enough, ours Utterances to decompose, in such a way that material things and expressions designating their qualities no longer appear in them, but instead a series of sensory data expressions; yes, material things are nothing more than a set of Propositions about sensory data. In addition to this sensory data phenomenalism, there is the adverbial phenomenalism: Material things are constructed solely on the basis of sensory experience, not sensory data, without in any way ascribing reality to them. Both varieties of phenomenalism completely deny the independent existence of material things; these are "not metaphysically real: things that are` out there`, which are the sort of things we think of as such they would exist even if there were no perceivers. " (Audi: 1998, p. 43).

Alfred Ayer has become involved in The Foundations of Empirical Knowledge busy with the various theories of perception. On the following pages I will outline Ayer's remarks. The starting point is the argument from illusion, to which a representation of the factual opinion on the one hand and the linguistic perception on the other hand connects. Ayer's concept of sensory data will be presented on the basis of this. Ayer is in a sense a defender of a sense data theory, especially against other theories of perception, but he also turns against sense data theorists themselves: he wants the sense data theory to be linguistically and contrary to opinions other than in fact Understand the theory to be interpreted. Finally, a separate chapter will be devoted to Ayer's confrontation with phenomenalism.

2. Ayer and the problem of sensory data

2.1. The argument from illusion

The argument from illusion represents the starting point of Ayer's remarks. Therefore, it should be briefly introduced at this point.

At the beginning of the argument it is stated that quite ordinary and by no means scary objects are capable of exhibiting different properties, depending on the number of observers or on the respective conditions under which the objects are observed. For the time being it does not matter what kind the phenomena are special: It can be both qualities are about objects, as in the example of the rod broken in water or the elliptical-looking coin, as well as the existence of things, as in the case of the mirage.

In the course of his argument, Ayer favors the example of the stick immersed in water: We hold a stick in our hands that we would not refuse to have the `straight` property. Now we dip it in water - and we are confronted with the following problem: the stick looks, as would he broken, but we knowledge it is pretty accurate that it was not harmed on the way from the air to the water, and the water itself was not in any way aggressive towards the staff. Also, when we feel the stick under water with our hand, so feel we that it still has a proud straight shape. So while we assume with certainty that the staff is straightso see we have him as broken Rod. “Then it follows that at least one of the visual appearances of the stick is delusive; for it cannot be both crooked and straight. " (FEK, p. 4).

By turning one of the two appearances into a delusion to explain, we have cleared a contradiction out of the way, but we still have to admit that we are still anything perceive. This anythingthat neither a Nothing still something real could be called, was baptized with the name "date of the senses". Sensory data are extremely practical in their use (and how could they not be, since they were introduced ad hoc): One perceives sensory data, “which are similar in character to what he would be experiencing if he were seeing a real oasis , but are delusive in the sense that the material thing which they appear to present is not actually there. "(FEK, p. 4).

We are still intact and not forced to take measures to defend our dusty and proven fortress, which is called everyday language and 'ensures' the existence of material things. Because we have no reason to assume that there would be no cases at all in which we perceive material things. However, there are philosophers who do exactly that the do: after them we take always only Sensory data, No way on the other hand material things are true. Your argument sounds very plausible prima facie: namely, they say there is none intrinsic difference between the deceptive and the veridic perceptions. When I perceive a stick, "which is refracted in water and so appears crooked, my experience is qualitatively the same as if I were looking at a stick that really was crooked." (FEK, p. 6). Disregarding the fact that the conditions both types of perception are different and that we often knowledgeThat we are dealing with a deception can of course be said that we cannot distinguish between the deceptive and the veridic perceptions. But even if we are able to separate the two cases, the reason for this separation does not lie in the quality of the perceptions themselves.

Ayer formulates the argument from illusion in such a way that the impression is created that he himself is in line with the radical view that material things are imperceptible for the reasons mentioned, but he objects that the statements just made require further examination .

2.2. A fundamental mistake and its correction

2.2.1. Criticism of the factual view

First and foremost, according to Ayer, it is about the following question: When we worry about whether we perceive sensory data or material things, then we try something Factual or something Linguistic to manage? It can be said in advance that Ayer's heart does not exactly beat for those philosophers who give their place to the first variant. They concluded from the argument from illusion, “not merely that it is linguistically inconvenient, but that it is false to say that we are ever directly aware of a material thing.” (FEK, p. 11f.).

