CHAPTER IV NON-EMPIRICAL CRITIQUE OF A PRIORI AND A POSTERIORI Introduction Empiricism, both in its classical and modern forms, gives importance to sense- experience. What is not obtained by senseexperience is treated as non-empirical. If this view is extended to the subject- matter of the disciplines such as religion, ethics, aesthetics, logic and mathematics, all these disciplines turn out to be non-sensual for their respective subject- matter is not derived from sense-experience. If logic and mathematics are equated with religion, ethics and aesthetics, then the entire exercise of empiricists to build up scientific knowledge collapses. Thus an exception is granted to the statements of logic and mathematics. Unlike the statement of religion, ethics and aesthetics, the statements of logic and mathematics are a priori in the sense that they are grounded in precise formulae. This view of empiricists is attacked and an attempt is made to show how Waismann attacks the distinction between a priori and a posteriori forms of knowledge.
Waismann s Critique of the Distinction Fredrich Waismann doubts the whole philosophy of empiricism as too dogmatic and in particular he dislikes the attempt made by empiricists to construe everyday statements on the model of the precise formula of mathematics and the advanced science and hence he wants to make a philosophy which has flexible terms and concepts with itself. Such a feeling makes him to reject the analytic (a priori) and synthetic (a posteriori) distinction as it is fundamentally grounded in linguistic concepts, which are vague as well as strictly rule bounded concepts. Waismann in his famous series of essays entitled Analyticsynthetic, analyses the application of analytic and synthetic terms in our day-to-day affairs through which he interprets the meaning arrived at different levels. These interpretations on the application of analytic and synthetic terms later made him to reject the distinction. In order to understand the reason why he rejects this distinction, one has to know his philosophical position. As it is generally claimed that analytic statements are grounded in meaning or a statement whose truth follows from the very meaning of their terms, which ultimately implies that it is part of its definition; it gives the impression that they can be precisely defined and hence they can be eliminated from other kind of statements. Waismann argues that if analytic statements 86
are taken as a part of a definition, then they must be taken as obscure and incomplete concepts as no one can define anything through any means such as ostensive, verbal, dictionary, etc. Waismann s concept of definition, though elongated, plays the central core of his philosophy. He proceeds from the Kantian approach of analytic-synthetic distinction. Kant held that in any analytic judgment the predicate term is contained in the subject term such that no new information is given to us. Kantian approach has been criticized by many philosophers including Waismann. The common view is that the concept of containment needs further clarification as it is obscure in relation of its constituent terms. Further, this definition of Kant cannot be applied to other type of statements that do not possess subject/s and predicate/s such as relational and existential statements. For all these reasons Waismann declared that Kantian approach basically uses only metaphorical terms and nothing else. Waismann gives a formal account on analytic statement. According to him, a statement can be said to be analytic if it can be turned into a truth of logic by means of mere definitions. 1 Obviously, this definition inherits the notion of definition. Waismann says if the concept of definition has no boundary, then analyticity cannot be taken into account. So, he rejects the analytic/synthetic distinction. 87
Waismann proceeds step by step in order to reject this distinction. From the given expression for analyticity, it is clear that an analytic statement is transferable into truth of logic. But transformation is a process wherein one cannot find out the line that separates the thing that has not been transformed and the thing that gets transformed. If it is so, how can one say that this point or that point is the logical form? One has to familiarize with the process of transformation in order to know the logical truth of the statement. The transformation of an analytic statement (p), All planets move round the sun into a logical truth as mentioned by Waismann is as follows: 2 Given (P) = All Planets move round the Sun - analytic statement. The first step is to translate the All idiom into the idiom there is no such thing that not. Let this transformation be T. Applying T to the sentence P in the way of an operator, we obtain, TP = There is no planet that does not move round the sun. The next step is to put the term planet in the place of a predicate, Let L symbolize this process and write the whole as LTP. 88
LTP = There is no thing such that it is a planet that does not move round the sun. Next step is to eliminate the last restrictive clause and put it conjunctively (operation R). RLTP = There is no thing such that it is a planet and that it does not move round the sun. Next step is to apply the principle of double negation (N). NRLTP = There is no thing such that it does not have the following property: It is a planet, and it does not move round the sun. Now the translate the idiom there is no such thing that not back into the All, or whatever idiom (transformation T-1) T-1NRLTP= whatever a thing maybe, it has not the following property; it is a planet and it does not move round the sun. If DeMorgan (M) rule is applied, we get, MT-1 NRLTP = whatever a thing may be, it has the following property; it is not a planet, or it is not the case that it does not move round the sun. Next apply N-1 (the converse of N). N-1 MT-1 NRLTP = whatever a thing may be, it has the following property; it is not a planet, or it moves round the sun. 89
