English Section

      Buddhism Today 

Vietnamese Section


...... ... .  . .  .  .
By Henry Cruise

One of the most recent major writers on Buddhism, David Kalupahana, has likened the approach of Early Buddhists(1) to that of the Logical Positivists: ... the Buddha confined himself to what is empirically given. Following a method comparable to that adopted by the Logical Positivists, he sometimes resorted to linguistic analysis and appeal to experience to demonstrate the futility of metaphysics.(2)

Whether the Buddha demonstrated the futility of metaphysics will not be our concern here. Our main concern will be to elucidate

I. The "empiricist" approach of Early Buddhism, and

II. What this approach lead to in the areas of

1. Causation

2. "Things" involved in causation (dharmas)

3. Nirvaa.na, Self, and the "unanswered" questions.



Early Buddhism rejected both authority and reason (specifically a priori reasoning), either separate or together, as sufficient bases for knowledge. They were rejected because, according to Jayatilleke, "beliefs based on authority or reason may turn out to be true or false."(3)

Authority and/or reason may give us true beliefs, but that they are true is not guaranteed by their being derived from reason and/or authority.

What guarantees the truth of a belief, and what constitutes knowledge, is that one has "'personal knowledge'... of it, taking into account the views of the wise."(4)

"Personal knowledge" of something (say 'P') would seem to be a necessary but not always sufficient condition for one to be said to have knowledge of 'P'. That a view is held by the wise does not appear to be a sufficient condition for one to be able to claim a view as knowledge, but it would seem in some cases that it is a necessary condition. How should we take this?

It should be noted that there would be a problem with the above if too much emphasis were placed on "the wise," for then one would be left with the question of how one can tell who are the wise, other than by ascertaining who possesses knowledge, and consequently be open to a vicious circularity. Given this and the already mentioned Buddhist rejection of authority, it seems reasonable to take the reference to "the wise" as a reminder to check one's personal knowledge with the personal knowledge of others.

The important question that arises is, what is to count as "personal knowledge"? Both Kalupahana and Jayatilleke agree that "personal knowledge" is acquired through perception, ordinary and extra-sensory, as well as by inference derived from such perceptions.

There are said to be six forms of higher knowledge (abhi~n~naa) one can acquire on reaching the fourth jhaana, or stage, of meditation. These are:

(i) iddhividhaa - psychokinesis

(ii) dibbasotadhaatu -clairaudience

(iii) cetopariya~naa.na - telepathic knowledge of various kinds

(iv) pubbenivaasaanussatin~naana - knowledge of past lives

(v) dibbacakkhu - knowledge of the decease and arising of beings

(vi) aasavakkhaya~naa.na - knowledge of the destruction of the defiling impulses.(5)

As Jayatilleke points out,(6) (i) is a matter of "knowing how", rather than "knowing that." We should also note that the rest, (ii)-(vi), are not different forms of knowledge per se, but different forms of perception.

I may have the ability to perceive that I have destroyed the defiling impulses, but, that I have so destroyed them, and further, known that I have, would appear to be separate things again. Similarly, I may have the ability to read minds, hear voices at a distance, and so forth, but this is distinct from any particular piece of knowledge I may acquire from such abilities.(7)

Jayatilleke claims, justifiably I think, that Buddhism "makes less obvious the gap between the empirical and the mystical."(8) The powers mentioned above are not seen in any way as ''supernatural," and the conditions for their attainment are listed by the Buddha. Both Jayatilleke and Kalupahana spend time discussing these psychic powers, but they ignore the higher stages (5th-9th) of Buddhist meditation and their role in Buddhist knowledge; Although these higher stages do not appear necessary for the verification of the four noble truths, they do seem to play a part in Buddhist knowledge, indeed, they did seem to have a part to play in the Buddha's own attainment of nibbaana.(9)

The major point that is to be drawn from the above is that the early Buddhist view of verification and knowledge is not as naive as Kalupahana's comparison of it with Logical Positivism would lead us to believe.(10)

The empiricism of Early Buddhism is like that of the Logical Positivists in that they both rejected a priori reasoning and authority as sources of knowledge, and tried to ground knowledge in experience. But this is where the similarity ends; there are major differences that highlight the Superficiality of the similarities so far cited.

Those Logical Positivists that were interested in knowledge (rather than in demarcation criteria for linguistic meaningfulness) were interested in it as a public endeavour, "science." Knowledge was seen as somehow being public property which a society built on. On the other hand, knowledge is a private thing for Early Buddhism. Even belief in the Four Noble Truths does not count as knowledge unless one has investigated them personally, verified them for oneself. For Early Buddhism, "public knowledge" would be a contradiction in terms.

Secondly, for Early Buddhism, grounding knowledge in experience was not a straightforward matter whereby this or that experience verified this or that proposition in a clear-cut manner. Rather, it was a matter of observing or seeing "with proper understanding."(11) This not only points one to the fact that people can misperceive; it is suggestive of a more significant point, that people can and do misinterpret what they observe consistently.

The abovementioned injunction to check one's personal knowledge can be seen as a caution stemming from the above fact. So, too, can the frequently stated claim in the Nikaayas that people hold views because of certain likes and dislikes be seen as another facet (in fact the cause) of people consistently misinterpreting the world.

Thirdly, not only did Early Buddhism consider the mind not to be a "blank slate" upon which the world wrote, it insisted that the "contents" of the mind were legitimate "objects" of observation. Seeing and knowing the world correctly involved eliminating the "theory ladenness" of our observations, and the only way one could eliminate such colouring of our experiences was to be aware of our likes and dislikes, the theories or views we held to. Thoughts, feelings, habitual tendencies, and so forth were all things we could and should observe and come to see and know correctly.