But what does it actually mean to regard a certain thing as either factual or linguistic? Well, in the first case, the respective apologist believes that the sentences uttered are referring to something 'in the world' that could be empirically proven. With this, of course, there is the possibility of showing that what we have so far unsuspectingly put up for the best does not correspond to the facts. On the other hand, anyone who understands something as a linguistic problem has neither interest nor an eye for the terribly common things around; at most those are of importance sentencesin which we `grasp` these things. The same cannot be proven or refuted either, because the facts are not even debated. What is brought up is that How to use of sentences; a corresponding analysis is only intended to emphasize the conditions of the sentences concerned, which may eventually lead to a new description of the sentences concerned. In short, the factual viewer embeds what he is talking about in causal Contexts that the advocate of the linguistic view tries against (conceptual) logical Establish connections.

Back to the philosophers mentioned above who understand the argument from illusion as an argument based on facts. According to them, the sentences of the argument should refer to something that can be discovered in the world; the expressions occurring in the sentences should actually existing entities describe. O b they actually exist is still open, but there is at least a claim to it; nor has their nature been clarified for the time being. The proponents of this view take, for example, the conclusion of the argument that we never perceive material things, but always only sensory data, at face value: Regardless of whether we fundamentally deny the independent existence of material things or admit them, our own perception does not play a role in this question, since the only thing that we can ever see, feel and taste in real life are individual things called sensory data that can be found in space and time.

Here again the steps of the argument from illusion in brief: Firstly, deceptive and veridical perceptions are qualitatively indistinguishable, secondly in the case of deceptive perceptions we believe (sometimes) that we perceive material things or real qualities, but this is not true. Thirdly, since we perceive mere sensory data in the case of a deception, it follows from all three premises that even in the case of veridic perception we are only dealing with sensory data and never with material things.3

At first glance, it doesn't seem like we can crack this flawless ball of iron logic anywhere. The argument is undoubtedly conclusive. Or maybe not? Ayer doesn't find the argument very convincing. He thinks that the conclusion on the part of the 'fact philosophers' can only be drawn under certain conditions; and there are these Conditions he is unwilling to accept. The difference between Ayer's conditions, which forbid him to draw the above conclusion, and those which the other side demands, can only be made by conceptual analysis Be Kind: Off propositional That is because there is nothing wrong with the argument. However, we walk into a propositional problem as soon as we refine the internal structure of the propositions through an analysis of the concepts.

Ayer doubts the correct understanding of the terms used in the argument: his analysis of the terms yields sentences different from what the fact philosophers would claim as the result of their conceptual analysis.

In the line of argument just outlined, the first premise cited was the qualitative indistinguishability between veridic and deceptive perceptions. It is only when one misconstrues this premise that one arrives at conclusions such as that we never material things and always only Perceive sensory data.This wrong conclusion in turn presupposes a mistake, namely qualitative Indifference as a lack of difference Art to interpret. Such an interpretation is no coincidence; certain ones are required for this Auxiliary argumentsas Ayer calls them. The 'proof' for the generic Indistinctness is achieved in an indirect way: the opposite of what is to be proven is assumed, and then unacceptable or supposedly unacceptable consequences (regardless of whether logical or factual) are drawn from these assumptions.

The following line of thought tries to convince us of the claim in question: Would be first, both types of perception perceptions generic different objects, these objects should be based on their qualities differentiate; however, this already contradicts the first and impeccable premise of the argument from illusion. Second, in the case of generic diversity, it would not be possible to form a continuous series with them, but this is actually possible4. Third, material things should be in causal independence exist by observers and have certain qualities; Not all fact-philosophers would agree with this assertion (for example, not the indirect realists, since they postulate causal independence), but it can still be seen as a weighty argument in favor of factual interpretation.

It seems, then, that it cannot be asserted without self-contradiction that the objects of veridic and deceptive perception are each of a different kind. For Ayer, on the other hand, the assumptions mentioned can be rejected without contradiction; i.e. we are able to show that the assertion that the objects of perception are of different kinds in both cases, Not implies the sentences put forward by the fact philosophers and thus none of the dreaded self-contradictions arise. It shows quite conclusively that the factual view refutes itself: it assumes certain things in order to ultimately conclude their negation with the aid of these premises. The focus is not on the first two implications assumed by the opposing side (on the grounds of Ayers that they are incapable of any empirical evidence), but the third.

