IN A RECENT BOOK--perhaps the most outstanding work of Buddhist scholarship
in recent decades--Professor Jayatilleke advances the thesis that there is a concept of meaningless statement in the Paali Nikaayas.
"The Buddha," he says (para. 536), "refers to statements (bhaasita.m)
of a certain character as 'appaa.tihiirakata.m ... sampajjati' [Diigha Nikaaya 1.
193, 194, 195, 239, 241, 242, 243, 244; Majjhima Nikaaya ii. 33, 41] and
statements of the opposite character as 'sappaa.tihiirakata.m (bhaasita.m)
"(Diigha Nikaaya 1. 198). His theory, in its initial formulation, is that
statements characterized in the former way "are in some sense 'tacking in
meaning' " (para. 540).
He first considers and rejects a number of translations and commentarial explanations of the phrase, and he finally settles on one commentarial explanation, Buddhaghosa's, as fitting the contexts best: "amuulaka.m niratthaka.m Sampajjati," which he translates as "becomes baseless and meaningless." His method of establishing his theory is to collect the contexts in which the phrase appears and is applied to statements, and then to see which translation makes the best sense of the phrase in those contexts. It is a method we shall ourselves follow in criticizing his theory.
AUTHOR'S NOTE: I should like to thank Miss I. B. Horner for her kindness in consenting to read the paper from the point of view of Paali scholarship, This is not to be construed, however, as implying her endorsement of the paper's contents.
K. N. Jayatilleke, Early Buddhist Theory of Knowledge (London: George Allen & Unwin Ltd., 1963). References to the book are to paragraph numbers.
 Two of the translations which Professor Jayatilleke reports are incorrectly given: "Talk without ground," which he attributes to Professor T. W. Rhys Davids, should be "talk without good ground"; and "incomprehensible talk," which he attributes to Miss I. B. Horner, should be "irresponsible talk" (para. 540A, I[a]).
There are six main contexts in which the phrase appears, but it is the first to which Jayatilleke devotes his greatest attention, and it is in his analysis of the first that we find his theory set forth in greatest detail. According to him, the standard example of an appaa.tihiirakata.m bhaasita.m is the following statement: "I like and am in love with the beauty queen of this country." He asks us to suppose that "the beauty queen of this country" has a denotation-that is, that there is someone answering to the description; and then he asks whether the statement is meaningful. According to him, the Paali Nikaayas say that the statement is meaningless under certain conditions: ". . . the context can render such a statement meaningless" (para. 548). In the case of the particular statement under consideration, "The context is one in which the person who makes this statement confesses that he does not know whether this beauty queen ... was 'a k.satriya, a brahmin, a vai`sya or `suudra'... does not know 'what family or personal name she had,' 'whether she was tall, short, dark, brunette or golden in colour or in what village or town or city she dwells'.... In other words he claims to like and love a person whom he has not 'seen or known' " (para. 548).
More generally and schematically: "X's statement 'I love Y' is meaningless since (i) one is not sure whether there is an instance of Y, and (ii) even if there is, it does not make sense for X to say that he loves Y unless he has some acquaintance direct or indirect with Y, such that he could specify at least one of the characteristics of Y.... In other words, there is no verifiable content to the statement from the point of view of X who is making it" (para. 549).
The resemblance of this account to the positivistic verifiability criterion of meaning is obvious, and it is a resemblance to which Professor Jayatilleke explicitly draws our attention (para. 558). It comes out perhaps more explicitly in the refrain with which he concludes his examination of each of five contexts of the phrase. Thus, in the case of the example of the beauty queen: "The Paali Nikaayas seem to be saying that a statement in which no verification or meaning is attached to one of its terms (by the speaker) is in fact meaningless" (para. 550); in the case of the example of the stairway to the mansion: "His statement is considered meaningless since he can attach no meaning or verification to the term 'mansion' " (para. 551) ; in the case of the example of the Brahmans on Brahmaa: "This statement is considered
Jayatilleke, op. cit., paragraphs 546, 550, 551-554.
I have omitted the sixth context because I do not understand the paragraph in which it occurs.