The "empiricism" of Early Buddhism is aptly encapsulated by Edward Conze when he claims that "like is known by like" in that there is a "hierarchy of insights dependent on spiritual maturity."(12) However, Conze does not see this as empirical at all; indeed, he views this as a demarcation between the spiritual and the empirical. I think he is right in that it is at this point that Early Buddhism diverges from the Logical Positivists, as it does take into account the above mentioned phenomena. But I think he is wrong in asserting that Early Buddhism is unempirical. It just had a more sophisticated idea of what it is to be empirical than had the Logical Positivists, the latter being Conze's yardstick of "empiricism."

There is nothing methodologically wrong with the fact that special training is needed in order for people to have certain observations and understandings. Nor is a training procedure unsound because not all people are suited to, capable of, or interested in that training. If this were the case then modern physics would be in disrepute. To someone who objected to this line of argument by saying that modern physics is involved in "public" (read "objective") observations, while Buddhists indulge in "private" (read "subjective," "unscientific") observations, Early Buddhism would reply: "All experiences are 'subjective', all knowledge is 'personal knowledge'."

The above is not meant to be a full blown defence of Early Buddhist empiricism; it is more a sketch of how one can defend such a position, and an indication of where I think the important issues lie. The main point is that there is a case to be made that Early Buddhism was empirical, in the way that modern science might be said to be empirical but not in the way in which "the Lord Buddha finds himself conscripted as a supporter of the British Philosophical tradition of empiricism'."(13)

With the above as background, I now propose to examine some other areas central to Early Buddhism.



It would appear that it is easier to say what the Early Buddhist theory of causation is not, or what it rejects, than what it actually proposes. Many of the scholars working in this area introduce the topic by detailing those ideas about causation that Early Buddhism repudiates (that is, cause through/by self other than self-caused, self and other-caused, and the thesis that there is no causation). These authors explain why such ideas were rejected (not only because they were supposedly factually false, but because they lead to the heretical views of either eternalism or annihilationism), and then state that Buddhism treads a middle path between these extremes. But what is this theory that treads a middle path?

Kalupahana believes that the Early Buddhist theory "transcends the commonsense notion"(14) where the "commonsense notion" is one that distinguishes causes and condition.(15)

While recognizing several factors that are necessary to produce an effect, it does not select one from a set of jointly sufficient conditions and present it as the cause of the effect. In speaking of causation, it recognizes a system whose parts are mutually dependent.(16)

Jayatilleke holds that the Early Buddhist view of causation "resembles the [modern day] Regularity theory except for the fact that it speaks of... empirical necessity."(17) Kalupahana disagrees with this latter point, and offers evidence in the form of a quotation from Buddhaghosa on 'necessity' which he thinks shows that 'necessity' just points to the existence of regularity: "Since there is no failure, even for a moment, to produce the events that arise when the conditions come together, there is said to be 'necessity.'"(18)

However, Kalupahana does not believe that this leaves Early Buddhism with a regularity theory or something closely akin to one.(19) In support of this he says:

"It is true that Early Buddhism depended on experience (i.e., 'contact,' phassa, ch'u or 'sensation, vedanaa, shou) to verify the nature of reality. But such experience was not considered momentary.... Therefore, the causal connection itself becomes an object of experience."(20)

Furthermore he cites the statement "On the arising of this, that arises" and the use of "dependent origination" to describe causation, as evidence of an attempt to bring the notion of "productivity" into causation.

Although the use of these terms and aphorisms might be seen as suggestive of more than regularity, it is difficult to understand what Kalupahana means by "productivity." Kalupahana himself points out in other parts of his book that Early Buddhism did not want to bring "metaphysical" concepts of "power to produce, " and so forth into causation.(21)

Jayatilleke also notes the same point: "The Buddhist theory is... empirical since it spoke only of observable causes without any metaphysical pre-suppositions of any substrata behind them."(22)

Given this, it would seem that the most plausible support Kalupahana has for holding that Early Buddhist views on causation were not ones of mere constant conjunction, akin to a regularity theory, is that Early Buddhists felt that they could perceive causal connections. However, nothing we have looked at so far would seem to imply that Early Buddhists could, or claimed they could, see causal connections. Kalupahana's point is that since Early Buddhists were able to resort to the sixfold higher knowledge, they could "perceive the relationship between two events that are separated in time and space."(23) But it is not at all obvious that one can perceive relationships, at least in the way that it is not obvious that "one is able to perceive the causal connection between two events that succeed one another without a pause or temporal gap (e.g., the connection between touching a live electric wire and getting a shock)."(24) Kalupahana claims one can see this connection, and with "higher knowledge" one can see in a similar fashion other connections.

Taking Kalupahana's example, one can perceive the touching of the wire and the shock, but whether one can perceive anything over and above this, a "causal connection," is doubtful.(25) But this claim appears even more dubious when one remembers that Early Buddhists do not, as Kalupahana himself notes, "select one from a set of jointly sufficient conditions and present it as the cause of the effect."(26)

In the above example, it would not just be the touching of the wire that caused the shock. The "cause" would be a set of jointly sufficient conditions: touching the wire with bare hands, you being "grounded," the wire connected to a generator that is functioning, and so on.

