International Journal for Philosophy of Religion 53: 93–110, 2003. © 2003 Kluwer Academic Publishers. Publishers. Printed in the Netherlands. Netherlands.
93
The logic and language of Nirvana: a¯ na: A contemporary interpretation
PETER KÜGLER Department of Philosophy, Philosophy, University of Innsbruck, A-6020 Innsbruck, Austria Abstract. In contrast to historically oriented approaches, this paper tackles the concept of
Nirv¯ Nirvana a¯ na from the perspective of contemporary philosophy of language. It focuses on four propositions: Nirv¯ Nirvana a¯ na exists; Nirv¯ Nirvana a¯ na does not exist; Nirv¯ Nirvana a¯ na exists and does not exist; Nirv¯ Nirvana a¯ na neither exists nor does not exist. The Buddha’s rejection of these propositions is interpreted by means of explicit and conditional definitions of existence. Stalnaker’s notion of pragmatic presupposition provides an explanation why the propositions are without meaning. After comparing the word “Nirv¯ “Nirvana” a¯ na” with indexicals, proper names and theoretical terms, it is finally asked what linguistic function the word has.
Meta-Affirmation and Meta-Negation 1. Affirmation, Negation, Meta-Affirmation
In Buddhist studies, the term “logic” designates a large field of historical, epistemological, metaphysical and, indeed, logical research, as exemplified by Theodor Stcherbatsky’s famous two volumes on Buddhist Logic.1 This paper, of course, has a much narrower scope. It is only concerned with the concepts of existence and Nirv¯ Nirv ana a¯ na (Nibb¯ (Nibbana), a¯ na), and with the presuppositions and consequences consequences of bringing these concepts concepts together. together. According to tradition, though, the Buddha has warned his disciples not to think about whether the Tath agata a¯ gata – the one who has reached Nirv ana a¯ na – (1) exists after death, (2) does not exist after death, (3) both exists and does not exist after death, (4) neither exists nor does not exist after death. In the Pali a¯ li Canon, this warning is typically justified by the claim that metaphysical speculation is of no use to those who strive for Nirv ana. a¯ na.2 The prospect of missing enlightenment, however, did not prevent Buddhists from speculating. Curiosity has often been a stronger motive. In addition, the idea has been put forward that speculation and meditation must go hand in hand. In order to get into the position of achieving Nirv ana, a¯ na, one must first know where to look and how to recognize it. In this process, one may emphasize either speculation or meditation. As Mircea Eliade noted, there have always been “Metaphysicians” along with “Yogins” in Buddhism.3 Provided that it is not useless to think about the four propositions cited above, what attitude should Buddhists take toward them? In classical logic,
94
PETER KÜGLER
there are two choices of distributing truth and falsity among the first two propositions. If the first is true, the second is false; and if the second is true, the first is false. Propositions 3 and 4, however, leave no choice, as they are logically false. Yet if we are ready to give up classical logic, any number of the propositions might be regarded as true. Conversely, all of them could be regarded as false. In fact, if we wish to assign truth-values at all, the latter seems to be a good interpretation, as it explains the Buddha’s warning. I will discuss this interpretation in section 3. In section 4, alternative approaches will be examined, including the one I consider to be the best. It is to treat all propositions as being without meaning and therefore as being neither true nor false. They are without meaning, because Nirv¯ana is inconceivable. Here, truth is understood in a realist sense. Roughly speaking, a sentence is true if it describes reality correctly. Truth in this sense requires referential relations between language and world. Meaning, on the other hand, is linguistic meaning; it is something which a speaker “has in mind”, or which he or she “grasps” when learning a new word. Linguistic meaning should not be construed too broadly. If we would identify it with use, for instance, the four propositions, as well as the word “Nirv¯ana”, would trivially have meaning, since they are used in some way or the other. (In which way “Nirv¯ana” is being used will be explained at the end of this paper.) However, as Ludwig Wittgenstein wrote, “we understand the meaning of a word when we hear or say it; we grasp it in a flash, and what we grasp in this way is surely something different from the ‘use’ which is extended in time!” 4 One should keep in mind, nevertheless, that the lack of linguistic meaning does not imply that considering the four propositions has no “meaning” for the pursuit of practical goals. If we follow the suggestion of meaninglessness, or alternatively the suggestion of falsity made before, it need not be useless to speculate about the four propositions. This speculation could be effective, for instance, if the goal is to achieve a higher state of being, or something less ambitious. But what we discover by that survey, according to the two suggestions, is that the propositions are either false or without linguistic meaning. Once we have understood this, we may decide not to worry about the Tath¯agata’s existence or non-existence after death any more. Because why worry about false or meaningless propositions? Thus our philosophical reflection would have been useful, and we would be able to agree with the Buddha’s warning, but this time for philosophical reasons. This has also been the preferred strategy of Buddhist philosophers. One of them, N¯ag¯arjuna, explicitly rejected the four propositions in his M ulamadhyamakak arik ¯ ¯ a¯ (XXV.17), arguing from the premise that all things are empty (s¯ ´ unya). I should note, however, that this is not the line of thought
¯ THE LOGIC AND LANGUAGE OF NIRVANA
95
to be followed in the present article. Rather than on emptiness, the focus will be on the definition of existence. Since Nirv¯ana is something like the “state” of the Tath a¯ gata after his corporeal death, we may talk about Nirv a¯ na and the Tath¯agata as if they were one and the same. Thus the four propositions on the existence of the Tath¯agata can also be read as being about Nirv a¯ na. Let us give them names for easy reference: Affirmation: Nirv¯ana exists. Negation: Nirv¯ana does not exist. Meta-Affirmation: Nirv¯ana exists and does not exist. Meta-Negation: Nirv¯ana neither exists nor does not exist.