Now the denial of if something is a planet, it moves round the sun is something is a planet, and it does not move round the sun, ; as the denial of the latter, according to M and N-1, is something is not a planet, or it does move round the sun, and as the two denials cancel (N-1), this must come to the same as the first sentence; which show that the if idiom can be transformed into the not and or idiom.(i). In ~1MT~1NRLTP = whatever a thing may be, it has the following property, if it is a planet, it does not move round the sun. Let finally D stand for the definition of planet, applying D as a sort of operator, we get, DIN~1MT~1NRLTp = whatever a thing may be it has the following property: if it is a heavenly body that moves round the sun, then it moves round the sun. Repeating the R step yields: RDIN~1MT~1NRLTP = whatever a thing may be, it has the following property: if it is a heavenly body and if it moves around the sun, then it moves round the sun. Waismann s contention is to equate the last sentence with the definite formula in phrase-marker (PM), i.e., whatever has the 90
property and the property has the property (tautology). This shows that we have arrived at the statement that is logically true. The truth that has been revealed through the above said transformation can be summarized in the following manner. On the whole there are many individual tools. R,D,I,N,M,T,L,R, T~1,N~1 were employed in the process of transformation. These tools can be broadly classified into two divisions, namely, definition and operators. Though the tools are different, the function is the same - to transform an analytic statement into a truth of logic. These two types of tools are not only closely connected with each other, but also connected with the conception of equivalence. Such a connection can be seen on the considerations of any equivalence of the form: P = P (N) Here transition of one expression to another without change of meaning is being made. Here is the place where the transition accounts for an operation. Thus an operator helps to bring out the process of transition from one expression to another in accordance with equivalence. Here it is seen obviously how equivalence and an operator mediate the process of transformation. Similarly, the relation between definition and equivalence can be seen as follows: it is said earlier that the function of both an operator and a 91
definition are one and the same. Then whatever function the operator does can also be done by the definition under the same condition in which an operator works. If operators are workable in accordance with equivalence then, according to our explanation, the definition too is workable in accordance with equivalence. As the connection between a definition and equivalence is so intimate the definition is always symbolized in the form of an expression propounding equivalence. Then the validity of a definition depends on the way a certain term (e.g. The word planet ) is used in English, without belonging to the body of truths provided once and for all by logic, to describe or to denote something. In such a case, two sentences express the same fact, by virtue of definition, such that they can be construed as a substitution license. This makes the possibility of definitional equivalence in any definition. 3 incidentally, definition differs from substitution license. The reason being the former can be applied to any logical equivalence as they are related intimately with one another (logical equivalence can also yield substitution license) whereas substitution license cannot be extended to any case as they are limited to a given formula only. It means, all definitions give rise to substitution license but all substitution licenses do not give a definition. 4 92
We are concerned about definitional equivalence and substitution license just to understand that definitions are substitution license and every substitution license can be rewritten as equivalence if we have equivalence, the applicability of logic is possible, such that, deduction can be made easily. The ultimate consequence will be a definition. That is, a definition is being derived from a definition. Waismann s expression of analytic depends on the notion of logical truth and operators. The accuracy of his definition depends on the clarity and precision of the terms used in the definition. His contention is that these terms are more or less blurred and consequently the conception of analytic cannot be precisely defined. 5 If it cannot be defined, its boundary is open such that no one can recognize an analytic statement at any given instance. The intention behind Waismann s view is that analytic statements, if they were defined, would be hard to understand. The reason being, any definition, if it is defined, gives only fixed and settled terms. He tries to keep the definition flexible so that the terms involved in it may adjust according to the new situations that may arise in the future. On the other hand, hard and fast rule would be turned into a pseudo-precision in due course of time. 93
One may notice two kinds of parallel attacks made by Waismann on analyticity. At the first level his attack is on the phenomenal level of definition. He says that the term definition is ambiguous and hence it is open in its texture. The other is within the concept of analyticity. It is said that analytic statements are statements that can be transformed into logical truth by means of definition. In the process of transformation, as it is said earlier, certain operators are involved and one among these operators is the definition. In the proceeding explanation it should be taken that the attack on definition is common to both the cases. What is a definition? Waismann considers different kinds of definition. He says all these definitions are blurred. By considering one by one through examples, he passed through a number of analytic statements. This helps us, as well, to understand, how analytic statements are applied. Explicit definition is a license which permits us to replace a word, or a symbol, by the definitions, i.e., to translate an expression into a different idiom. 6 As it is applicable to trivial cases it does not give rise to great disputes. His concern is mainly on recursive definitions in use. To see whether such definitions really define anything concisely and precisely forms the content of the proceeding paragraphs. 94