It might seem plausible to assert that one can see the connection between a cause and an effect: A------causes------B

But, it does not appear possible that one can view the relationship between mutually dependent conditions and an effect. It is this latter view of causation that the Early Buddhists held.

All the above would seem to argue for our rejecting Kalupahana's claim that Early Buddhists could perceive causal connections, and that their theory of causation was something more than a regularity theory. However, despite this, I think Kalupahana is correct, or at least his two conclusions are correct; the reasons he gives for his conclusions seem to me less than adequate in places.

The first point to be made is that Jayatilleke is right in claiming that Early Buddhism spoke of necessity, and it is prejudice that makes Kalupahana attempt twisting the idea of "necessity"' into the idea of regularity.(27) This is a crucial point to make, for if one allows that there are no empirical or "natural" necessities, one gives the "in principle argument'' of Hume (mentioned by Siderits in my note 25) a foothold.

Hume himself acknowledged that we have the "impression" that we experience causal connections, but argues that since "empirical necessities" are impossible in principle, we must be mistaken, and these "impressions" must somehow have to do with habit and expectations, and they could only arise through our projecting them (albeit unconsciously) onto the world. An Early Buddhist would have rejected this however, for he would have given priority to experience and claimed that the "in principle" argument is sophistry that needs to be rejected in favour of experience. We have already noted that Early Buddhism considered the observation of "internal" objects of experience, thoughts,feelings, perceptions, and so forth as important as the observation of "external" objects. Because of this, Early Buddhists would have had little reason to doubt that they did experience causal connections. The arising of pain when one puts one's hand into a fire is not just constant conjunction, it is an experience of the causal phenomenon "fire" in relation to what it is to be the causal phenomenon "human being." We all experience this causal connection. We shall say more about this in discussing dharmas; for the moment, I would like to say a bit more about the Humean arguments that might make what we have already said somewhat more palatable.

It might be thought that the rejection of Hume's argument on the above grounds is cavalier, if not foolhardy, even if it is consistent with the Early Buddhist epistemological position. Some might say that if this is what the Early Buddhist epistemological stance condones, so much the worse for that stance. However, the work of some modern day philosophers, particularly that of Saul Kripke, (28) should make the above seem less cavalier, particularly in relation to the idea that empirical necessities exist. In summarizing Kripke's views Stephen P. Schwartz says:

Kripke claims that most recent and contemporary philosophers have failed to distinguish the metaphysical notion of necessity from the epistemological notion of prioricity and the linguistic notion of analyticity. If "necessarily true" means true in all possible worlds and a priori means knowable independently of experience, then we are talking about two very different notions, and there is no reason to suppose that their extensions have to be the same... Of course, the claim that there can be synthetic necessary propositions is startling to most contemporary analytic philosophers, but given the persuasiveness of Kripke's arguments and examples the claim must be taken seriously.(29)

This is not the place to discuss Kripke's arguments; the main point is that Hume's thesis that there are no such things as "empirical necessities" is arguably false, and out experiencing such things need not be an illusion, a product of habit and expectations. The other strand of Hume's argument, that any single experience is insufficient in giving us experience of "causal powers, " is equally suspect. Not is it with the identification of necessity with a prioricity,(30) but some authors also have argued that:

... Hume's conclusion is already presupposed by his atomistic epistemic assumptions. Since he takes punciform and atomistic sensations as epistemically basic, it would be impossible for any single impression to be the original of any relational concept, let alone the original of 'the action of causal powers'. This epistemic assumption, however, seems more dubious to us than the fact of the ordinary experience of the action of causal powers....(31)

As we have already noted, Early Buddhists did not hold such epistemic assumptions, and would consequently not be swayed by the above Humean argument, nor be inclined to view their observation of things in its light.

"Things" Involved in Causation (Dharmas)

All dhammas (Pali) are said to be characterised by:

(a) impermanence (anicca)

(b) unsatisfactoriness (dukkha)

(c) nonsubstantiality (anatta).

Kalupahana sees (a) as the essential characteristic and (b) and (c) as corollaries. He also believes that for Early Buddhism the claim "all things are impermanent'' is not based on the view that things are momentary, which Kalupahana says would be seen as a speculative metaphysical opinion by Early Buddhists, but on an empirical claim about objects of experience, which can be verified by seeing that all things are subject to birth, decay, and destruction, arising and passing away.

The term dhamma, when applied to empirical things, is always used in the sense of 'causally conditioned dhammas' (paticcasamupanna-dhamma)."(32)

Putting these ideas about "things" and "causation" together, we see that for the Early Buddhists all things are conditioned, and all conditions are themselves conditioned. It is essential that we keep this in mind, for in my opinion it is central to Early Buddhism. "Causation" is not one thing and "things involved in causation" another; we can differentiate them for ease of discussion, but ontologically the are not separate or separable.

'Things' are not static, 'things' to which something happens or which "bump into" other things; to be a thing is to be a causal thing, to be conditioned and a condition. This is an active, dynamic understanding of causation and of causal things. Thus a fire is not one thing and heat another, something a fire "has." "Fire" is a name we give a complex causal phenomenon that has certain conditions for its arising and certain necessary effects. To perceive a fire is to perceive a causal phenomenon, to perceive causal powers. For this reason it would be impossible for Early Buddhists to hold a regularity theory of causation along the lines suggested by Hume, in that "we may define a cause to be an object followed by another, and where all objects, similar to the first, are followed by objects similar to the second."(33) An Early Buddhist would understand the notion of one object being similar to another as one causal phenomenon being similar to another. For us to perceive two objects as similar is for us to be in similar causal relationships with two assumedly similar causal phenomena.