These propositions depend on each other in an obvious way: Negation negates Affirmation; Meta-Affirmation affirms both Affirmation and Negation; Meta Negation negates Affirmation and Negation. As it is well known, propositions comparable to these have been considered not only in Buddhism, but in other religious and philosophical systems too. In the western world, negative theology has been formulated in a similar style. Meta-Negation, for instance, reminds of Pseudo-Dionysius who claimed, in the final lines of his Mystical Theology, that God is beyond affirmation and negation. However, there is a distinction which has not been made by the author of the Mystical Theology, but which later became central to negative theology. It is the distinction between God’s essence and his existence, the former comprising several attributes, such as omnipotence, omniscience, goodness and unity. In the Judeo-Christian tradition, negative theology was only directed towards God’s essence. Theologians denied that this essence can be described by any positive attributes, but usually they did not hesitate to assert God’s existence . Moses Maimonides, who influenced Thomas Aquinas in this matter, wrote as clearly as one could wish “that we comprehend only the fact that He exists, not His essence. Consequently it is a false assumption to hold that He has any positive attribute.” 5 So the mainstream of negative theology in the Judeo-Christian tradition has been no negative approach at all, as far as God’s existence is concerned. In terms of the propositions above, negative theology has not exceeded level one, Affirmation. Putting it differently, if in Affirmation the word “Nirv¯ana” is replaced by “God”, the outcome would have been accepted by Maimonides, Thomas and their followers. They believed that God exists; they only denied that his essence can be positively described.
96
PETER KÜGLER
2. Approximations of existence
Apart from any comparison with Buddhism, negative theology has a serious defect of its own. Existence without essence is unintelligible. It does not make sense to assume that something exists without having any positive attributes. Consider two examples: physical and mental things. Physical things exist in space and time, if they exist at all. In this case, to exist means to have spatial and temporal properties. If materialism is wrong, physical things are not the only ones that exist. There are also mental “things” like feelings, sensations or thoughts. For these things, to exist is to have mental properties – phenomenal features in the case of feelings and sensations, intentional content in the case of thoughts. These examples of existence lead to the idea to define the concept of existence by the very notion of having properties. Without properties there can be no existence. Yet if something does have properties, existence can be ascribed to it as well. Strictly speaking, existence does not presuppose that a thing has a plurality of properties. One property alone would also do. Therefore, in the following definition, “X has properties” may be read as “X has at least one property”. 6 Formally, the definition is an explicit one: “just in case” is meant as expressing an equivalence. The variable X stands for anything whatsoever. Definition 1: X exists just in case X has properties. Although this definition is a good starting point, it is not the best analysis we could get. Actually, it is just a first approximation. In order to improve the definition, I shall first distinguish between three kinds of properties which may be called material, mental and subject-relative properties. The first type includes properties like being round , which is a property of spheres, and being solid , which is a property of solid bodies. Setting aside the claim of idealistic philosophers that all properties exist relative to subjects – if they exist at all – we may characterize material properties by the fact that they exist independently of any subjects. At least this is how they are usually construed in philosophy. In contrast to material properties, the existence of mental properties does depend on the existence of subjects. In fact, properties of the second type are those which make something a “subject”. Nothing can be a subject unless it possesses mental properties like perceiving, imagining and thinking. Properties of the third type, subject-relative properties, existentially depend on mental properties. When a subject perceives a thing, that thing has the subject-relative property of being perceived . If the thing is imagined by somebody, it has the subject-relative property of being imagined . When you see a ball in front of you, this ball has not only the material property of being round, but also the subject-relative property of being perceived by you,
¯ THE LOGIC AND LANGUAGE OF NIRVANA
97
since you have the mental property of perceiving the ball. The ball loses this subject-relative property as soon as you lose the respective mental property. The properties mentioned in the above definition must be of first or second type. The reason for this is that being perceived, being imagined or being thought of, or the possession of any other subject-relative property, does not imply existence. You may perceive (imagine, think of) a thing even though it does not exist. It might just be a hallucination or a dream. Only possession of material or mental properties guarantees existence, that is, material or mental existence. Sense impressions, mental images and thoughts have mental existence; material things have material existence. 