Definitions in use are those definitions which get their meaning by understanding the instances in which they are applied. One such case is the concept of Time. What is time? Each and everybody knows what time is. Yet, if we were asked what it is, we are reduced to speechlessness. Perhaps one may say time is the form of becoming. If it is taken as the definition of our case that we are considering now, then according to the principle of definition, the phrase the form of becoming can be replaced by the word Time, whenever and wherever it comes in our way without a change in meaning. Obviously this substitution license will not give sense at any time. We are confronted with no combinatory word which does the job the noun Time does. One may give numerous cases where in the word Time can be used in a different way other than the word Time. Examples for such cases are, by this time = now, a long time ago = long ago, behind time = late, from time to time = now and then, now is your time = now is your opportunity; had a good time = enjoyed myself etc. 7 But the difficulty is to go on with one translation consistently for all the phrases in which the word occurs, i.e., to make the translation to prolong throughout. The fact is that there is no standard translation that can serve to eliminate the word from any context whatsoever and replace it by the definition. 95
For these reasons one may conclude that the word time cannot be defined through definition in use. Its use can no longer be distilled in rules and hence it seems counter to the doctrine, meaning = use. The reason being no one can limit the concept of use to give a set of words. On the other hand the concept use is an incomplete process as it is going to apply to numerous cases in the future. Waismann considers whether dictionary definitions can be accounted as a genuine case. The problem that arises if we take dictionary definitions as definitions is that it turns into a vicious circle in course of the analysis of the involved and begs the question. As we are looking for logical strictness we cannot account for these types of circular definitions. Further dictionary will give expressions which come as near as possible to the meaning of a word without coinciding with it completely. This enables us to gather more information about a particular word. For this reason we shall have to preclude dictionary definitions. Waismann s attention then turns towards ostensive definitions. There is a controversy whether non-verbal definitions are definitions. Ostensive definitions are indeed non-verbal definitions but they have to behave as verbal definitions at certain instances. Pointing an object together with pronouncing the word is exactly as a definition. In this respect, percept serves as the media of understanding the object that has been pointed and so 96
perception can be followed, broken and referred to in order to justify the word s use. 8 All these methods of apprehension have a close resemblance to the constituent elements used in understanding a verbal definition. But in ostensive definition the act of pointing is not the only means to understand something. In this context, it is necessary to recall Wittgenstein s attack on learning a language ostensively. Language is like a box of tools. This shows that mere pointing is not enough; rather one has to look for the context in which the act of pointing was made. In verbal definitions, one is covering all the bounds of language. But in non-verbal definitions like ostensive definitions, one has to exceed the realm of words. The most important aspect is that an ostensive definition can no longer be used as a substitution license. 9 If I come to know a man by name John with the help of another man who pointed and showed John s physical body to me, then I cannot substitute the word John with the physical John in terms of verbal definition. If ostensive definitions were admitted as a means of defining something, one has to work within the length that the percept can reach. Further, pointing something and saying red would not help someone to understand the colour red. In this respect, ostensive definition is incomplete. One has to search for the connective words such as red means ; is called red, it designates, etc., which makes easier to catch the meaning. So it 97
becomes necessary to understand these connective words which are difficult to understand through ostensive definition. So in this definition we are confronted with words that cannot be defined ostensive. If it is argued that these connective words can be understood in course of training one has to get used in an ostensive definition where the training part ceases and the definition part begins, which is, of course, impossible to know. Many more examples can reveal the truth that ostensive definitions are not definitions at all. All these show that ostensive definitions cannot give rise to analytic statements. It may be argued through examples that ostensive definitions give rise to analytic statements. But on strict scrutiny it reveals the truth that it wouldn t give rise to analytic statement. 10 Consider the following example, if a meter is defined as the length of the standard meter in Paris then the statement the standard meter is 1 meter long is an analytic, for the definition of the meter when substituted in our case turns into a logical truth. This statement is true by ostensive definition. It is so because as the word meter is significant, there must be at least one thing that is one meter long by pointing to which the meaning of meter can be explained. But the crucial point here is that a statement is analytic if it is transformable by means of definitions into a truth of logic. On analysis, one could find out that no such transformation has 98