Our earlier criticism of Kalupahana gained its persuasiveness because "things" were understood as "static things'' which were somehow mysteriously "joined" in a causal situation. However, once we switch to looking at "things" as "causal" things, as complex causal phenomena for which we have certain names, the mystery disappears.

This also explains why the Early Buddhists were reluctant to allow ideas of "power to produce" into the area of causation, for this brings us back to the static notion of things, which supposedly possess these powers. To repeat, a thing is nothing other than a 'causal thing', a causal phenomenon. As Harre and Madden point out: "The exercise of causal power or efficacy is nothing in general; it is precisely the relationship of production between specific and potent objects... and revealed in the experience of daily life...."(34)

These views are, I think, a direct result of the Early Buddhist theory of knowledge. This empirical attitude also leads to a concentration on the practical aspects of attaining nibbaana, the conditions that need to be met for its attainment, while excluding all 'speculative' views. The major drawback of this attitude is that Early Buddhism does not, and indeed cannot, give causal explanations about much that we experience.

Causal explanations involve indulging in speculative views, theorising about causal mechanisms, suggesting hypothetical, theoretical entities. Early Buddhism cannot give such explanations. They often list conditions that need to be satisfied before certain effects occur, but they rarely are in a position to discuss what is significant about such conditions such that they do give rise to effects. The distinction is between being told that x, y, z, and so on need to be done before one's car will start, and being told what it is about x, y, z, and so on that causes one's car to start. The former is a listing of conditions, the latter an attempt at a causal explanation.

This is a problem I see with most Early Buddhist "explanations." Jayatilleke notes this about one aspect of Early Buddhism when he says, with regard to the ability to remember past lives, that "the pali Nikaaya's are apparently not interested in accounting for this memory by a theory, but in merely stating that it is a faculty that can be evoked.'"(35) In many ways, however, this is not a disaster for Early Buddhism, as it was not interested in explaining everything; rather it was concerned with explaining suffering and its cessation, and in this area it did provide explanations.

Nirvaa.na, Self, and the Unanswered Questions Our discussion of causation should lead us to expect Early Buddhism to see the "self" (or at least a certain understanding of the self) as a causal phenomenon. Kalupahana succinctly summarises this view saying:

According to him (the Buddha) this (the concept 'man') was merely a 'bundle of perceptions', or a group of aggregates, not discrete and discontinuous, but connected by way of causation.(36) Somewhat reluctantly he also notes that one could take the stance that:

What the Buddha denied was that the five aggregates (khandaa) are the permanent and eternal 'self' (aatman) and that the Buddha did not actually deny a 'self' over and above, or not identical with, the aggregates.(37)

According to Kalupahana this leads us to the problem of the unanswered questions. I do not wish to discuss these in detail yet, but we can note that Kalupahana's position is that to the Buddha the 'self', whether it is identical with the body or different from the body, is a metaphysical entity. It is a metaphysical entity solely because it is unverifiable, either through sense perception or through extrasensory perception. In short, it is not given in experience (avisaya), and therefore the Buddha left these questions undeclared.(38)

Kalupahana disagrees with any suggestion that Early Buddhism implies the existence of a "transcendent" self, or a "transcendental" nirvaa.na state. It is the latter claim I wish to tackle first. Kalupahana's first step in arguing for this is to claim that Early Buddhism equates on the one hand,

(i) Nirvaa.na obtained in this life, with,

(ii) Nirvaa.na substrate left (saupaadisesa); while on the other hand he equace,

(III) Nirvaa.na after death, with,

(iv) Nirvaa.na substrate (anupaadisesa).(39)

That is, Kalupahana is taking the meaning of saupaadisesa (and its negation) , literally. Saupaadisesa meaning "having substratum of life remaining"(40) is understood by Kalupahana to refer to bodily existence, and not to something like "potential for rebirth'' or "having inclination or craving for life remaining."

The point of these distinctions is that, given them, Kalupahana can claim that there is nothing essentially transcendental about one who has attained nirvaa.na in this life. Further, the only reason we can say nothing about a dead arahant (which for Kalupahana is the same as one who has attained nirvaa.na without substrate) is that one should not speculate about unverifiable matters.

That is, no experience of such a one can be had through any form of perception, and the nature of one who has attained nirvaa.na without substrate is not unverifiable because he is transcendent, and so forth, but because such a one is dead and beyond empirical investigation.

The above argument seems to me to be questionable on a number of points. First, it is not obvious that one can make the equations (i) = (ii) and (iii) = (iv), above (that is, nirvaa.na with substrate left, meaning "nirvaa.na in this life" and nirvaa.na without substrate, meaning "nirvaa.na after death") . If in fact this cannot be substantiated, then Kalupahana can no longer maintain that the only reason one cannot say anything about one who has attained nirvaa.na without substrate is because such a one is dead, and as such, beyond empirical investigation.