7 Thus, on our second try, the explicit definition of existence looks like this: Definition 2: X exists just in case X has material or mental properties. Now we are ready to return to the four propositions on the Tath a¯ gata or, for that matter, on Nirv¯ana. In Samyutta Nik aya ¯ XXII.86, these propositions are discussed in terms of the five skandhas (khandhas). It is said that the Tath¯agata cannot be found within the skandhas, which is why the four propositions do not apply to him. The first skandha comprises all material properties. It is denoted by the word “r u¯ pa”, which is commonly translated as “form” or “matter”. Mental properties belong to the remaining skandhas: a), perception or cognition ( samjñ¯ a, saññ¯ a), sensation or feeling (vedan¯ ar a¯ ) and consciousness mental formations or dispositions ( samsk ara ¯ , sankh¯ 8 ana, viññ¯ ana). (vijñ¯ As the five skandhas include all material and mental properties, we are able to give an explanation of why Affirmation is false. It is false because Nirv¯ana is outside the five skandhas, having no material or mental property whatsoever. For the same reason, the second definition implies the falsity of Affirmation. As said in section 1, we want the other propositions to be false too. Unfortunately there is no easy way from the falsity of Affirmation to the falsity of Negation, Meta-Affirmation and Meta-Negation. But one could try the following line of reasoning: Assume that existence itself is a material or mental property. Assume further that if a property is material or mental, so is its opposite. The opposite of existence is non-existence, therefore nonexistence is a material or mental property too. From these assumptions it follows that the second proposition, Negation, is false. The argument leading to this conclusion takes the form of an indirect proof ( reductio ad absurdum). In short, Negation is false because it yields a contradiction. Here are the details of this indirect proof: Suppose that Negation is true, which is to say that Nirv a¯ na does not exist. By the assumption made before, the non-existence of Nirv¯ana is a material or mental property. Thus, by defini-
98
PETER KÜGLER
tion 2, Nirv¯ana exists, which contradicts the premise that Nirv¯ana does not exist. Therefore this premise is false; in other words, Negation is false. Given the falsity of Affirmation and Negation – which has just been demonstrated – it is straightforward to demonstrate the falsity of Meta Affirmation and Meta-Negation too, but we need not care about that, because the above argument has several weak points. First of all, the argument could be applied to any non-existing thing. Dragons do not exist. If non-existence were a material or mental property, we could prove, again with the help of definition 2, that dragons exist, as they have the property of non-existence. In this way, we would get a similar contradiction as in the case of Nirv a¯ na. Secondly, the argument implicitly presupposes the law of excluded middle (tertium non datur ). The fact that Negation implies a contradiction is reason enough to believe that it is not true. But not being true need not be the same as being false. Later we will test the idea that Negation, as well as the other propositions, is without meaning and therefore neither true nor false. If there is room for such a third alternative, the fact that Negation is not true does not imply that it is false. It could also be neither true nor false. A third objection deals with definition 2, where existence is construed as a property of second order. 9 Existence figures as a relation between certain properties and all things having these kinds of properties. Yet it does not itself belong to these properties. 10 The property of existence is on a higher level than material and mental properties. In the above argument, however, the level was wrongly assumed to be same for all properties, as we supposed that existence and non-existence are among the properties mentioned in the right-hand side of definition 2. It seems that the argument leading from the falsity of Affirmation to the falsity of Negation was flawed right from the start. But how then can the falsity of Negation be justified? A proposal is being made in the next section. 3. Falsities
As Heinrich Zimmer put it, “the ultimate and real task of philosophy, according to Indian thought, and to such classical Occidental philosophers as Plato, transcends the power and task of reason. Access to truth demands a passage beyond the compass of ordered thought.” 11 This is true of western mysticism, inspired to a large extent by Plato, and it is even more obvious in the case of Buddhism. Western mysticism and Buddhism are among those approaches which deny that God, Brahman, Nirv a¯ na, or whatever, can be understood by rational means. Instead some kind of “direct intuition” is favoured, which is preceded, accompanied or followed by a transformation of the entire personality. Moreover, it is claimed that neither the intuition
¯ THE LOGIC AND LANGUAGE OF NIRVANA
99
itself, nor its “content”, nor the result of the transformation, can be grasped by language or thought. The ineffability and inconceivability of Nirv a¯ na may also explain why Affirmation, Negation, Meta-Affirmation and Meta-Negation are false. The point of departure for this explanation is the notion of pragmatic presupposition, which can be defined in different ways. For our purposes, however, differences in details do not matter, so we may rely on Robert Stalnaker’s description: “A proposition P is a pragmatic presupposition of a speaker in a given context just in case the speaker assumes or believes that P, assumes or believes that his addressee assumes or believes that P, and assumes or believes that his addressee recognizes that he is making these assumptions, or has these beliefs.”12 On this account, it is reasonable that in an ascription of existence or non-existence, the pragmatic presupposition is made that the speaker has a conception of the thing which is said to exist or not to exist. Needless to say that usually this presupposition is made only implicitly. If the presupposition is not fulfilled, however, it has a good chance to become explicit, for then the speaker can be criticized for not knowing what he or she is talking about. If someone says that lizards exist and dragons do not, a plausible assumption of both speaker and addressee is that the speaker knows what lizards and dragons are. If the addressee has reason to believe that the speaker has no conception of lizards and dragons, the presupposition itself is likely to become a subject of debate between the two persons. Stalnaker’s description of the concept of pragmatic presupposition – which he considers not to be a definition or analysis, but something less precise – is incomplete because it does not indicate how the pragmatic presupposition is logically related to the assertion. In the next section, this presupposition