The factual philosophers accused us that our assumption under discussion, namely that we must regard the objects as generically different in both cases of perception, entails the claim that material things exist with their own qualities independent of any observer. In order to refute our infamous initial assertion, they negate this statement; they claim to say something about material things is no more than hypothetical sentences to say about our sensory experience that we have in certain situations. Ayer notes at this point that we should realize the fact that we are only going through such hypothetical judgments induction come, with which he wants to promote the assertion of the independent existence of material things, but at the same time he limits his own argumentation: “Though this is perhaps the most natural way to interpret this proposition, and the one that I shall adopt myself, it is not the interpretation that is required by the argument from illusion. "(FEK, p. 12f.). For such sentences can also be true of things that we currently perceive; thus it is inadmissible to infer from the postulation of independently existing material things as hypothetical propositions about our sensory experience, which are supposed to be both true and unfulfilled, that we are in the case of actual sensory perception no perceive material things. But if that logical it follows from the definition of material things in the type mentioned, need we no longer have any empirical evidence, but by definition the fact philosopher has to look for such evidence. Thus, the factual view has itself for the first time superfluous made. For the second time it must yield to its fate if we consider the consequences that should follow from the definition of independent material things: If they are to be independent of us and we can never perceive them in the way we are familiar with, then we must not call the things around us material things. This leads us into an absurdity, however, in that the argument from illusion, which is both our basis and that of the factual philosophers, always comes from such things went out is. Anyone who wants to formulate a factual view on the basis of the argument from illusion grinds the assumption of material things through the entire argument in order to, among other things, dispute or refute the existence of material things with the help of this assumption: Such an argument but in propositional logic, explicitly means nothing other than a simple logical contradiction. But this should not cause us to negate the initial assumption, namely that there is such a thing as material things. What needs to be changed is the way in which the argument from illusion is viewed.

According to Ayer, none of the implications mentioned at the beginning of the chapter follow logical from the assertion that the objects of our perception are of different kinds in both cases. First and second, do not follow because that would presuppose that we could distinguish veridic and deceptive perceptions from the outset, which cannot be correct, since otherwise we would not either term of deception (because this makes us believe, we're taking something Real thing true). Likewise, thirdly, it does not follow from the assertion, although Ayer himself does not reject the assertion that there are material things. However, he defines it independently: "It is in every instance a matter of our being able to establish a certain order among our experiences." (FEK, p. 16). To say that the existence of material things follows solely from the statement that we have different objects in front of us in both cases of perception requires that we consider it a self-contradiction to ascribe different properties to a thing. "But here it may be objected that these contradictions cannot, in fact, be derived from the nature of our perceptions alone." (FEK, p. 14). Contradictions only arise when we want to understand the respective object as a thing that always remains the same in its properties. We don't usually do that; we can very well agree that one and the same thing looks different to different observers, or to the same observer at different times. It is the introduction of such implicit parameters that are able to resolve the contradictions.

2.2.2. The linguistic perception

But what does it actually mean to interpret the argument from illusion and the associated sense-data language not, as Ayer refuses, factually, but linguistically? An answer to this requires the prior elimination of misunderstandings.

Given the argument from illusion, the question arises whether we shouldn't subject our everyday language to a general maintenance; for, as we have seen, we come up against certain limits with it when it comes to distinguishing between veridic and deceptive perceptions. In both cases we are talking about 'perceiving' something, although we mean something completely different in each case, i.e. the conditions are completely different in both cases. And it seems that this problem can only be curbed by introducing a new concept, namely that of the sense datum.

If we stick to our usual conception of the word 'perceive': if 'perceive' implies that the perceived object always also exists or is endowed with this or that property, we consequently have to deny that it is deceptions there is, or to call it a mistake, to say that the perceived objects are in most cases material things or that they mostly have the claimed properties. In fact, and fortunately, we do not fail too often because of the stubbornness of our language; she even behaves very benevolently. "Thus, to return to the familiar examples, if I say that I am seeing a stick which looks crooked, I do not imply that anything really is crooked, or if [...] I say that I am perceiving two pieces of paper , I need not be implying that there really are two pieces of paper there. " (FEK, p. 20f.). But that seems to be just a play on words that requires a distraction, because we can still be held responsible for the fact that, despite everything, we have perceived something that must exist in some sense.

The mistake, according to Ayer, is to assume that the word 'perceive' has an existential implication. However, I can consistently claim that I saw two pieces of paper without implying that two pieces of paper really existed. However, should we decide to equip the word 'perceive' with an existential implication, we will of course get into disagreements when we speak of 'perceive' in the case of deceptions. Therefore it is then only possible for us to say, for example, we would have thought that ß we saw two pieces of paper. To ask at this point what we would have actually perceived is to assume that we are asking for verifiable facts. There is nothing we can check, however; the only thing that could be raised to the rank of a subject of dispute are the various ways of describing facts about which we, for our part, are fundamentally in agreement and about which we do not know how to argue about them.