In the "Discourse on the Relays of Chariots" (Rathaviniitasutta), in the Majjhima Nikaaya, it is said that one lives under the Buddha for Utter Nibbaana without attachment (hereafter U.N.W.A.), and U.N.W.A. is said to be none of:

(a) purity of moral habit,

(b) purity of mind,

(c) purity of view,

(d) purity through crossing over doubt,

(e) purity of knowledge and insight into the way and what is not the way,

(f) purity of knowledge and insight into the course,

(g) purity arising from knowledge and insight.(41)

Each of these states is said to be a purpose for the next state (presumably each is a necessary condition for progressing to the next); for example, "purity of moral habit" is a purpose for (necessary for) "purity of mind." A footnote referring to the list above explains:

Whatever is purity of mind, this is the goal (attha), this the peak, this the culmination of purity of moral habit.(42)

The main thing to note about the above is that there is no mention of death as a necessary condition for U.N.W.A. (and (a)(g) does seem to be a list of necessary conditions), and the last "state" in the list, (g), is said to be a purpose for U.N.W.A. Also we should note that "without attachment" is a translation of anupaadaa, which the Pali-English Dictionary says is a gerund of an upaadiyati, the latter meaning "to take hold of, to grasp, cling to, show attachment (to the world)." This suggests a less literal translation of anupaadisesa (without substrate), than that given by Kalupahana, referred to above. Moreover, the following quotations from Early Buddhist texts seem to deny Kalupahana's equations, (i) = (ii),

(iii) = (iv); in these texts it is implied that U.N.W.A. can be attained in this life. At the end of the "Discourse of Kii.taagiri" (Kii.taagirisutta) the Buddha says:

For a disciple who has faith in the Teacher's instructions and lives in unison with it, monks, one of two fruits is to be expected: profound knowledge here and now, or, if there is any basis (for rebirth remaining), the state of no-return.(43)

The Buddha expands on this at the end of "The Parable of the Water-Snake" (Alagadduupamasutta). Here, after rejecting the view that he is a nihilist, the Buddha lists six possibilities for his followers. It is worthwhile quoting the significant parts of the first two at length:

(1) ... those monks who are perfected ones, the cankers destroyed, who have lived the life, done what has to be done, laid down the burden, attained their own goal, the fetter of becoming utterly destroyed, and who are by perfect profound knowledge the track of these cannot be discerned.

(2) ... those monks in whom the five fetters binding the lower (shore) are got rid of all these are of spontaneous uprising, they are attainers of utter nibbaana there, not liable to return from that world.(44)

The Buddha goes on to list four more groups of attainment for his followers, briefly these are:

(3) once returners, not liable to downfall

(4) bound for awakening

(5) if striving for faith, bound for awakening

(6) bound for heaven. (It is to be noted that monks are mentioned in all but the last group.)

Keeping in mind that the above is a descending order of accomplishment (from 1-6), we can note some of the similarities and differences between the first two groups. It is stated that the second group are non-returners, consequently the first must also be non-returners. The reason for this is that since the above is a descending order, whatever a lower group has attained a higher will have attained and gone beyond. This is one similarity; the differences are:


Have perfect profound No mention of yet having

knowledge and are freed perfect profound knowledge and freedom


Said to be untraceable (tracks said to attain utter nibbaana cannot be discerned) there (after death ?)

It seems clear from the above that attaining the state of non-returner, and having the knowledge that one is a non-returner, is not the attainment of the goal of Buddhism (although it is an indication that one is near the goal, certain to attain it).

In contrast to the "lower" second group, the first has attained U.N.W.A. here, in this life. This is a direct contradiction of Kalupahana's claim.

What also is clear is that knowledge is required for U.N.W.A. (Recall (a)-(g) above, especially (g).) That knowledge is a requirement for U.N.W.A.

should not come as a surprise, as the Buddhists hold that ultimately it is ignorance that is responsible for craving, grasping, old age, death, and so forth. That in the end ignorance can only be eradicated by knowledge seems entirely consistent and to be expected. We might surmise that for some, death can aid in acquiring such knowledge and thus attaining U.N.W.A. But death is not the only way to such knowledge, and indeed it is the "perfect profound knowledge" that is important for U.N.W.A., not death.

If one can attain U.N.W.A. in this life, then Kalupahana can no longer maintain that the only reason the Buddha said little about such a one is that such a one is beyond the scope of knowledge because he is dead and the Buddha did not want to involve himself in mere speculation.

It is my contention that there is little essential difference between a living and a dead Tathaagata (one who has attained U.N.W.A.), and the reason the Buddha said little about such a being must be seen not only as because of the scope and limits of knowledge, but, more importantly, because of the "nature" of one who has attained U.N.W.A. Problems about this "nature" will have little to do with problems of tracing one who has died. The same conclusion, that there is no essential difference between a living and a dead Tathaagata, is held by RuneJohansson.(45) Kalupahana's judgement on this conclusion is that this is a very superficial comparison.... A living arahant cannot be known easily by an ordinary person, nor even gods (Indra, Brahma or Prajaapati- M 1.40), because his ways are very different from their own.... But an arahant can be known by another arahant. On the other hand, the nature of a dead arahant cannot be known even by an arahant.(46)

In this instance I think Kalupahana slides over quite a few problems much too quickly and ends up with a number of erroneous claims. The full excerpt from the Majjhima Nikaaya used by Kalupahana in the above is:

Monks, when a monk's mind is free thus (by getting rid of the conceit 'I am') the devas those with Indra, those with Brahmaa, those with Pajaapati, do not succeed in their search if they think: 'This is the discriminative consciousness attached to the Tathaagata. What is the reason for this?(47)

The reason given is certainly not the one presented by Kalupahana above ("his ways are very different from their own"); it is (Horner:)(48) "I say here and now that a Tathaagata is untraceable." (Johansson:)(49) "I say that a Tathaagata cannot be known even in this life."