taken place. We do not require a connection between the word meter and the actual standard in Paris. What we do require is that verbal description cannot take the place of the definition itself. For all these reasons, it becomes necessary for Waismann to reject ostensive definitions as the producer of analytic statement. By definition, we gathered a number of explanations that do not go in hand with others, but all of them are bound together in certain relations. If we group cases of definitions together according to their similarity, it is easy to see that the similarity resolves itself into a bundle of relations, holding at times only between the forms of the definitions, at other times between the various jobs they perform. It is in this connection that he asks how much definition is to be taken to cover in the formula just quoted, and draws attention to the wide variety of cases to which, owing to the ambiguity of this term and the lack of suitable definition, it is hard to say whether the formula applies or not, in the example, Time is measurable, etc. Thus Waismann s whole philosophical outlook is as was that of later Wittgenstein. His rejection of the analytic-synthetic dichotomy is to be connected with his rejection of the view that whatever we apply a common name to, we must be thinking that things or situations to which we apply it literally have some feature/s in common. 11 It is so because no precise boundary can be 99
set to the applicability of such terms as it is possible for the coming up of fresh instance which tempts us to widen their extension. Contingent and Necessary The difficulty in answering whether the statement Time is measurable is analytic, results in the indeterminacy of considering this statement as contingent or necessary. One may be inclined to say that it is just an accidental fact as no one is trying to affirm or deny any fact at all. 12 On the other hand, one is never intended to object any of the facts which make it possible to measure time. So it is a puzzling statement where we cannot decide any one of the truths. Waismann says that the reason for our bewilderment towards this problem is due to the case in which the philosophical antithesis contingent- necessary loses its edge. No one can pick out the border line that separates analytic-synthetic, necessarycontingent, a priori-a posteriori statements. To get the clear idea that the line of separation cannot be traced, Waismann analyzed and interpreted a number of examples such as I see with my eyes and hear with my ears; every experience is some one s experience; two people cannot share a pain; though it may well be that they have pains, which are in every respect alike, etc., for the sake of explanation, we will take some of these examples. 100
Whatever may be the case, any statement of any matter can be taken, more to say, used in two senses, namely; a trivial and important. 13 The statement I see with my eyes on the moment of my utterance, would not give any extraordinary implication, as one who hears it may simply exclaim, of course it is; but what of it?. The reply seems to reject my utterance as a sort of platitude. In this respect, the statement that I uttered is taken in a trivial sense. On the other hand, if he asked me why I should see with my eyes? How the eye is made use of in seeing something? What do you mean by seeing? Why the eye should always be correlate with seeing? Can I see with an organ other than an eye? etc., then these questions make me think and raise metaphysical questions on the statement I see with my eyes. In these contexts the statement would then have to be treated in an important sense. The same shift of meaning may occur when we consider an expression like every experience is the experience of a person, i.e., it occurs in somebody s mind. 14 For a moment this sounds like a truism. Yet on further penetration into the assertion with a puzzling gesture it reveals the truth that experiences do not occur by themselves and it is a significant outcome that may again be used as an input of an inquiry. In all these examples certain ways of thinking convert trivial statements into an important statement. It is also possible to move in the opposite direction, i.e., from important to trivial. For example, so long as we think that the 101
earth is unique among the heavenly bodies, we are tempted to attribute it a deep, metaphysical significance which, however, evaporates, once the earth is seen to be just one planet amongst others. All these statements may be looked both from trivial and important angles. These statements and the examples of puzzling expressions do not combine neither of the labels contingent and necessary. Waismann found out these considerations in his further courses of his discussion. Take the sentence, I see with my eyes. At first, it seems to be contingent as I cannot see anything when I shut my eyes. But if one looks into the meaning of sentences of similar sort like I eat with a knife and fork then the sentence I see with my eyes will no longer be contingent as in the former case there are numerous alternatives, to perform the job. Whereas in the case of the statement in our consideration, only one object, Eye fits into the meaning of the sentence and nothing else. If any one says, I see you quite distinctly with my eyes shut, then, it is certainly a contradiction as seeing and closed eyes are negating each other. Leaving aside the questions regarding seeing an after image with one s eyes shut, we will proceed further into the meaning of the sentence in our considering. This sentence can be seen as a factual as well as a non-factual statement. It is factual when I see this in a trivial sense. On the other hand when I say that it is part 102
of the meaning of seeing that I can see a thing only when my eyes are open, and when I look at it, it will become clear that this applies to one sense of the verb only and in this sense, then saying, I see with my eyes shut must be some sort of a contradiction; if so, saying, when I see you, my eyes are open must be a nonfactual a priori statement. Here the existence of the latter is a must for the act of the former and this shows that this connection is not an accidental one, but a necessary one. We have also seen that the statement I see you as both of them are conveying a truism. All these explanations bring out our prior assumption that certain statements like I see with my eyes do fall into a class where neither necessary not quit contingent truths can be attributed. 15 But it does not follow that neither of them is absent in this statement. Both empirical and analytical characters are inherited within them but within an unequal distribution. One among them predominates over the other. This resulted in the declining of the strength of one of these characters. Waismann s conclusion (towards the statements we have just analyzed) is that one cannot classify analytic and synthetic statement with precision, such that one excludes the other completely. When he says something is analytic, it means that he is taking the situation in which these statements are uttered as constant so that his utterance can be related more in terms of analytic with the constant and also can be related less in terms of synthetic with the constant. 103