will be embedded in a conditional definition of existence in order to explicate this relation. But first I shall try an alternative method, which is to form the conjunction of the assertion and the presupposition. As to the assertion itself, we are mainly interested in ascriptions of existence or non-existence. It should also be noted that every assertion depends on a variety of presuppositions, but here we need only consider the requirement that people have an idea of what they are talking about. With this caveat in mind, consider the following example: If I say that lizards exist, what I am really saying is, “lizards exist, and I have a conception of lizards.” In general, “X exists” translates as the conjunction of “X exists” and “I have a conception of X” (or, depending on perspective, “the speaker has a conception of X”). Here a “conception” is meant to include imagination, in the sense of forming mental images or models, as well as more rational ways of conceiving. In the terminology employed by Descartes in the Sixth Medita-
100
PETER KÜGLER
tion, to conceive ( cogitare) means either imaginari or intelligere – or both,
provided that these faculties can also be at work simultaneously. Thus the presupposition mentioned before could also be expressed in this way: Whenever an ascription of existence or non-existence is made, the speaker must have cognitive access to what is said to exist or not to exist. This access is through imagination and/or reason, by forming mental images and/or by appealing to intellectual concepts. Now back to Buddhism: in Samyutta Nik aya ¯ XXXV.83, we read that the one who has reached Nirv¯ana is inaccessible to the sense organs, including eye, ear, nose, tongue, body (touch) and mind. If this is true, it is also plausible that no conception of Nirv a¯ na, in the sense just described, can be formed by human beings who ponder the existence or non-existence of Nirv a¯ na. Mental imagination derives from perception; one is able to create a visual image of a lizard only if one knows how lizards look like. In contrast to lizards, however, Nirv¯ana cannot be imagined, as it cannot be perceived, neither by the eye, nor by the ear, nose, tongue or touch. A similar point can be made for concepts (dharma, dhamma), which are the objects of the mind ( manas, mano).13 Since Nirv¯ana cannot be grasped by the mind either, a “conceptual” comprehension is impossible as well. Therefore the falsity of the four propositions dealing with the existence of Nirv¯ana may be due to the fact that a certain pragmatic presupposition is not satisfied. Let us formulate this presupposition in the most general way, abstracting from any specific context of discourse: The presupposition is not that a particular speaker or addressee has a conception of Nirv a¯ na; it suffices to demand that it be possible at all to have a conception of Nirv a¯ na. For sake of simplicity, this presupposition may be abbreviated as C (which replaces the variable P in Stalnaker’s description). Then the four propositions can be written as conjunctions of the original propositions, respectively, and the condition C: Affirmation: Nirv¯ana exists, and C. Negation: Nirv¯ana does not exist, and C. Meta-Affirmation: Nirv¯ana exists and does not exist, and C. Meta-Negation: Nirv¯ana neither exists nor does not exist, and C.
But C is false, as a conception of Nirv a¯ na is an impossibility. Thus each of the four propositions is false, since for a conjunction to be false it is sufficient (and necessary, by the way) that it has a false conjunct. 14 This completes one of the tasks set at the beginning, to interpret Affirmation, Negation, Meta Affirmation and Meta-Negation as false statements. “Nirv¯ana exists” is false because what it really means is, “Nirv¯ana exists, and it is possible to have a conception of Nirv¯ana”. The same goes for the other three propositions.
¯ THE LOGIC AND LANGUAGE OF NIRVANA
101
4. Conditional definition of existence
As I also said at the beginning, to regard the four propositions as false is just one possibility. The other is to assume that they have no linguistic meaning at all. This assumption can be supported by handling the pragmatic presupposition in a slightly different way. This time the presupposition is added to the propositions, not as a conjunct, but as a condition. We get what logicians and mathematicians call a “conditional definition”. For the discussion that follows, we need to have a general definition, so we will work with our wellknown variable X. Accordingly, the presupposition that it is possible to have a conception of Nirv¯ana, which was denoted by C, is replaced by the condition that it is possible to have a conception of X, which we may abbreviate as C(X). Conditional definition: If C(X), then X exists just in case X has material or mental properties. The crucial feature of this definition is that the concept of existence is well defined (by the equivalence after the comma) only if condition C(X) is fulfilled, that is, only if a conception of X is possible. If X is identical to Nirv¯ana, however, the condition is not satisfied, since it is not possible to have a conception of Nirv¯ana. This gives rise to an interpretive problem. To solve it, let us see how it is tackled in arithmetics. The best known arithmetical example requiring a conditional definition is division. The condition is that the divisor must be unequal to zero. Patrick Suppes has suggested four ways to handle the cases in which this condition is not met, that is, in which the divisor does equal zero. For the sake of the argument, I will list these suggestions in the opposite order as they appear in Suppes’s Introduction to Logic:15 According to the first, the result of a division by zero (x/0) is not a real number, but some other object (which is the same for all numbers x). The second suggestion amounts to the arbitrary decision to make x/0 equal to 0 for all x. The third suggestion is to stipulate that x/0 is a real number, although it cannot be decided which one it is. The fourth suggestion, finally, has it that x/0 is without meaning. 16 As this paper is not on mathematics but on Buddhism, it would be pointless to deal with these suggestions themselves. Rather I will formulate analogous approaches that are applicable to our problem, the interpretation of “Nirv¯ana exists”. The idea behind this analogy is that the division by zero corresponds to the falsity of condition C(X), and that the assignment of numbers to x/0 corresponds to the assignment of truth values to “X exists”. 17 First approach: If X is inconceivable, “X exists” is neither true nor false – it has some other truth value. This means adopting a three-valued logic. The nature of the third truth value, however, is completely unclear. In the