This seems a peculiar use of the term "verification"; in its usual sense it makes no sense to apply it to terms but only to statements.
meaningless, since no meaning or verification is attached by those who make it to the term 'Brahmaa' " (para. 552) ; in the case of the example of the soul after death: the statement is "considered meaningless ... because those who make it ... can attach no verifiable content to the statement or to the concepts in it" (para. 553) ; and finally, in the case of the example of the highest color: "... the statement, 'that colour than which there is no other colour which is higher or better is the highest colour' is said to be 'meaningless' ... because what is meant by 'the highest colour' is not specified. No meaning by way of anything verifiable is attached to the phrase 'the highest colour' in this context" (para. 555). From this recital we can easily see that "all the ... statements have a certain characteristic in common, namely that no verifiable content is attached by the speaker to some of the symbols or words contained in them" (para. 556).
There are, however, differences between the two criteria of meaning, according to Jayatilleke, and it will be instructive to consider them. The first is that the positivistic criterion specifies, while the Nikaayas do not, that verification be in terms of sense-experience. More important, for our purposes, is the second difference: that the positivist criterion makes no reference, while the Nikaayas do, to utterer and utterance-context, "In the above contexts," writes Jayatilleke, "the statements were not considered apart from the speaker and the context in which they were uttered." Thus a statement, for the Nikaayas, is not, as it is for the positivist, intrinsically meaningless but only relatively so: in relation to one speaker and one context it will be meaningless; in relation to another speaker and context it will not. We seem thus to have a relational or relativistic criterion of meaning in contrast to the absolutistic criterion of the positivists.
We have to be clear about this relativistic criterion, for there are numerous points at which Jayatilleke ascribes a nonrelational conception to the Nikaayas: statements are said to be meaningless simply, without any reference to utterer and utterance-context. They are not characterized as meaningless-in-relation-to- one-utterer-and-his-utterance-context, and thus meaningful in relation to another. Just how, then, are we to understand this conception? So far as I can see, it means just two things: (1) either the speaker can attach a meaning to all the terms of his statement, or he cannot attach a meaning to one or more of them; and (2) either the speaker can correctly claim that he has
It is perhaps worth noting the shift from verification to verifiability; two very different criteria.
Jayatilleke, op. cit., paragraph 558. We may note in passing that neither were they ever so considered according to the positivistic criterion. But limitations of space prevent extended clarification of the point.
See the texts quoted above, pp. 68 f.
"seen or known" the subject of his statement, or he cannot. The former concerns the utterer, and the latter the utterance-context. The reference to the speaker is a reference to his ability to attach a "verifiable" meaning to the terms of his statement; the reference to the context is a reference to the speaker's perception or knowledge of the subject of his statement. If the speaker cannot correctly claim that he has seen or known the subject of his statement, and if he cannot attach a verifiable meaning to one or more of the terms of his statement, then his statement is meaningless-under-those-conditions; otherwise meaningful.
That Jayatilleke's theory is unsound appears from a number of considerations. There is, first, the fact that it seems to be based on a misinterpretation of the contexts in which the disputed phrase occurs. Thus for him all the contexts are either implicitly or explicitly semantical; this appears throughout his discussion. But it seems clear from a closer reading of those contexts that the discussions in them are all, on the contrary, epistemic or evidential. The considerations to which appeal is made and on the basis of which the questionable bhaasita.m is characterized as appaa.tihiirakata.m are invariably epistemic or evidential ones. The complaint invariably is that the person involved is ignorant of what he is talking about, and the complaint invariably concludes that he has an attitude toward something he has not seen or known. There is no word about meanings, no reference anywhere to terms or statements, no protest at any point that the various persons involved are using meaningless terms or uttering meaningless statements. Thus, as anyone may verify for himself, the Canonical passages say nothing about the term 'mansion,' they say nothing about the term 'Brahmaa,' nothing about the term 'extremely happy,' and nothing about the term 'the highest colour'; nor do they say anything which would warrant Jayatilleke's conclusion, for each statement containing these terms, that the disputed "statement is considered meaningless since (the utterer of it) can attach no meaning or verification to the term." Whatever the Canonical passages say of those statements, they say nothing about the terms they contain.