This would be clear if we consider the example, space has three dimensions. This statement is hardly accepted as a contingent statement for we fail to see why space should have three dimensions and not four. In addition to that, this statement cannot be put into any rational insight. The reason being we are conceiving this statement as such without any question. Still, we cannot say that it is a necessary statement. If it is a necessary statement, one understands what it is that makes it a necessary truth. In the case of space has three dimensions, no one understands the subject space as something constituting itself three dimensionally. This makes us to say that this statement is not a necessary statement. Some may consider this statement as a contingent fact as it is opaque to the statement once considered as necessary and later as contingent. Yet it is not quiet contingent, and not quite necessary, not quite empirical and not quite logical, they are half way between. So far we have seen that the accounts on facts, such as necessary contingent, analytic-synthetic etc., failed to limit its realm within their own concepts. Ultimately, it leads us to the point that there is only one kind of fact and not two as supposed earlier. Facts are what they are, contingent. 16 But statement of the form half of a half is a quarter is a necessary truth. Note the two different kinds of words viz., fact and truth that have been used here. The difference between a fact and a truth is such that the 104
former describes something while the later is not used descriptively at all. The difficulty that confronts us here can be brought out as follows. When we speak of a fact, we seem to have in mind something that is asserted in an empirical statement the truth of which depends on experience. So long as we have not turned to experience we do not know whether that statement is true or not; it may conceivably be either. And even if we know that it is true, we can think about what would be the case if it is false. So a fact is what has alternatives. That is to say contingent fact has alternatives. Now look at the statement, space is three dimensional. Can we think about a four dimensional space? It is impossible. Thus there is no alternative available here and hence the mystery still remains. Waismann s appeal to ordinary language rests on the bewilderment about the notion of fact. The so called claims of facts, in statements like space is three dimensional lead us to the confusion of the whole system, for we get confused about ordinary usage of language in our day-to-day affairs. Is there any remedy to this confusion such that our language can be conceived free from any confusion? It is possible only if language has a change in meaning from time to time. 17 Because we never stick to the same meaning of a particular word in a particular language always. The reason for assigning new 105
meaning is that the use of a particular word changes from one society to another society and in such a case if there we put into existence a new society with new culture and tradition there arises the need to assign a new meaning to a word that was held with some other meaning. For example, the word romantic initially had the meaning false, imaginary, unnatural, etc., and later it becomes a term for a literary emotion, applied to nature seen through the medium of poetry and fiction. This type of reshaping the language results in a chance of the world picture embodied in it. Waismann s intention being that a philosopher need not bound him/herself to the actual use of language. 18 He is supporting this view strongly, for he is encouraging this change as the criteria to be achieved. If it is achieved, then it will give way for the development of any philosophy. To talk of the ordinary use of language, according to him, is unrealistic. If at anytime he is adhered to the ordinary use of language, then he is restricting himself to the study of one language to the exclusion of the other. This results in the obscurity of ubiquitous features of their own language. So one s own language is noticed and recited in their full significance only when it is compared with other languages. Waismann s approach towards language as a tool that can be applied to new cases of future inventions in the language world, has been strongly criticized. His view on language as 106
flexible has to limit its conceptual structure at least in one point, or else it will not be possible to put into use in the courses of effective thinking. W.H. Walsh in his articles, analytic-synthetic criticized the open-texture concept of Waismann. 19 He said that empirical concepts are open in its texture such that no precise boundaries can be set up that results in the in-definability of empirical concepts, if it is so, it is not possible to know whether a statement is analytic or not, if that statement inherits empirical concepts with itself. Whether this restriction is applicable to natural as well as artificial language is our prime concern. Walsh agrees with the limitations expressed in artificial language, but he is going against the view that natural language is devoid of limitations as expressed by Waismann. The difference of opinion arises from the fact whether all the things denoted by the same fact have anything in common. Now it is evident that the artificial language does not possess the open texture principles, whereas the natural languages do. To illustrate consider the following examples of Walsh, a triangle which is an artificial language, has no chance of including a principle, other than the one that exists now, i.e., a triangle must have three angles, in the future and hence it does not follow, the open-texture concept of Waismann. But a natural 107