Whether or not we get caught up in contradictions depends entirely on our determination of the meanings we attach to certain terms. If perception is supposed to imply existence, it is not correct to say that I have perceived two pieces of paper; I either saw a piece of paper or, in the case of a complete hallucination, no piece of paper at all. "I was indeed having an experience that could properly be described as perceptual in one sense of the word." (FEK, p. 22). However, this type of perception cannot be described in terms of perception, if these are provided with the existence implication. In these cases, we usually use the term `appearing` to distinguish it from the term` being` (`is`), both when it comes to the qualities of an object even when it comes to its existence. When a thing 'appears' to us one way or another, then we say that we perceive something, but not that such and such a property or even the object itself exist; we can even say that we 'see' something without wanting to imply something 'real': it is' not ordinarily a source of confusion to us, because we are able to tell from the context what is the sense in which such words are intended to be understood [emphasis added by me]. ”(FEK, p. 24).

Since we are not always able to name the circumstances precisely and accordingly differentiate between veridic and deceptive perceptions, the concept of the sense datum was introduced. At the same time, this also means, as we have seen in the argument from illusion, that we perceive sensory data in both cases of perception, since they are qualitatively indistinguishable. The assertion that in both cases we perceive sensory data and not material things is sometimes very useful and can be described as acceptable as long as we do not actually understand it (see Chapter 2.2.1.).

In the case of a factual apprehension of the conclusion drawn, it should be possible to observe effects in such a way that certain of our previously indisputable statements suddenly turn out to be not correct turn out. In other words, the fact-philosopher would have to show us what he will not be able to do, precisely because we are not dealing here with a dispute over empirically verifiable facts, but only with the use of our language and an analysis of the same. According to Ayer, the sensory data language simply fulfills a specific purpose: “it is useful for us to have a terminology that enables us to refer to the contents of our experiences independently of the material things that they are taken to present. And this the sensatum language provides. ”(FEK, p. 26).

2.3. Comments on the problem of sensory data

2.3.1. The concept of sensory data

In philosophy there is no clarity even about the basic concept of the sense datum. The different ways of understanding him to lead however, not directly into a certain kind of perception theory, but vice versa, the respective conception of the concept of sensory data is nothing else than Expression some understanding of perception.

This is also the case with Ayer, and he gives the following definition for the term `sensory date`:" A. is sensing a sense date s, which really has the quality x, and which belongs to M.. "(FEK, p. 58). Depending on whether we give the word 'perceive' an existence implication or not, we may have to add a restriction to the definition: If no existence of the material thing or property (s) is to be implied, we say that A. assumes s would to M. belong, which leaves the possibility that M. does not exist or does not exist with these properties. Sentences in which I communicate my sensory experience are therefore at least sometimes true, even if the objects or properties are objects of a deceptive perception.

There are now philosophers who give the concept of the sense datum a completely different meaning: They regard sensory datum as something that the observer owns right true becomes. At first glance, such a definition does not seem problematic, but there is actually a catch here. Because not only should we designate the things of which we are aware as sensory data, we should rather any Call an object of which we can become aware a sense datum. But this has the following unpleasant consequence: "There is no accepted meaning of the expression` direct awareness` by reference to which it can be made clear without further explanation what is to be meant by the word `sense-datum`." (FEK, p. 60). The problem is: circular reasoning. We cannot define the concept of sensory datum using a concept, namely that of direct awareness, since this is itself defined with the help of the concept of sensory datum.

Ayer therefore proposes to understand direct awareness in such a way that the belief in the existence of the respective object is based on sensory experience is based and it does not come about because we are referring to the object from sensory data shut down. That means, of course, that we material things are directly aware.

The other side interjects here that we can only perceive sensory data and never material things, because we can always be wrong in our perception. But this argumentation presupposes that direct awareness is understood in such a way that it implies the existence of a material thing or the asserted properties. And in order to avoid the accusation that the existence of non-existent things or properties was postulated, they introduced the concept of sensory data which exactly fulfills the desired condition, namely admitting the existence of 'something' without adding it to the rank of ' Reality '. "That is to say, the expressions ´direct awareness` and ´sense-datum` are to be regarded as correlative [...], it is not satisfactory merely to define one in terms of the other." (FEK, P. 61). Neither the one nor the other term can therefore be defined by the respective other term; both require an independent way of understanding.