Given this together with our previous discussion, it seems clear that a Tathaagata cannot be known if looked for by another. Further, a

Tathaagata cannot be known by another, because of his nature and not for any straight-forward empirical reason, such as his being dead.

What are we to say about Kalupahana's other point, that an arahant can be known by another arahant in this life? It would appear that if this were true, one could not reasonably keep the position I have been advocating.

I would like to tackle this by noting that there are at least three things we could mean by the above claim:

(i) That an arahant can be known to be an arahant, by another arahant. That is, an arahant can know that another is an arahant, in this life.

(ii) That an arahant can know the nature of another who is known to be an arahant (i,e, (i)), in this life.

(iii) That an arahant can have direct perception of the arahant nature of another arahant, in this life.

I wish to argue that (i) and (ii) are not problems for the sort of position I wish to advance. These are not problematic, for, remembering that Early Buddhism allowed inference based on experience as a form of knowledge, (i) and (ii) could be ture, and it also could be true that a living arahant is unknowable by another.

I could know (i) through inference, and partly because an arahant is unknowable by another. For example:

Premise(1) Normal people have certain identifiable characteristics, egos the conceit "I am" habits of mind, and so forth, Premise(2) The only people who do not have such are arahants known from my own experience if

I am an arahant. Premise (3) This person has no such haracteristics-known by psychic powers.

Therefore this person is known by me to be an arahant.

Similarly, if I am an arahant, and assumedly know my "own" nature, I can therefore, knowing (i), inferentially know (ii), given that all arahants have a similar, or the same, nature. Which of (i), (ii), or (iii) is claimed by Early Buddhism? It is difficult to know where Kalupahana's claim comes from, for he cites no texts at this point in his work, but the section of Johansson's book he is commenting on does mention a text. It says:The venerable Maha-Moggollana saw with his mind (ceto) that their minds (citta) were freed without basis (for rebirth).(50)

This is certainly compatible with holding that an arahant is unknowable by another. I can know that someone is freed (for example, by seeing an empty prison cell) and "know" where he is (that is, outside the cell) without implying I can trace or directly perceive where, or what, he is.

Since a living arahant has been claimed to be unknowable by another, it should follow that a dead arahant is also. However, Johansson cites texts

which he interprets as indicating that: at least the Buddha himself claimed the ability to identify and report about dead arahants.... the Buddha himself was able to trace an arahant after death.(51)

I think Johansson is drawing the wrong conclusions from these texts, mainly because he does not make the sort of distinctions just made (that

is, (i), (ii), (iii) above). for there is a world of difference between being able to report about, on the one hand, and identify or trace, on the other.

The texts he cites relate how the Buddha reported that certain monks who died in a fire had attained parinibbaana.

If an arahant is unknowable by another, one could still know inferentially that someone had attained nibbaana. That is, the Buddha could report that someone had attained nibbaana, without it being the case that he could directly perceive ("know" directly) the nature of his being. The Buddha could know, through inference, that someone had attained nibbaana after death because:

(a) only beings who have attained nibbana are not reborn (known through experience and inference), (b) a certain being cannot be seen to arise (known through psychic power clairvoyance the ability to see the demise and arising of beings), (c) one can 'know', and report, that this certain being had attained nibbaana.

We can pause here and note some of the conclusions we have come to:

(1) Although Early Buddhism rejected a concept of self composed of the aggregates, there seem to be suggestions that there is in some sense a self, but not an "I."

(2) For Early Buddhism, nibbaana is something more than just escape from the cycle of rebirth, more than becoming a non-returner. Further, attaining nibbaana is intimately entwined with knowledge.

(3) U.N.W.A. is attainable in this life, and it would appear to be an untraceable "state," that is, (it is) not directly perceivable by another.

What we cannot conclude from the above discussion is that Early Buddhism has implications of the existence of a "transcendental" nibbaana state, if we understand by "transcendental" the claim that it is a state that is not knowable. We have seen that this state is not directly perceivable by another, but it is "knowable'' if we follow Early Buddhist epistemological criteria. That is, this state is knowable in the sense that all things are, in that we can have personal experience of it, we can experience nibbaana. It is only when one unconsciously imports foreign epistemological criteria that the debate about a transcendental state of nibbaana arises. Nibbaana a transcendental state, if by this we mean a state that surpasses others; it is not if we mean a state that is beyond knowledge, and hold consistently to the Early Buddhist criteria for knowledge. It is neither sufficient(52) nor necessary that something be able to be talked about, or described, before that something comes into the province of human knowledge, as far as the Early Buddhists are concerned. It is the Buddha's refusal to talk about, or describe, the nibbaana state and the Tathaagata that sparks much of the debate about "transcendentalism" in Early Buddhism.

With this discussion in mind we can now approach those unanswered questions of the Buddha that have to do with the Tathaagata. When asked to reply to the statements:

a) The saint exists after death

b) The saint does not exist after death

c) The saint does and does not exist after death

d) The saint neither exists nor does not exist after death the Buddha responded by saying "I do not say this" to each in turn.

I think we have said enough to enable us to dismiss Kalupahana's explanation of the Buddha's silence on these questions. Kalupahana's explanation was that "the silence of the Buddha was due to his awareness of the limitations of empiricism, rather than concepts."(53) Is then the silence of the Buddha as a response to the questions about the Tathaagata due to the limitations of concepts as Murti and Jayatilleke amongst others believe? I think it is.