The treatment of the example concerning "the highest colour" is a particularly clear case of the systematic misinterpretation of the Canonical passages. For according to Jayatilleke the statement "That colour than which there is no other colour which is higher or better is the highest colour" is said to be meaningless (appaa.tihiirakata.m) "because what is meant by 'the highest colour' is not specified" (para. 555). But this seems clearly erroneous on several
Limitations of space exclude extensive quotation of the relevant passages as evidence for the following arguments, but the interested reader may find them in the Paali Canon or in Jayatilleke's book.
counts. For, first of all, precisely what is specified by that statement is the meaning of "the highest colour": the statement is nothing else than a definition of the term. Professor Jayatilleke himself points this out when, characterizing the statement as meaningless, he protests that it is a tautology or a definition "when the answer [to the question asked] should be in the form of a contingent proposition." Meaningless such a statement may be, according to some theories of tautology and definition; but that it is a definition, and thus a specification of the meaning of "the highest colour," is undeniable. What the reply, by means of this definition, to the Buddha's request (for a specification of the highest color) does not do is simply to specify the highest color. As Jayatilleke's own translation (para. 554) puts the Buddha's complaint: "But you do not specify that colour" (italics added); and as I. B. Horner's translation of the Majjhima Nikaaya (Middle Length Sayings ii. 230) puts it: "but you do not point to this lustre" (italics added). Thus the Buddha's complaint is not that the meaning of "the highest colour" is not specified but that the highest color--the highest color itself--is not specified. The discussion never rises to the semantic level; the reference is never to meanings. The discussion here, as in all the other relevant contexts, remains on a factual level; the reference is always to things.
That the discussion of the statements cited is mostly if not always epistemic and not semantic is, inconsistently enough, admitted by Jayatilleke; for he explicitly recognizes that "the statements which are called appaa.tihiirakata--are baseless (amuulaka.m ...) in that the persons making them do not do so on the basis of any evidence or arguments which justify their assertion" (para. 557). The inconsistency is evident: baselessness implies meaningfulness. Only a meaningful statement can be baseless. To hold that a statement is baseless is to admit that it is meaningful, for it makes no sense to say of a meaningless string of sounds or marks that it is baseless. A baseless statement is, in the context of an argument, simply a meaningful one for which no ground or reason or evidence has been advanced.
Jayatilleke concludes, however, that the admitted baselessness of these statements is "incidental in that this explanation does not hold good for the example" of the highest color (para. 557). But we may note that if the "baseless" explanation does not hold for the single example of the highest color, the "meaningless" explanation cannot hold for the other five examples; and since there are only six examples, his theory, at best, fails for five-sixths of the cases.
By contemporary logical usage it would be reckoned as a definition, not a tautology; for a tautology, in that usage, is a truth-functional logical truth, which this statement is not.
It is not the case, however, that the "baseless" interpretation fails for the example of the highest color; for the Buddha's complaint, as we saw earlier, is not that the meaning of a term is not specified but that a thing is not specified: the highest color. In the context of the discussion concerned, specification of the highest color would, if otherwise satisfactory, have provided a basis for Udayin's claims as to the nature of his teacher's teaching. To have specified the highest color (in Jayatilleke's translation) or pointed to the lustre (in Miss Horner's) would have been to have provided a basis for his contention; it would have been to have provided the very subject of his teacher's teaching. To have failed to do that would have been to have failed to provide even a subject for discussion. To reply continually in terms of a definition to the Buddha's continual request for a specification of the subject was continually to fail to provide a subject for discussion; it was continually to fail to identify the subject of his teacher's teaching. For a definition tells us nothing about the world; it informs us only about the meanings of words. It is perfectly possible, without any experience or knowledge of anything whatever, to know and to assert--as Udayin repeatedly does--that the highest color is the color than which no other color is higher. One needs no experience or knowledge of the highest color to know and to be able to say this; for in saying in these terms what the highest color is one is merely giving its analytical equivalent. Thus, to say that the highest mountain in the world is the mountain higher than any other is not in the slightest to identify that mountain for us; it is merely in effect to tell us what "the highest mountain" means: it means "the mountain higher than any other." But one needs no knowledge of the mountains of the world to know this. To identify the highest mountain, however, is a very different matter; it requires knowledge of all the mountains of the world and of their relative heights. And if the question is which is the highest mountain in the world, the most fundamental basis for discussion would be a proposed identification of that mountain. Thus Udayin's replies were repeatedly baseless.
I have so far argued that Professor Jayatilleke's theory is itself baseless and inconsistent. But perhaps the most serious objection to it is that it saddles the Paali Nikaayas with what seems to be an unsound criterion of meaninglessness. To see this we must recall the general schematic formulation of the criterion (p. 68, above) and generalize it a little further.