The sense and purpose of the concept of sensory data was to be able to express perceptions without having to take on the burden of the implication of existence; thereby we are able both to say that we have perceived something and that this perception was deceptive. We must not apply the distinction between 'appearance' and 'reality' to the sense data themselves. In the event that we do, "the terminology of sense-data becomes superfluous." (FEK, p. 69). For the sensory data language was introduced in order to be able to ignore the difference between reality and appearance; it is fundamentally different Art as the language in which we actually refer to material things. That means in consequence: It is one sufficient Condition for the existence of a sense datum is that it is just perceived, because the question does not arise whether we are dealing here with a veridical or deceptive perception.But is it the other way around, that if we assert the existence of a sense datum, we would have to infer from it that we are actually perceiving it? Then, however, we would no longer have the opportunity to go beyond possible To speak sensory data. However, according to the other side, we lost the ability to analyze propositions about material things, since we are not only adding reports about our current perception here, but also hypothetical Finding mention of sentences. However, according to Ayer, such an approach is difficult to apply to sensory data. We just wanted to Not define the existence of sensory data using the same criteria as material things. Instead of several hypothetical sentences, in the case of the existence of sensory data, we can only specify a single hypothetical sentence as the criterion of existence; "And in that case we shall be able to express no more by asserting that the sensibile exists than we are already able to express by asserting the hypothetical proposition in question." (FEK, p. 71). If the existence of the sense datum follows from the definite hypothetical proposition (as a criterion of existence), and if existence is not supposed to claim anything beyond this hypothetical proposition, then it can first be concluded that, conversely, from the existence of the sense datum this a hypothetical sentence follows. Since nothing else can follow from the hypothetical proposition other than the current perception of a sense datum, the principle of transitivity means that the existence of a sensory datum implies the current perception. And that in turn means that this one necessary The condition that represents: Current perception and the existence of a sense date therefore mutually imply one another. Ayer suggests that we should therefore speak of their existence only with reference to just perceived sensory data, not against it when we speak of possible sensory data.

A brief consideration is the question of whether we can speak of 'knowledge' with regard to the sensory data. If, according to Ayer, the object of our knowledge is to exist independently of us, then sense data are nothing knowable, "for I have made it a necessary and sufficient condition of the existence of sense-data that they should in fact be experienced." (FEK, p. 78f.). However, he objects on the spot that the independence of the object to be known is not a necessary condition of knowledge: for we apply the concept of knowledge both to the cases of propositional truth, which actually presuppose the independence in question, and to it acquaintance with individual things that do not imply this independence. The woman now asks herself whether we can equate such an acquaintance with knowledge: to make acquaintance with a sense datum means the direct awareness of a sense datum; But since knowledge relates to propositions, but awareness itself does not represent a proposition, the expression 'knowledge' cannot be applied to the awareness of sensory data. The relationship between the awareness of sensory data on the one hand and propositions or knowledge on the other hand is a topic that has not yet been dealt with at this point.

2.3.2. Apparitions and propositions

Let us look again at the argument from illusion: The conclusion that we perceive sensory data in both veridic and deceptive perception was drawn on the basis of an obvious contradiction. This arises from the assertion of qualitatively identical perceptions in both cases, while at the same time it is established that only in the case of veridic perception is there actually a material thing. The concept of sensory data had to be introduced here, since nothing in our experience or perception can resolve the contradiction; the contradiction came through a description comes about and can apparently only be removed again by such.

However, there are philosophers who apply the concept of contradiction to appearances self apply: Contradictions are not considered to be a linguistic problem, but are given a factual status. With the result, “that our ideas of secondary qualities are not resemblances of any real qualities of material things” (FEK, p. 28). If one and the same thing is supposed to have, or apparently actually has, contradicting qualities, then these qualities must not be associated with the respective material things really get.

Ayer wonders, however, whether we are entitled to make such an assertion. Appearances are not one of the kinds of things that could contradict one another. You can unite what Ayer is criticizing here Category error call. But it only comes about when we talk about the phenomena in fact Understand: A stick is still straight now, but the next moment, when it is immersed in the water, it looks broken. The dilemma that the stick should now have two irreconcilable properties is now attempted by declaring these properties to be independently existing entities that contradict each other, and then denying them to the material thing, the stick.

According to Ayer, the proponents of this thesis commit the mistake of granting appearances the possibility of being able to contradict one another. But - "they simply occur." (FEK, p. 29). Contradictions do not arise until appearances are used as partial expressions in propositions, but it is not even now a necessary condition that we are faced with contradictions. The other side may then reply that the reason that we do not have to get into contradictions is that we do not treat the phenomena as realities, as we should; if we did so, they think there would be contradictions, since, as Bradley says, a thing must be consistent in itself and therefore cannot have contradicting properties. Ayer also rejects this objection: A language that makes no distinction between appearances and realities does not necessarily imply contradictions, "provided that the language also contained suitable criteria, which would, of course, be different from the criteria that we now employ, for determining when a thing changed its qualities and when two appearances were appearances of the same thing. " (FEK, p. 31).