102
PETER KÜGLER
usual semantics of three-valued logic, the third value is assigned, for instance, to statements with vague terms, statements about the future, or paradoxical statements. Accordingly, the third value is called “vague”, “undetermined” or “paradoxical”. The features denoted by these words are related to truth and falsity in one or the other way, and this is the reason why they can be regarded as truth values. To begin with vague terms: they can often be replaced by more precise terms expressing the same content in a more accurate way. These latter terms can in turn be used to build statements which are true or false. As to the second example, suppose that future events are not determined by present facts, which is why we put the label “undetermined” to statements in future tense. Even on this assumption, our predictions will come out right or wrong. As the future becomes the present, undetermination turns into truth or falsity. And finally, a paradoxical statement is such that its truth implies its falsity, and vice versa. For this reason the notion of paradox, too, depends on the notions of truth and falsity. No such relationship to truth and falsity exists in the case of “X exists”, when X is inconceivable. Neither can the sentence be replaced by another one which is true or false, nor will it become true or false as time goes by, nor do truth and falsity interfere with each other as in a paradox. So it is hard to see how the third value assigned to “X exists” could be a truth value at all. Second approach: If X is inconceivable, we could arbitrarily assign the value false to “X exists”. On the other hand, we could also choose the value true. None of these decisions is of much interest as long as it cannot be justified by good reasons. But what reason could there be for choosing true instead of false, or false instead of true? The decision seems to be completely arbitrary, so the second approach is another dead-end street. Third approach: Some may want to claim that “X exists” is either true or false, though we do not know which it is. Unfortunately this approach is as little convincing as the previous one. Provided that X is inconceivable, we are not able to give a content to any word which is intended to denote X, which becomes clear when we replace the variable by a neologism like “hadbud”. No one knows what hadbud is (as I hope, otherwise feel free to create your own word). In fact, we do not even know whether “hadbud” is a proper name or a general term. So perhaps we should put an indefinite article in front of the word and ask ourselves what a hadbud is. In any case, it is not reasonable to claim that “hadbud exists” or “a hadbud exists” is either true or false and that no decision can be made. As far as conceivability is concerned, “Nirv a¯ na” is on a par with “hadbud”, therefore the third approach has to be rejected too. The fourth approach is already implicitly present in the previous argument against approach number three. It says that if X is inconceivable, “X
¯ THE LOGIC AND LANGUAGE OF NIRVANA
103
exists” has no meaning. As Nirv¯ana is inconceivable, “Nirv¯ana exists” has no meaning and is therefore neither true nor false. The same holds for the other propositions, which are built upon the first one: “Nirv a¯ na does not exist”, “Nirv¯ana exists and does not exist”, “Nirv a¯ na neither exists nor does not exist”, are meaningless too. Compare again: “hadbud does not exist”, “hadbud exists and does not exist”, and “hadbud neither exists nor does not exist”, are as meaningless as “hadbud exists” (or “a hadbud exists”). In comparison, the fourth approach has the most plausibility, but nevertheless it depends on two assumptions about linguistic meaning. First, it must be assumed that reference presupposes meaning. Otherwise it could be objected that despite its having no meaning, the word “Nirv a¯ na” does have a referent, which would guarantee the existence of a truth value for “Nirv a¯ na exists”. The second assumption is that meaning presupposes conceivability. Only then we are justified to infer the absence of meaning from the absence of conceivability, as it was done in approach number four. 18 Reference presupposes meaning; meaning presupposes conceivability. At first sight, both assumptions seem to be alright, but there are three types of words which raise doubts about their validity. Two of them – indexicals and proper names – are discussed in the next section. They are sometimes regarded as counterexamples to the first assumption. Theoretical terms, which are the third type, may be suspected of contradicting the second assumption. This will be the topic of section 6. 5. Indexicals and proper names
The first assumption says that reference presupposes meaning. This appears to be true of all kinds of words, yet indexicals – like “I”, “you”, “here” and “now” – are claimed by some philosophers of language to have no meaning, but only reference. Moreover, it might be that the word “Nirv a¯ na” has semantic features similar to those of indexicals. Putting it more bluntly, it might be that the word “Nirv a¯ na” is an indexical term. In this case, provided that indexicals have reference without having meaning, we would not be justified to argue from the meaninglessness of “Nirv¯ana” to its having no reference. However, the premise that indexicals have no meaning is erroneous. The meaning of the word “I”, for instance, is to pick out the speaker of the assertion in which the word occurs. David Kaplan calls this the indexical’s character , which is a function that determines the referent of the indexical in varying contexts.19 Thus the character has the status of a context-independent meaning. It does not change as the context changes – only reference does.