language like game (as given by Wittgenstein ) does not limit the number of games that are in practice today, but it is ready to include in the future any sort of thing that can said to be a game. This makes Waismann to say that natural language is flexible and no precision on its part can be made. Waismann digests the view of natural language as the one which leaves everything as such. But the question Walsh raises is whether the phenomenon remains the same after the reshaping of a natural language, if it is admitted to go like that? He points out the example given by Waismann such as romantic, revolution, etc., and says that the phenomenon will not remain as such, which implies that the adaptable nature of natural language loses its purpose in course of time. So Walsh argues that at one point or another, the natural language has to limit its boundary or else they would become empty. He gives the ways of limiting a natural language. This is possible from a personal judgment on concrete cases as a natural language coincides with the precise boundary that we set up for analytic statements. That is, it is possible to decide upon the empirical concepts in the future. Having made clear that the natural language is definable as it is bound with the decision of ours, it is not possible to stick with the view that a formal definition is not possible and hence analytic is not possible. The implication behind this is that the 108
distinction between analytic and synthetic can be made, as it is said in the previous chapter, depending upon specific circumstances. It is so because this distinction depends upon the decision of a person with respect to the circumstances under which these statements are given to the deciding person. This dilutes our issue regarding the problem of language expressed in terms of analytic and synthetic distinction. In contrast to the views of Waismann, W.V. O. Quine makes an attempt to dissolve the distinction between analytic (a priori) and synthetic (a posteriori) statements from the scientific point of view. The empirical approach of Quine led him to reject the analytic -synthetic distinction. He claims that no one can define analyticity without presupposing the notion of analyticity. Quine s Critique of the Distinction We have just seen that the traditional philosophy distinguishes all propositions into two distinct classes. They are the statements classified as analytic and the statements classified as synthetic. W.V.O.Quine in his essay Two Dogmas of Empiricism brings forth the problem in distinguishing analytic and synthetic statements. His problems related to such distinction rest on the problem of meaning involved in the notion of analytic statements. He concludes that there is no such distinction as conceived by the traditional philosophers. 109
This makes Quine to say that the two dogmas of empiricism are logically ill-founded.20 One of the two dogmas of empiricism is a belief in the cleavage between truths which are classified as analytic and truths which are classified as synthetic. Analytic truths are grounded in meaning, independent of matters of fact, always true and they are necessary statements. Synthetic truths are subject to experience, they may be true or false and they are contingent statements. The other dogma is reductionism. Reductionism is the principle through which one can reduce nonempirical statement into empirical statement. Both these dogmas are connected with one another through the verification principle. Since we cannot speak of the distinction of statements as analytic and synthetic, Quine refuted them on the following grounds. Quine traced back Kantian approach to this problem and he affirmed that Kant conceived a statement as analytic only if it attributes to its predicate what is conceptually contained in the subject itself. This yields two shortcomings, viz., it restricts itself to subject- predicate form and it appeals to a notion of containment which itself need clarification. For this reason it is better to leave the definition part of Kantian views, instead one may take the level at which Kant made use of the notion of analyticity. At this point Quine rephrased the description as follows: A statement is analytic when it is true by virtues of meanings and independently of fact. 110
Here it is one s duty to examine the concept of meaning. According to Quine, meaning is not identical with naming. It shows that he accepts Frege s distinction between sense and reference. 21 Frege s example of morning star and evening star illustrates that terms can name the same thing but differ in the meaning. It applied both to the level of abstract and general. The terms 9 and the number of planets name one and the same abstract entity but are unlike in meaning because astronomical observation was needed to determine the sameness of the entity in question. This is with regard to the case of singular and abstract terms. Similar is the case with creature with heart and creature with kidney which are alike in their extension but different in their meaning. Problem of Synonymy Quine further says that if it is possible to make distinction between meaning and reference, we presuppose analyticity. 22 But the question that we asked earlier leads us only to obscure talk about entities or the Aristotelian notion of essence as it presupposes the modern notion of intension or meaning. Quine says, Meaning is what essences become when it is divorced from the object of reference and wedded to the world. 23 This makes Quine to say that the meaning as some intermediary entities can be abandoned. 111
Here we are facing the problem of analyticity; Quine divided analytical statements into two classes. The first one may be called logically true statements. It can be understood with the following examples. 1. No unmarried man is married. Here the truth remains the same under all interpretations of its components namely man and unmarried. The other class of analytic statements is based on synonym which is typed by: 2. No bachelor is married. This statement is turned into a logically true one by substituting synonyms for synonyms. In this way the second class of statements can be transferred into first class by substituting unmarried man for its synonym bachelor. Apparently it results in a tautology. Hence, according to Quine, we lack a proper characterization of analytic statements of the second class. They mainly depend on synonyms, which in turn presuppose analyticity. The notion of synonymy is in no less need of clarification than that of analyticity itself. Synonymy as Definition Even the definition of bachelor as unmarried man does not serve the purpose. Because, bachelor is defined as unmarried man. Who defined so and why? According to Quine, we naturally tend to refer to the dictionary and accept the lexicographer s explicit definition as law; for the lexicographer is merely an empirical 112