In addition, the conception of appearances as mere appearances does not protect us at all from self-contradictions: an object without a mention of specific ones appropriate circumstances Ascribing both veridic and deceptive properties in the form of appearances puts us in an awkward position, as does the treatment of qualities as realities, since we attribute them to a thing indiscriminately to speak. Ayer does not understand properties in such a way: "What we actually do is to de- fine the real qualities of a material thing in terms of the qualities of certain privileged ap- pearances." (FEK, p. 31). They are privileged insofar as we ascribe to them the status of objects of veridic perception, which is because, according to Ayer, they are under certain Standard conditions happen.

Bradley had, as was briefly mentioned, requested that the definition of material things include consistency in their properties. However, since the facts are bogus and one and the same thing presents itself to us one and the other differently, depending on the circumstances of the situation and the observers, he concluded that these properties do not really belong to the material thing. Ayer clearly rejects this argument: for him the properties of a thing can be defined in terms of the appearances, whereby the thing itself does not represent anything separate from its appearances; the question, how these appearances do not appear here at all. Bradley wants to know that the coming about is causally understood, depending on the external conditions and the states of the observer. A kind of dependency can even be admitted without our being compelled to say that the properties of a thing are not its real properties: Certain conditions must be fulfilled in order to be able to perceive certain phenomena. That is why we are not yet justified in claiming that these properties would then not be real; this also applies if the conditions in question are not met at the moment and the respective phenomena cannot be observed. Because "all that this involves is that the hypothetical propositions, which assert that the appearances would be manifested if the conditions were fulfilled, remain true even when their protases happen not to be realized." (FEK, p. 35f.).

In addition, a much more fundamental question arises: To what extent is it possible for us to have 'knowledge of sensory data'? In the previous chapter the difference between knowledge and acquaintance was briefly mentioned: only propositions are knowable; Objects of our acquaintance in the sense of direct awareness can only be sensory data. But sense data are logically primarily compared to the propositions: "Whenever we are directly aware of a sense-datum, it follows that we know some proposition which describes the sense-datum to be true." (FEK, p. 80). Ayer now declares these empirical propositions, which relate to sensory data, to be 'indubitable'. Sentences, however, cannot be doubted in various ways, and the question arises what special sense Ayer had in mind here: There is on the one hand analytical statementswho are beyond doubt because the assertion of their opposite would generate self-contradiction; this case does not apply to the empirical propositions, since it is possible to assert a statement or its negation, depending on the circumstances. Ayer also rejects that psychological Interpretation of the indubitable, since this would then also have to apply to propositions about material things, from which propositions about sensory data are supposed to be different. What is meant here is that logical The unquestionability of such sentences: “We imply, not merely that our belief in the truth of such proposition is not, in fact, mistaken, but that it could not conceivably be mistaken. And it is this view that is sometimes expressed by the assertion that such propositions are `indubitable´ or` incorrigible´. " (FEK, p. 80). If I can make a mistake here at all, it is this mistake more verbal Kind and as such refers to the Kind of description of the respective sense datum, which comes about as a misdescription because I use an expression that I consider to be applicable to this or that sense datum, which is illegitimate, contrary to the rules of the respective language. I can also make mistakes in relation to sentences about material things, namely when I have accepted certain additional sensory data which, however, were not fulfilled in the following. An uncorrectable proposition, on the other hand, “is completely verified by the existence of the sense-datum it describes; and so it is inferred that to doubt the truth of such a proposition is not merely irrational but meaningless; for it is only significant to doubt where there is a logical possibility of error [emphasis mine]. " (FEK, p. 83). In order to establish the truth of this proposition, no further, hypothetical statements about sensory data are required; furthermore, they themselves do not imply any further sentences. What has just been said even applies, although it is of course a factual matter whether a sentence about a sense datum is true or false. What is to be disputed is that this decision can be made in the same way as in cases where we have to decide about the existence of a material thing.

2.4. Ayer's Critique of Phenomenalism

It is the phenomenalists, among others, who advocate a factual interpretation of the argument from illusion. On the one hand there is Sensory data phenomenaliststhat although material things for knowable keep this, however, with the sensory data identify: "Our only genuine, certain knowledge of perceptibles is restricted to what directly appears to us and would be as it is even if we should be hal- lucinating." (Audi: 1998, p. 42). The reason for this claim seems to be that we usually think of material things as entities lying 'outside' the sensory data, but a simple train of thought tries to prove us wrong here: “If you imagine subtracting the book's sensory properties one by one - its color, shape, weight, and so on - what is left on it? [...] It is like stripping layer after layer from an onion until nothing remains. " (Audi: 1998, p. 42).