104
PETER KÜGLER
Since indexicals have meanings which are as context-independent as the meanings of other words, they cannot be cited as counterexamples to the assumption that reference presupposes meaning. I conclude that even if the word “Nirv¯ana” were an indexical – which I would like to leave open here – we would be allowed to argue from its having no meaning to the absence of reference and truth. In contrast to familiar indexicals like “I”, “you”, “here” and “now”, the word “Nirv¯ana” has no “character” (to use Kaplan’s term) that could determine the referent of the word in a given context. So, if “Nirv a¯ na” is an indexical, it has neither meaning nor reference, and “Nirv a¯ na exists” has no truth value. Let us now turn to another argument against the assumption that reference presupposes meaning. According to a popular theory, proper names have meanings which are associated with the descriptions the speaker has in mind ¯ when he or she uses the name. If you know, for instance, that Ananda was the follower of the Buddha who was said to have had the best memory, you ¯ are able to refer to the right person by associating the name “ Ananda” with the description just given. The meaning of the name is (partly) expressed by “the follower of the Buddha who was said to have had the best memory”. In this account, it is not denied that proper names have meanings. So it does not jeopardize the assumption that reference presupposes meaning, which is why we need not consider it any further.20 What we need to consider, however, is Kripke’s rival account, often called causal, or causal-historical, theory.21 On this theory, proper names get their references fixed by an initial act of baptism, which need not be a real baptism in any ritual sense, but only the first occasion when the name is used for its object. After this occasion, the word is “handed over” from one person to the other, until it may be used by people who have never experienced, and may not even have a conception of, the carrier of the name. Thus people are ¯ able to refer to the right person by using the name “ Ananda”, even if they do ¯ not know that Ananda was a follower of the Buddha, or anything else about ¯ Ananda. Adherents of this causal theory of reference could maintain that “Nirv a¯ na” is a proper name and that the causal theory applies to it. From this they could conclude that “Nirv¯ana” possibly has reference without having meaning. This argument, however, though superficially appealing, misses an important difference between proper names and the word “Nirv¯ana”: proper names have a kind of meaning which “Nirv a¯ na” does not have. Although the causal theory implies that proper names have no contextindependent meaning – in contrast to indexicals, as we have just seen – these names possess meanings within certain contexts. In other words, proper names have context-dependent meanings, at least on some occasions. A
¯ THE LOGIC AND LANGUAGE OF NIRVANA
105
meaning of this kind exists, above all, if a speaker associates certain descriptions with the name and is able to produce these descriptions when being asked to what object the name refers. According to the causal theory, to be sure, this ability is no necessary condition for using the name correctly. But there is at least one occasion in which the usage of the name depends on the availability of descriptions of the object. This occasion is the initial baptism. As Kripke has put it, descriptions do not give meanings, but they determine the referent of the proper name in the initial baptism. 22 Proper names can only have reference, if a description of the object is available to the speaker during the baptism. Let me add two things that may be important for understanding the issue. First, when Kripke claims that descriptions do not give meanings, what he has in mind is context- independent meaning. Therefore his claim is compatible with proper names having context- dependent meanings. Secondly, among the descriptions that may be used for fixing the reference of a proper name, there are also descriptions containing indexicals. If someone says, “I call this child ¯ Ananda”, the indexical “this”, as part of the descriptive phrase “this child”, ¯ is used to fix the reference of the name “ Ananda”. In our terminology, “this child” expresses the context-dependent meaning of the name, which is only valid within the context of the baptism. In conclusion, even if we suppose that Kripke is right and proper names carry no context-independent meaning, they certainly have meanings in some contexts, which are given by the descriptions the speakers have in mind or actually use in these contexts. If there had never been any information about the object, it would not have been possible to fix the reference of the name. If Kripke is right, a single person need not have such information, but once there must have been a person who knew what he or she was talking about when using the name for the first time. This is the crucial difference between typical proper names and the term “Nirv¯ana”. Given that nobody (including the Buddha) ever had a conception of Nirv¯ana, nobody ever knew what he or she was talking about when using this word. Nobody (including the Buddha) has ever been able to provide a description of Nirv¯ana that could have been used to fix the reference of the word. So, on Kripke’s account, if “Nirv a¯ na” is a proper name, it is a name without reference. And as there is no reference, there is also no truth-value of “Nirv¯ana exists”. In this section, we considered the idea that “Nirv¯ana” is either an indexical term or a proper name. We have discovered differences which may persuade us that “Nirv¯ana” must not be assigned to any of these categories. But even if we treat the word as an indexical or as a proper name, we are justified to argue from its meaninglessness to its having no reference. Because also in
106
PETER KÜGLER
the case of indexicals and proper names, reference presupposes some kind of meaning, which “Nirv¯ana” does not possess. 6. Theoretical terms
After having defended the assumption that reference presupposes meaning, let us consider the second assumption. In order to give meaning to a word, it is necessary to have a conception of its object. It is only on this supposition that the inconceivability of Nirv¯ana yields the conclusion that the word “Nirv¯ana” has no meaning. In the eyes of some philosophers, however, this second assumption is threatened by theoretical terms. After all, we seem not to be able to build a full conception of the objects of some theoretical terms. According to quantum theory, for instance, there is an uncertainty in the position and momentum of a particle. They cannot both be determined simultaneously. The particle has no definite position at a given time. This is not meant in a purely epistemological sense. It is not that the particle’s position is definite yet unknown to us. Rather there is no such thing as a definite position at a definite time. We are not able, however, to conceive a particle with indefinite position. To be sure, we often imagine things without imagining their positions in space. I may think of a friend without thinking of where he is right now. But in cases like this, we assume, implicitly at least, that the object of our thought does have a definite