There are in any well-developed language a number of terms, in important respects like "liking" and "loving," which in contemporary Western philos-
From all this it may be seen that where Jayatilleke finds the similarity of the examples of the beauty queen and the highest color to reside in their meaninglessness, I find it in their baselessness.
ophy are called intentional words: e.g., "believing," "desiring," "seeking," "expecting," "thinking of" and the like. These words express attitudes or states of mind which have the following important characteristic: one can have such psychological attitudes toward an object even though the object does not exist. As Professor R. M. Chisholm puts it:
Psychical phenomena, according to Brentano's thesis, are those "which intentionally contain an object in themselves." ... The psychical phenomena which most clearly illustrate his thesis are what are sometimes called psychological attitudes, e.g., believing, desiring, hoping, wishing, and the like. When he said that they are characterized by "intentional inexistence," he was referring to the fact that these attitudes can be truly said to have objects even though the objects which they can be said to have do not exist. Even if there weren't any honest men, for example, it would be quite possible for Diogenes to look for one. Diogenes' quest has an object, namely an honest man, but, on our supposition, there aren't any honest men. The horse can expect to receive his oats in ten minutes, even though the event which he expects does not in fact occur. William James would be able to believe that there are tigers in India, even if there weren't any tigers in India. Even if there aren't any disembodied spirits, it is quite possible for someone to take something to be one. But mere physical phenomena, on the other hand, cannot thus "intentionally contain an object in themselves." In order for Diogenes to sit in his tub, for example, there must be a tub for him to sit in. In order for the horse to eat 'his oats, there must be oats for him to eat. In order for William James to shoot a tiger in India, there must be a tiger for him to shoot. And so on.
Without in any way distorting Jayatilleke's formulation of his criterion, we may, I think, generalize and reformulate it as follows: "X's statement, 'I have a certain psychological attitude (liking, loving, believing, desiring, seeking, expecting, thinking of, etc.) toward Y' is meaningless if (i) X is not sure of the existence of Y, and (ii) even if Y exists, X lacks any acquaintance direct or indirect with Y such that he could specify at least one of the characteristics of Y."
Now let us consider the two clauses one by one. According to (i) X's doubt about the existence of any object Y toward which X purports to have a psychological attitude renders the psychological attitude impossible; for it renders meaningless any statement in which X might attempt to express that attitude. Thus for anyone to say that he believes in without being sure of the existence of God is impossible, for the corresponding propositional formulation of that belief is meaningless. All belief is ruled out, for all belief, by definition, is unsure of the existence of the thing it believes in. If belief were certain, it wouldn't be belief but knowledge. But if doubt about the existence
R. M; Chisholm, "Intentionality and the Theory of Signs," Philosophical Studies, III (June, 1952), 56-63; the quoted passage is on pp. 56-57.
of anything renders a psychological attitude toward it impossible---or the corresponding propositional formulation of it meaningless--then much of scientific investigation would be impossible; for theoretical scientists postulate objects about whose existence they are in doubt; explorers seek legendary cities about whose existence they are in doubt; detectives hunt murderers about whose existence they are in doubt. In many cases the postulated objects are found; the cities are discovered; the murderers are captured. But according to the first clause of Jayatilleke's criterion, all these activities are impossible and the corresponding propositional expressions of them meaningless.
According to (ii), X's inability to specify even a single characteristic of Y on the basis of direct or indirect acquaintance with Y renders it impossible for X to have a psychological attitude toward Y. Perhaps the first thing to observe about this clause is that it seems unwarranted by the Canonical text; it goes far beyond it. All that the Canonical passage (Diigha Nikaaya 1. 241) shows is that the person purporting to be in love with the beauty queen is ignorant of a good many--it never says all-of her characteristics (social class, name, height, complexion, and geographical location), and then it concludes that he claims to like and love someone he has neither seen nor known. This pretty clearly implies "direct acquaintance," but it is far from clear that it implies "indirect acquaintance." But let us evaluate the condition itself, regardless of its basis in the Canon. Here again the condition has two parts, so we shall consider them one by one. According to (ii, a) if X cannot specify at least one characteristic of Y on the basis of direct acquaintance with Y, then X cannot have a psychological attitude toward Y. In this context "direct acquaintance" with Y presumably means direct sensory or extrasensory experience of Y.
To see that this is unacceptable we have only to remember that no Buddhist now living--excluding those rare persons, if any, whose memories extend back over sufficient lifetimes--has any direct acquaintance, in this sense, with the Buddha. Yet he believes in the existence of the Buddha; he reveres the Buddha; he has various other psychological attitudes toward the Buddha. But nothing is changed if we shift our discussion to the psychological attitude consisting of love toward a beloved. The belief and reverence of living Buddhists toward the Buddha are aroused by documents, the Suutta and Vinaya Pi.taakas, purporting to be true records of the Buddha's sayings and doings. The love of a man toward a woman may be aroused quite as easily
You can know about, and you can know a great deal about, someone you have never seen or known.