The second, albeit less common, variety is adverbial phenomenalismwho interprets the perception of material things as having a certain experience; we don't even need sensory data, just a 'perceiver' and the way in which he perceives something in order to be able to speak of sensory experience.

But let's stick with the former type of phenomenalism. Material things and a respective set of sensory data sets are for him, as I said, identical. Ayer admits that there are both material things and sensory data, but denies that we are looking for a kind inner relationship can search between these two object classes. Ultimately, phenomenalism carries out no other procedure than this, even if it ultimately denies the existence of material things and sets the individual expressions that are supposed to stand for material things to be synonymous with sensory data.

Against this equation, Ayer says: "The term` material thing` is not synonymous with any term or set of terms that stand for species of sense-data. " (FEK, p. 229). The relationship 'perception of sensory data' and 'perception of a material thing' runs subjunctionally in only one direction, namely from the perception of a material thing to the perception of sensory data: whenever we say that we perceive a material thing, we imply that we also perceive sensory data, but not the other way around. The reason for this is the qualitative indistinguishability between veridic and deceptive perceptions cited in the argument from illusion; and a lack of difference in the qualities means a lack of difference in the perceived sensory data, which do not give us any information about whether we are dealing with an illusion or not. No matter how many sensory data that appear to be certain can be given: “Whatever the strength of this evidence may be, it will always be logically compatible with the hypo-thesis that this material thing is not in all respects what it appears to be, or even that it does not exist at all. " (FEK, p. 230). Not only is deception still possible, our evidence regarding the material things in question will never focus on anything other than further sensory data refer, since we are not allowed to draw a conclusion from the observable sense data to the principally unobservable material things: "All that the evidence in question will be evicence for or against is the possible occurrence of further sense-data still." (FEK, p. 231). However, when we talk about material things, such sentences must always be translated into sentences about sensory data, otherwise our speech does not make any sense. And from here it is concluded that material things are nothing else be can be used as collections of current and possible sensory data.

Objections have been made against this phenomenalist type of analysis, which, however, must be regarded as untenable because they are based on the same and incorrect basic assumption as phenomenalism. The mistake is to understand the speech about material things and about sense data as a factual speech, ie the raised objections “are founded upon the mistaken assumption that a material thing is supposed to consist of sense-data, as a patchwork quilt consists of different colored pieces of silk. ”(FEK, p. 232). It is feared that the reduction of material things to a set of sensory data represents an unacceptable robbery of their substantiality and unity, which must be prevented.

But both sides, the phenomenalists and their critics, allow themselves to be carried away into accepting a momentous condition. Ayer also includes the distinction between material things and sensory data in his philosophical repertoire, "but the purpose of making this distinction was simply to increase the utility and clarity of the sense-datum language by ensuring that its sentences should not be of the same logical form as those that refer to material things. " (FEK, S, 232f.).To say that both types of sentences, that is, those that relate to material things and those that relate to sensory data, have the same logical form, is to say that they are able to replace one another. That may still be bearable as long as it is the mere designation is about objects, but becomes a problem as soon as we want to demonstrate the interchangeability using practical and everyday sentences: “To say, for example, that this was being written with a` pennish` group of sense-data, instead of saying that it was being written with a pen, would be neither true nor false but nonsensical. "(FEK, p. 233).

The attempt of some philosophers to proclaim interchangeability and to regard material things as mere 'logical constructions' has therefore failed; a sentence about a material thing cannot be translated into a finite set of sensory data sentences, without to experience a change in meaning. And again the other side unabashedly steps into the footsteps of the phenomenalists, who lead them astray just as they do: according to the factual form, it denies the phenomenalist claim that we only perceive sensory data, countering this by saying that we very much material Perceiving things, of course without denying that in the case of deceptive perception we only perceive sensory data. Both sides wrongly assume that they are talking about something that can be proven or refuted; in fact, however, they differ only in the way they speak with regard to sensory perception.

However, if we follow Ayer, the real problem is another one. Let's look back at the argument from illusion: It was formulated because there are cases in which we do something, namely to perceive a sense datum, but not a material thing. That is, if we make a sentence in which we name the object of our perception in the form of sensory data, it does not follow that we really have a material thing in front of us, even if that is possible. Conversely, it follows from the assertion that we perceive a material thing that we also perceive sensory data. It would therefore be presumptuous and reckless to simply replace sense-data expressions in certain sentences with expressions that designate material things.