position. I believe that my friend is somewhere. I do not believe that he has an “uncertain position” or something like that. Indeed, the latter would not make much sense. But even if we have no conception of a particle, say an electron, with an indefinite position, the word “electron” certainly has meaning, whatever this meaning may be.23 So it seems that meaning is not necessarily accompanied by conceivability. And as to the word “Nirv¯ana”, supposing that it is a theoretical term, we cannot exclude that it has semantic features similar to those of “electron”. Just as the inconceivability of electrons would not imply that statements about the existence of electrons are meaningless, the inconceivability of Nirv¯ana would not imply that the sentence “Nirv a¯ na exists” is meaningless. To evaluate this idea, we have to reconsider the so-called “inconceivability” of electrons as compared with the inconceivability of Nirv a¯ na (having been developed in section 3). Again, as in the case of indexicals and proper names, this comparison will reveal an important disanalogy. It is probably true that we have no complete conception of electrons, given that nobody is able to imagine an electron having no definite position. But nevertheless we have a partial conception. We know, for instance, that electrons are much smaller than rocks, pebbles and grains of sand. We also know
¯ THE LOGIC AND LANGUAGE OF NIRVANA
107
what it means that an electron has a definite position. If we did not know this, we would not be able to realize the conceptual difficulties connected with the absence of a definite position. Moreover, there are mathematical descriptions of electrons in quantum physics, and finally, the word “electron” is used for various explanatory purposes. That a lamp lights up when the switch is turned on, or that a Geiger counter is clicking, can partly be explained by moving electrons. By studying empirical facts, we get an idea of how electrons interact with other parts of nature. This, too, adds to our conception of electrons, which might necessarily be a partial one – but a conception it is. Nothing like that with Nirv¯ana. Here, not even a partial conception is possible. At best there are inadequate metaphors. Sometimes Nirv a¯ na is compared to the void. The only comprehensible notion of the void, however, is that of an empty space. Yet Nirv a¯ na certainly is no empty space, for it is no space at all. Furthermore, if Nirv a¯ na is identified with nothingness, as in some philosophical interpretations of Buddhism, it must be acknowledged that the concept of nothingness depends on the concept of non-existence. If it makes sense at all to talk about nothingness, it must be identified with the non-existence of anything whatsoever. In short, to talk about nothingness is a metaphysically queer way of talking about a situation in which nothing exists. Thus, if Nirv¯ana could be identified with nothingness, it would be correct to claim that the Tath¯agata, who has reached Nirv a¯ na, does not exist after death. But this is explicitly rejected by Buddhists, which means that the identity of Nirv¯ana and nothingness is just another wrong interpretation of Buddhist thought. Metaphorical accounts of Nirv¯ana do not even provide partial or somewhat inaccurate conceptions, rather they are completely beside the point. Hence they cannot be equated with scientific models of subatomic particles, which do provide partial conceptions of the material world. Moreover, the notion of Nirv¯ana does not enter into explanations, although the word “Nirv a¯ na” is sometimes misused in this way. There is, for instance, the question of reincarnation. If it is claimed that a person is not reborn again, and that this is because he or she has reached Nirv¯ana, the word “Nirv¯ana” is employed like a theoretical term to explain an aspect of reincarnation. But this explanation, again, reflects an inadequate understanding of Buddhism. In Buddhist scriptures, Nirv¯ana is not meant to be a tool of explanation, but a worthwhile goal of human conduct and a result of right action and right thought. (We will see below, however, that these descriptions are questionable too.) Summing up this section, since the word “Nirv¯ana” is not used as theoretical terms are being used, we should not assimilate its semantics to that of theoretical terms. Nirv¯ana cannot be conceived of by the human mind, not even in a partial or inaccurate way; nor do statements containing the word
108
PETER KÜGLER
“Nirv¯ana” explain anything. Hence the inference from the inconceivability of Nirv¯ana to the meaninglessness of the word “Nirv¯ana” cannot be objected on the ground that “Nirv¯ana” is a theoretical term. 7. The raft and the other shore
This paper was devoted to the question why Buddhists refrain, or should refrain, from investigating the four propositions listed at the outset. The first step in answering this question was to give two tentative definitions of existence, following the idea that what is to be defined is a second-order concept. Moreover, when people talk about the existence or non-existence of things, the pragmatic presupposition is made that the speaker has a conception of what he or she is talking about. By combining this presupposition with the four propositions about Nirv¯ana, and by adopting the Buddhist view that Nirv¯ana is inconceivable, we were able to interpret all propositions as being false. Alternative interpretations derived from a conditional definition of existence. After discussing several approaches to what happens when the condition of this definition is not satisfied, I suggested that in this case any ascription of existence or non-existence is without meaning. More precisely, the argument went from the inconceivability of Nirv a¯ na to the meaninglessness of the word “Nirv¯ana”, and in turn, from the meaninglessness of the word to its having no reference. Because “Nirv a¯ na” has no reference, the four propositions are neither true nor false. In defense of this reasoning, I dwelled on three types of words which are sometimes said to have reference without having meaning, or to have meaning but no conceivable object. I argued that even if the word “Nirv¯ana” were an indexical term or a proper name, it could only have reference if it had meaning. Furthermore, disanalogies between the usage of the word “Nirv¯ana” and the usage of theoretical terms suggest that “Nirv¯ana” does not belong to the latter. All in all, the inconceivability of Nirv¯ana implies that the word “Nirv a¯ na” does not refer to anything. On the other hand, the word “Nirv a¯ na” has been used by many people, in verbal and written communication, and it is still being used so. This indicates that the word probably does have a linguistic function. To delineate this function, we may think of the well-known parable of the raft, presented in the P a¯ li Canon of Therav¯ada Buddhism,24 as well as in the Diamond Sutra, which is one of the most influential texts of Mah a¯ y¯ana. According to these texts, the Buddha has taught that his teaching should be regarded as a raft which must be left behind after the river is crossed. In this picture, most scholars of Buddhism identify Nirv¯ana with the other shore that has been reached by the traveler, implying that the word “Nirv¯ana” denotes the other shore.