scientist who records antecedent facts. He happens to gloss bachelor as unmarried man, it is because of his belief in his empirical observation that there is a relation of synonymy between these two terms implicit in general or preferred usage prior to his own work. 24 Therefore, the notion of synonymy presupposed here still has to be clarified. The kind of criteria Quine demands is a behavioral one. According to him, two terms are synonymous if and only if a native speaker assents to both when presented with one and the same stimulus. However, Quine makes an exception here. That is, definition can be accepted as transparent case of synonymy only in the case where notions are created for the sake of abbreviation. 25 Synonymy as Interchangeability Others try to define synonymy in terms of interchangeable salva veritate (truth being preserved) as used by Leibniz. This would hold true if and only if something is substituted for another in any sentence without changing the truth value of the sentence. But it is not quite true that synonyms bachelor and unmarried man are interchangeable salva veritate. For example, we cannot say bachelors of arts is one and the same as the unmarried man of arts. Quine goes further and says bachelor has less than ten letters and hence in its place we cannot substitute unmarried 113
man. However, he feels that still a criterion could be extracted out of this definition by making a provision that interchangeability salva veritate is not supposed to apply such fragmentary occurrences within a word. 26 This principle may be defined as cognitive synonymy assuming analyticity. Superficially looked at, this principle does not seem to presuppose analyticity. For example, terms like bachelor and unmarried man are interchanged salva veritate in the following sentences: To say bachelor and unmarried man are cognitively synonymous is to say: 1) All and only bachelors are unmarried man is analytic. Resuming the sufficient conditions of interchangeability we can say that, 2) Necessarily all and only bachelors are bachelors. This sentence is true if we suppose necessity as narrowly applicable only to analytic statements. If bachelor and unmarried man are interchangeable then, 3) Necessarily all and only bachelors are unmarried men must be true. So, to say that (3) is true is to say that (1) is analytic. Therefore, bachelor and unmarried man are synonymous. But a closer analysis reveals that we are confronted with the idea that this principle viz., cognitive synonymy, does presuppose 114
analyticity. Our business here is to give an account of cognitive synonymy not presupposing analyticity. Quine says that the condition of salva veritate varies in its force with variations in the richness of the language at hand. 27 That is, if a criterion is to be extracted from Leibniz s criterion we have to consider the language to which it is applied and in particular whether the language under question is extensional or intensional. For Quine, a language is extensional if any two predicates which agree extensionally (that is true of the same object) are interchangeable salva veritate. Any formal language would fit into this category. Therefore in an extensional language interchangeability or salva veritate is no assurance of the kind of synonymy required. We can have the assurance only with regard to the truth of the statement. For example, in a statement where such a substitution has taken place, we know only that it is true and nothing else. There is no common assurance whatsoever that the above statement is true by virtue of bachelors and unmarried man being synonymous. So it made us think that it is merely an accidental matter of fact. For instance, morning star and evening star will in an extensional language be interchangeable salva veritate and they are not synonymous in the ordinary sense but they surely agree extensionally. For this reason, Frege s identification of morning star and evening star is not a necessary synonymy but merely accidental. Similarly, for the term, 115
philosopher extensive synonymy is Socrates or Plato and the intensive synonymy is lover of wisdom. It implies that it is merely a matter of contingent fact that morning star is identical with evening star. But the type of synonymy we expect is a sort of interchangeability or salva veritate, which is necessarily feasible, but not contingently so. In the case of a natural language like English, which is intensional, where an adverb like necessarily is used, matters are still worse. Here the definition of synonymy assuming analyticity involves the notion of necessity and it shows that we are working with a language rich enough to contain the adverb necessarily. This adverb yields truth when and only when applied to an analytic statement. This idea of necessity presupposes the notion of analyticity. How? We can ask the question whether this adverb makes any sense. To suppose that it does is to suppose that we already know the meaning of analyticity. Therefore, the Leibnizian criterion loses the battle. So we go back to the beginning. We define analyticity in terms of synonyms. These synonyms are defined in terms of necessity and this necessity presupposes analyticity. Thus we are in a vicious circle. So there is no clear-cut demarcation between analytic and the synthetic statements. 116