In addition, there is a second problem: Whoever claims that sentences about material things can be replaced by sentences about sensory data, who therefore has one between them Equivalence asserted, must show that both types of sentences mutually imply each other. But that, according to Ayer, is precisely not possible. "For when we try to reproduce the content of a statement about a material thing by specifying the empirical situations that would furnish us with direct tests of its validity, we find that the number of these possible tests is infinite." (FEK, p. 239). So should someone try to verify the existence of a material thing on the basis of his current sensory experience, he can indeed form a chain of sensory data sets and extend it at will through further sensory data perceptions; but the hope of reaching a point at some point where the possibility of an opposing sensory perception that would bring down all previously provided verification sentences is ruled out - this hope is deceptive. "He will never be in a position to demonstrate that he will not subsequently have experiences that will entitle him to conclude that his original statement was false after all." (FEK, p. 239f.).

No matter how long a string of sensory data sets is sufficient to legitimately imply the existence of a material thing. Conversely, a sentence about a material thing does not result in a finite set of sensory data sentences. However, it does not follow either absolutely no Sentence about sensory data; if so, we would not have to struggle to distinguish between delusions and veridic perceptions. "But there is no set of statements about the oc- currence of particular sense-data of which it can truly be said that precisely this is entailed by a given statement about a material thing." (FEK, p. 240). The presence of a material thing implies a lot different Sensory data sets, depending on the circumstances in which it is perceived; their number is ultimately infinite.

From all this, the conclusion based on a factual conception must not be drawn that to talk about material things would mean to talk about something 'other' than about sensory data, or exclusively about sensory data and further 'nothing else'. According to Ayer, it is still possible to analyze material things in terms of sensory data; What Not should be tried is, "to seek in any such analysis a means of distinguishing one material thing from another." (FEK, p. 242).

3. Conclusion

Ayer is sitting between the chairs. He approaches certain positions and makes some concessions to each, but is far from being absorbed by them. What is striking about his argument is the lively discussion of possible objections from the other side and, in general, his preoccupation with opposing views. It seems like he's working on them to find his own point of view. For the reader, his view only gradually emerges; Ayer occasionally makes strange remarks, which, however, can always be consistently woven into his option.

What he shares with the theories of sensory data is the conviction that talking about sensory data makes sense. According to Ayer, 'sensory data' also solve the problem of how we can still speak of perception in the case of deceptive perception. But he does not take the decisive step of the sensory data theorists of restricting perception to mere sensory data in the case of veridic perception as well. Rather, he re-coins the concept of the sense datum and declares it to be an aid in the analysis of sentences that relate to material things. That is, he in no way denies that we perceive material things. He only denies that from the analysis of propositions about material things in the form of sentences about sensory data it follows that we do not perceive these material things. Ayer constantly insists on the distinction between the linguistic and the factual conception of the argument from illusion, which was the occasion to introduce a ´Sense data language`. A factual view can only lead to either denying that we perceive material things or to question the existence of material things at all. In addition, this position produces problems regarding the differentiation between veridic and deceptive perception, since both forms of perception are adjusted to one another and we have another problem of differentiation. If, on the other hand, we take the argument from illusion and the sense data language linguistically, then, according to Ayer, we can both recognize the existence of material things and at the same time solve the problems with regard to the distinction between deceptive and veridic perception. Since Ayer speaks of standard conditions in the identification of material things in several places, we cannot even solve them without this recognition of material things.


Audi, Robert. Epistemology. A Contemporary Introduction to the Theory of Knowledge. New York 1998

Ayer, Alfred J. The Foundations of Empirical Knowledge. London 1964


1 Ayer, Alfred J. The Foundations of Empirical Knowledge. London 1964, p. 2. (The passages cited below are taken from The Foundations of Empirical Knowledge indicated by paragraphs in brackets in the text; The Foundations of Empirical Knowledge is abbreviated by "FEK", followed by the respective page number.)

2 This term may sound a bit artificial; this is because he is a more or less arbitrary translation of the English term perceptual realism represents that in Audi, Robert. Epistemology. A Contemporary Introduction to the Theory of Knowledge. New York 1998, p. 28 and was not known to me.

3 A more detailed and informative presentation of the `arguments from hallucination` can be found in Audi 1998, p. 31ff.

4 This second counter-argument basically represents an instantiation of the first. See FEK, p. 8f. 7th