If we look at the context of the text in which these questions are put to the Buddha by Vacchalotta, in the "Discourse to the Vacchagotta on Fire," we find the Buddha saying "does not apply" to each of the alternatives (a-d) and then saying: "this dhamma is deep, difficult to see, difficult to understand, peaceful, excellent, beyond dialectic."(54) By way of explanation the Buddha asks Vaccha what Vaccha would say on being asked which direction (North, South, East, or West) the fire in front of him went on being quenched. The answer to be given to this question containing four alternatives (North, South, East, West) is the same as the one that is appropriate in response to the questions about the Tathaagata, which also entertain four alternatives (a-d), that answer being, the question does not apply.

The point of the analogy would seem to be that in both cases, what is assumed by the very framing of the question ('existence' applying to the Tathaagata, spatial direction' applying to a spent fire) is unacceptable. So much so that to deny the question is not enough; one must deny the assumption in the question, and one does this by rejecting the question as not legitimate. These are sometimes called category mistakes, but all questions of the sort we have been talking about need not involve category mistakes; for example, ''Have you stopped beating your wife?" is a question that might fit into the category of those which involve assumptions one wants to reject, but does not involve a category mistake.

All this clearly suggests that the silence of the Buddha is likely to be due to the lack of adequate concepts, due to the uncharacterizable, non-describable (but not non-knowable) nature of the Tathaagata. However, in a recent article Mark Siderits asserts that while we can see the unanswered questions as pointing to the Buddha being aware that a category mistake is being made, we should also note that: "the Buddha is not saying that the state of the arahat after death is indescribable or ineffable. This possibility is represented by the fourth of the four alternatives, which the Buddha rejected."(55)

It seems to me that Siderits is plainly mistaken, The fourth alternative does not cover the possibility that the arahat after death is indescribable: it covers the possibility that after death the arahat is describable neither as existing nor as not existing. That is, accepting the fourth alternative does not entail that one accept that the arahat is indescribable after death, it entails that one accept that the arahat is not describable as either existing or not existing after death. One could still accept that the arahat was not describable in these terms and at the same time hold that the arahat was describable some other way. So rejecting the fourth alternative is not tantamount to rejecting indescribability per se. The point of merit in Siderits' paper is the suggestion that the case for the indescribability of

the arahat after death cannot rest just on the Buddha's silence in the face of these questions. And my point is that it does not. Gathering together the salient features of the above discussion, the substance of my argument is:

(1) Early Buddhism placed fundamental importance on knowledge, understood as personal experience, and so forth. One could accept the Four Noble Truths as truths, but one did not know them until one had personal experience of them, and it was knowledge of them that was essentially important.

(2) A certain sort of theory of meaning seems to follow from this, and it would appear to be something like that suggested by Jayatilleke, namely, that if one had no personal knowledge of the things referred to in a statement, then that statement was meaningless to one.

(3) If one accepts, both that nibbaana is a "state" that surpasses others, and that one can only have knowledge of it by experiencing it for oneself, that it is unlike all other "states," then:

(4) All statements about that "state" will be meaningless to one, if one has not experienced that "state."

Now someone might point out in reply to this that what the above argues for is not the indescribability of the nibbaana state but its indescribability to all those who have not experienced it, and there may be true but to us meaningless descriptions of the state, so that it is therefore describable. It is at this point that the unanswered questions have importance, for they supplement point 3 above, and indicate how unlike all other states the nibbaana state is. They point to the fact that nibbaana is unlike all other states to the extent that all words and descriptions that derive from and refer to experiences that are other than nibbaana are inappropriate for describing nibbaana. And, if something as basic as existential language, exists/does not exist, constitutes a category mistake, I am mystified as to what further categories our language and experience leave us as alternative options of description.(56)

In summary then, my conclusions are that nibbaana is:

(a) Knowable, but not directly perceivable in another,

(b) Not meaningfully describable to one who has not experienced it,

(c) Not describable using language which gains its reference from experiences that are 'other-than' nibbaana, that is, it is not describable.



(1) Kalupahana takes "Early Buddhism" to be that which is presented in the Paali Nikaayas and the Chinese AAgamas. I shall follow this usage.

(2) David Kalupahana, Causality: The Central Philosophy of Buddhism (Honolulu: The University Press of Hawaii, 1975), p. 185.

(3) K. N. Jayatilleke, Early Buddhist Theory of Knowledge (London: George Alien & Unwin Ltd., 1963), p. 416.

(4) Ibid.

(5) See Kalupahana, p. 104, and Jayatilieke, p. 438, where these are noted and discussed.

(6) Jayatilleke, p. 439.

(7) Having any such ability may be a form of knowledge itself, but, following Jayatilleke's distinction, this too is a matter of 'knowing how' (e.g., to read minds) rather than 'knowing that'.

(8). Jayatilleke, p. 439. Jayatilleke goes further than this and later claims that "causal empirical explanations were everywhere substituted (e.g., theories of perception, knowledge, consciousness, etc.), for prevalent metaphysical theories." (Ibid., p. 452). This further claim is incorrect. in my opinion, as I argue below.

(9) Among other places this is mentioned in 'The discourse on the Ariyan Quest' in the Majjhima Nikaaya. I do not wish to go into the status of these higher levels of meditation here, but one need keep in mind that they are mentioned in Early Buddhist texts, and they do have some part to play in Early Buddhist knowledge, a part that has been virtually ignored by modern scholars.

(10) I do not wish to imply that Kalupahana is alone; many scholars have held that Early Buddhism maintained a naive form of empiricism.

(11) Kalupahana, p. 11.