We have seen the problem so far in question only with regard to the ordinary language and here we made the conclusion that the distinction between analytic and synthetic statements is not possible because of the vagueness of the ordinary language, for we do not have precise semantic rules. Quine s problem with ordinary language results in the confusion of understanding a statement, whether it is analytic or not. For example, Quine does not know whether the statement everything green is extended is analytic. The problem here is not with the meaning of the words, green or extended but with the meaning of analytic. So it is claimed that the difficulty in separating analytic from synthetic statements vanishes once we have a precise artificial language with explicit semantic rules. But such a notion according to Quine rests on confusion. This is due to our conception of the semantic rules with the assumption of analyticity of the former s application. Reductionism The dogma that the cleavage between analytic and synthetic truths is linked with the second dogma of empiricism, namely, reductionism through the verification principle. Reductionism is the belief that each meaningful statement is equivalent to some logical construct upon terms which refer to immediate experience i.e., each non-empirical statement can be reduced to empirical statement. The verification theory is that the meaning of a 117
statement is the method of empirically confirming or informing it. 28 Now, what is the nature of the relation between a statement and the experiences which contribute to or detract from its confirmation? The most naïve view of the relation is that it is a direct report. This is radical reductionism. i.e., every meaningful statement is to be translatable into a statement about immediate experience. This radical reductionism in one form or another well antedates the verification theory of meaning. This theory can be rephrased as follows: that a term to be significant it must be either a name of a sense datum or a compound of such names. Quine rejects this view because (a) this doctrine is ambiguous as sense data, as sensory events and sensory qualities, (b) it also remains vague as to the admissible ways of compounding and (c) it is very restrictive in its term by term critique which it imposes. 29 Now Quine suggests that we must take the full statement as our significant unit and not individual terms. Thus demanding that our statements as a whole be transferable into sense datum and not that they be translatable term by term. Quine goes one step further in saying that each statement, taken in isolation from its fellows, cannot admit of confirmation - as it is supposed in the reductionism, but our statements about the external world face the tribunal of sense experience not individually but only as a corporate body (holistic view) of 118
language world. 30 So, according to Quine, there is no individual or a set of statements which is very basic or primary because individual statements mean nothing in isolation as language is interconnected. This idea of rejection of reductionism leads to the holistic view of language. Quine says that the whole language confronts the world and not individual statements. Language is a network of statements and the reflection of one sentence about the world is supported by other statements. He says that language is like a field of force with empirical statements at the periphery and logical statements at the core. They are all interconnected. A conflict with experience at the periphery results in readjusting the interior of the field. Revaluation of some statements entails the revaluation of others. The reason being that they are logically inter-connected with each other. If we accept this view, then it becomes fully to speak of a distinction between analytic and synthetic truths, because any can be held true (come what may) by making adjustments in the system. So also there is no absolutely contingent/synthetic statement. So also there is no statement which is immune to revision. So there is no absolutely analytic/synthetic statement. For example, the statement, it s green is a true empirical statement. Now we can change the truth of this statement by changing some inner principle and formulating it as everything you see in an illusion. Now one can easily doubt the truth of the 119
empirical statement. So also, one can accommodate a recalcitrant event by adjusting inner principles. For example, one accommodates the idea that a tree is a tree and a stone by readjusting the principle of identity at the core. Therefore the whole language (world) is interrelated and there is no statement which is fundamental and immune to revision. This therefore rejects the distinction between analytic/necessary statement and synthetic /contingent statement. This goes against the reductionism as it says that verification is not word by word or statement by statement but it is the whole of language confronting the empirical world. Quine s idea against synonymy and the holistic view of language leads to the indeterminacy of translation. For him, statements mean something only in the context of a whole set of interconnected sentences. Therefore two statements taken in isolation cannot be synonymous. If one cannot speak of synonymy even within the same language, then how can he talk about synonymy in two different languages? This makes him to conclude that translation is strictly impossible. One can talk about synonymy only in extensional context and not in intent(s)ional context i.e., in the sense of meaning same as, for example, Number of planets =9 (synonymous extensionally) Number 9 number 7 120
Therefore,number of planets 7 This is non-sensual because it is not necessarily true but it is empirically true. Since synonymy is not possible in intentional context what we do in translation is we are imposing our ontology on the language of the other. The new language is understood in terms of our ontology. One is not learning the other language. For example, the meaning of two sentences remain the same (and hence analytical) if the words in the sentence have the same meaning. Let s say there are two sets of sentences. I sentence II sentence X Y Z P Q R A A The word X in sentence has a correlation, (i.e., meaning) A: similarly the word P in some other new sentence has the same correlation A. These two sentences can be analytical. But the problem here is these words are not similar and analytical by themselves but they are similar only on the condition of my application of the assumed meaning. Similarly, in learning a new language one has no chance of having a priori understanding that a word in a new language has the same correlation as his language. 121