Jayatilleke does not say what he means by this crucial phrase. This is a part of his theory which might well be explained and clarified.
by documents, e.g., letters and books, purporting to be true descriptions of the loved person. The recipient of letters or other such documents glowingly but falsely describing a woman as a person of great beauty, charm, and nobility may succumb to the description and fall in love with the woman without ever having seen or known her. History and literature are full of examples of the phenomenon. We admire Socrates on the basis of Plate's account, scorn him on the basis of Aristophanes', and doubt his very existence on the basis of contemporary historical investigations; yet our psychological attitudes--admiration, scorn, and doubt--remain, generated not by direct acquaintance (none of us has ever seen or known him) but by historical or semi-historical or even wholly fictional literary or epistolary accounts. In none of these cases is it possible to say that the psychological attitude is aroused by direct acquaintance with the object of the attitude; in none of these cases is there any relation of direct sensory or extrasensory experience. In none of them is it possible to say that X can correctly specify at least one of Y's characteristics on that basis; but that the lover feels love toward the object purportedly described, as we feel admiration toward Socrates and reverence toward the Buddha, is undeniable. But such psychological attitudes are impossible according to the first part of Jayatilleke's second condition.
Appeal to the second part (ii, b) of the condition will not extricate us from the essential difficulty. According to it, X must be able correctly to specify at least one of Y's characteristics on the basis of indirect acquaintance with Y. What indirect acquaintance would be Professor Jayatilleke has not said, We may suppose, however, that by indirect acquaintance with Y he means direct acquaintance with other items evidentially related to Y--evidences of Y's existence or character not based on direct sensory or extrasensory experience of Y: historical records, contemporary documents and reports, monuments, tablets, relics, personal effects, and in general all evidences of the kind sought out by archaeologists, historians, reporters, detectives, and so forth.
One difficulty with this condition, as so formulated, is that it rules out the possibility of total error about Y. On the basis of his perception and interpretation of the items evidentially related to Y, X's conclusions about Y's character may be completely mistaken: Y may have none of the characteristics X thinks he has. This might happen for either of two reasons: the evidence is misleading, or the evidence, though itself not misleading, is misinterpreted by X. In either case he will come to conclusions at variance with Y's actual character. If a man works single-mindedly to accumulate great wealth, we naturally tend to think of him--because the evidence is misleading--as greedy
 See previous footnote.
for money, power, and luxury; but he may be working selflessly, while living ascetically, to leave his wealth to the poor. When wealthy men donate large sums of money to charity, it is natural to infer philanthropy; but they may be trying to escape taxation. When film stars donate their services to soldiers at war, it is natural to infer generosity; but they may be seeking publicity. But if we can be mistaken about any characteristic of Y, on the basis of items evidentially related to Y, we can be mistaken about all: we may be unable to specify a single characteristic of Y's correctly.
But even more basically, indirect acquaintance itself is not necessary for the existence of psychological attitudes. As we saw in the quotation from Chisholm, we can have psychological attitudes toward objects even though those objects do not exist. To say that there are items evidentially related to Y is to assume the existence of Y, and that is an assumption that is not necessary to the existence of psychological attitudes. Whether God exists or doesn't--though we may believe He does-we have various psychological attitudes toward Him. Whether Socrates existed or didn't--though we may believe he did--we admire or scorn him, depending on whether we trust Plato's or Aristophanes' portrait. Whether Gautama, or Jesus, or Shakespeare ever existed or didn't--though we may believe they did--we venerate them. But the possible discovery some day that no such persons ever existed will not affect the fact that while we believed in their existence we had various other psychological attitudes--liking' and loving among them--toward them. Thus it is not only unnecessary, for X to have a psychological attitude toward Y, to be able to specify at least a single characteristic of Y; it is unnecessary to have any indirect acquaintance at all. The existence of Y is unnecessary for the existence of a psychological attitude toward Y. Thus it does make sense for X to say that he loves Y even if he has no acquaintance direct or indirect with Y and cannot correctly specify any of Y's characteristics. That is the peculiar nature of psychological attitudes.
If all this is so, then the criterion of meaninglessness which Professor Jayatilleke ascribes to the Paali Nikaayas is an unsound criterion. But I see no reason to believe that the criterion is to be found there.