(12) Edward Conze, Buddhist Thought in India (London: George Allen & Unwin, 1962), pp. 26 and 17, respectively.

(13) Ibid.,p.26.

(14) Kalupahana, p. 59.

(15) Kalupahana criticises Murti for not distinguishing between the Early Buddhist and the later theories of the Abhidharma. He, rightly, I think, points out that the former did not have a theory of moment, or atoms, nor did it distinguish between cause and conditions. See The Central Philosophy of Buddhism (London: Allen & Unwin, 1955).

(16) Kalupahana, p. 59.

(17) Jayatilleke, p. 453.

(18) Kalupahana, p. 93.

(19) Jayatilleke. on the other hand, is prepared to accept that the Early Buddhist theory is a type of regularity theory, and this is open to such a theory's traditional criticisms that its characterization cannot distinguish between accidental constant conjunctions and causal situations. This seems to be the crux of E. J. Thomas' criticism of Early Buddhist ideas on causation. See The History of Buddhist Thought (New York, 1933), p. 62.

(20) Kalupahana, p. 96.

(21) Ibid., p.73.

(22) Jayatilleke, p. 453.

(23) Kalupahana, p. 104.

(24) Ibid.

(25) Mark Siderits, in reviewing Kalupahana's book points out that "no empiricist may consistently speak of... an experience of causal connection... no amount of additional kinds of sensa can render perceptible what is in principle imperceptible." See The Journal of Indian Philosophy 8:194.

(26) Kalupahana, p. 59.

(27) Siderits claims that "Kalupahana represents the Buddha as holding that there is a necessary connection between cause and effect" (p. 96) .

However, Kalupahana does not claim this. What he does say is that Early Buddhism viewed causation as more than constant conjunction, but what this specifically means is never spelt out. Kalupahana draws back from claiming that Early Buddhists held to empirical necessity as we noted above, and I think Kalupahana is wrong so to draw back. See Journal of Indian Philosophy 8:193.

(28) See "Naming and Necessity" in The Semantics of Natural Language, edited by Donald Davidson and Gilbert Harman (Dordrecht: D. Reidel, 1972).

(29) Stephen P. Schwartz, "Intorduction" to Naming, Necessity and Natural Kinds, edited by Stephen P. Schwartz (Ithaca: Cornell University Press, 1977), pp. 25-26.

(30) Part of the force of Hume's epistemic argument comes from the idea that since it is only with repeated exposure that we discover certain

things about objects (e.g., fires), things that are not discoverable at first glance; such things can only be 'empirical', meaning 'not necessary'. One must also resist the temptation to confuse problems about empirical necessities with problems about induction. To say that fires must necessarily burn unclad human beings is not to claim that fires must always (or necessarily) exist, nor is it to claim that things that look like fires must burn things that look like unclad human beings.

(31) R. Harre and E. H. Madden, Causal Powers (Oxford: Basil Blackwood, 1975), p. 55.

(32) Kalupahana, p. 84.

(33) An Enquiry Concerning Human Understanding, Section VII.

(34) Harre and Madden, p. 57.

(35) Jayatilleke, p. 440.

(36) David Kalupahana, Buddhist Philosophy: A Historical Analysis (Honolulu: University Press of Hawaii, 1976), p. 39.

(37) Ibid.

(38) Ibid., p.41.

(39) Ibid., p. 69.

(40) T. W. Rhys Davids and W. Stede, eds., The Pali Text Society's Pali-English Dictionary (London, 1921-1925), p. 655.

(41) Majjhima Nikaaya, in The Middle Length Sayings, vols. I-III, trans. I. B. Horner (London: Luzac & Co., Ltd., 1967), I, p. 193.

(42) Ibid.

(43) Ibid., II, p. 156.

(44) Ibid., I. pp. 181-182.

(45) Rune Johansson, The Psychology of Nirvana (London: George Allen & Unwin Ltd., 1969), p. 61.

(46) Kalupahana, Buddhist Philosophy, p. 83. Emphasis mine.

(47) Majjhima Nikaaya I, p. 179.

(48) Ibid.

(49) Johansson, p. 61

(50) Ibid., p. 62

(51) Ibid., p. 63

(52) Jayatilleke discusses the Early Buddhist theory of meaning and claims that for them a statement is meaningless if no experiential content is attached by the speaker or hearer of the words used in the statement. See Early Buddhist Theory of Knowledge, p. 321.

(53) Kalupahana, Buddhist Philosophy, p. 180.

(54) Majjhima Nikaaya, II, p. 165.

(55) Mark Siderits, "A Note on the Early Huddhist Theory of Truth," Philosophy East and West 29, no. 4 (October 1979): 491-99.

(56) I am thus puzzled by Siderits' claim that "Once we see... that such predicates as 'reborn' simply do not apply to the arahat and that the deceased arahat is subsumed under a different category, the seeming oddity of the position (the Buddha rejecting the four alternatives) vanishes."

(Ibid., p. 494). I would like to know what this "different category" is. I would also like to know why, if the Buddha did see these questionsjust in terms of rebirth (i.e., questions about whether the Tathaagata exists after death were to be understood as asking if the Tathaagata was reborn), he was perfectly happy on some occasions to say plainly that the Tathaagata was not reborn, while on others he was extremely coy, and why these latter occasions coincided with those times he was asked, not about rebirth, but about existing after death.

Transcribed for Buddhism Today by Bhikkhuni Lien Hoa


Updated: 1-2-2001

Return to "Buddhist Philosophy"

Top of Page