¯ THE LOGIC AND LANGUAGE OF NIRVANA
109
Yet the analysis given in the previous parts of this paper suggests a different interpretation of the parable. Although the word “Nirv a¯ na” may be regarded as a tool, it is none of description. To put it differently, the word does not refer to the other shore, nor does it refer to the raft, it is the raft. Speaking less metaphorically, the word has no referential (descriptive) function; it only has transformative function, as Frederick Streng called it. The word “Nirv¯ana” is “useful to effect a spiritual change”, because “apparently descriptive terms as ‘Buddhahood’, ‘Tathagata’ . . ., ‘Emptiness’, and ‘allknowledge’ do not primarily function to describe an assumed subjective state or objective entity; rather these terms are often used to evoke insight into the conditions of experience which (when combined with meditation and moral action) will free a person from attachments to conventional fears and false expectations.”25 This quote from Streng’s ‘Language and Mystical Awareness’ needs just one qualification: whereas Streng thought that descriptive and transformative functions “can be found to some degree in the same expression” (p. 150) and that the terms listed above do not primarily function in a descriptive way, I have argued that “Nirv a¯ na” has no descriptive function at all. The transformation of life and personality is the sole aim of the language of Nirv¯ana. Notes 1. Dheli: Motilal Banarsidass, 1993. 2. Cf. Samyutta Nikaya ¯ LVI.7 and 41; Majjhima Nikaya ¯ 63 and 72. See also Mircea Eliade, Yoga: Immortality and Freedom (Princeton, NJ: Princeton University Press, 1969), p. 166. Assuming the distinction between Nirv¯ana and Parinirv¯ana, or the state of the Buddha after his enlightenment and the state of the Buddha after his death, the four propositions must be read as dealing with the latter. 3. Cf. Eliade, op. cit ., pp. 173–177. 4. Philosophical Investigations (Oxford: Basil Blackwell, 1953), §138. 5. Moses Maimonides, The Guide for the Perplexed (New York: Dover, 1956), I 58. 6. For our purposes, we need not settle the question whether it is possible at all to have only one property. 7. What about fictitious objects? Does Pegasus have the material property of being a horse (which would imply its existence)? No, Pegasus just has the subject-relative property of being conceived as a horse. A horse it is only in a derivative sense. 8. Cf. Heinrich Zimmer, Philosophies of India (Princeton, NJ: Princeton University Press, 1951), pp. 540–541, fn. 88. Although it is straightforward to assign material properties to r¯ upa and mental properties to the other four skandhas, this division is not essential to our investigation. All we need to assume is that every material or mental property can be found somewhere within the five skandhas.
110
PETER KÜGLER
9. A view which has also been held by Bolzano, Frege and Russell. Cf. Edgar Morscher, ‘Was Existence Ever a Predicate?’, Grazer Philosophische Studien 25/26 (1985/86): 269– 284. 10. Thus existence is not identical to the property of being causally efficient, which has been suggested in ancient Greek philosophy as well as in some schools of Mah¯ay¯ana Buddhism. Cf. Plato’s Sophist 247e and Stcherbatsky, op. cit ., I, p. 69. 11. Zimmer, op. cit ., p. 312. 12. Robert C. Stalnaker, ‘Pragmatic Presuppositions’, in Milton K. Munitz and Peter K. Unger (eds.), Semantics and Philosophy (New York: New York University Press, 1974), pp. 197– 213; p. 200. 13. Cf. David J. Kalupahana, The Principles of Buddhist Psychology (Albany, NY: State University of New York Press, 1987), chapter 6, which also includes a comparison of the six sense organs, and their objects, with the five skandhas. 14. Though I will later reject the interpretation of “Nirv¯ana exists” in terms of three-valued logics, it is worth noting that in these systems a conjunction is usually assigned the value false if it has a false conjunct. So even if “Nirv¯ana exists” had a third value, “Nirv¯ana exists, and C” would be false. 15. Princeton, NJ: Van Nostrand, 1957. Cf. pp. 166–168. 16. There is also a fifth approach in Suppes’s book, which is a method of doing away with the condition and using an explicit definition instead. In this respect, and because “x/0 = y” is false for all x and y, the fifth approach is akin to the strategy adopted in the previous section. 17. More precisely, 0 corresponds to the value false, all other real numbers to the value true. If x/0 is claimed to be no real number at all – as in the first approach – “X exists” has a third truth value in the analogy. 18. As to the second assumption, remember that one may conceive an object although it does not exist, since “being conceived” is a subject-relative property. For instance, it is possible to conceive the object of the word “Pegasus”. 19. Cf. David Kaplan, ‘Demonstratives’, in Joseph Almog, John Perry and Howard Wettstein (eds.), Themes From Kaplan (New York: Oxford University Press, 1989), pp. 481–563. 20. Although reference presupposes meaning, there can be meaning without reference. The non-denoting term “Pegasus” is associated with certain descriptions, in contrast to the non-denoting term “Nirv¯ana”. It is possible to describe Pegasus, but it is not possible to describe Nirv¯ana. So the semantics of “Nirv¯ana” is different from the semantics of fiction. 21. Cf. Saul A. Kripke, ‘Naming and Necessity’, in Donald Davidson and Gilbert Harman (eds.), Semantics of Natural Language (Dordrecht: Reidel, 1972), pp. 253–355. 22. Cf. Kripke, op. cit ., pp. 258–259 and 276–277. 23. I hope the reader will accept the premise that theoretical terms have meanings. A full account of the semantics of theoretical science would go far beyond the scope of this paper. 24. Majjhima Nikaya ¯ 22; cf. Zimmer, op. cit ., pp. 477–478. 25. Frederick J. Streng, ‘Language and Mystical Awareness’, in Stephen T. Katz (ed.), Mysticism and Philosophical Analysis (London: Sheldon Press, 1978), pp. 141–169; p. 154.
