BuddhaSasana Home Page English Section

Cultivation of Moral Concern in Theravāda Buddhism: 
Toward a Theory of the Relation
between Tranquility and Insight

Ethan Mills
Augsburg College

 Font VU-Times (Viet-Pali Unicode)


There are two groups of scholars writing on the two main types of Buddhist meditation: one group that considers insight (vipassanā) to be essential and tranquility (samatha) to be inessential in the pursuit of nirvana, and a second group that views both samatha and vipassanā to be essential. I approach an answer to the question of which group is correct in two steps: (1), an outline of the disagreement between Paul Griffiths (of the first group) and Damien Keown (of the second group); and (2), an augmentation of Keown's assertion that samatha can cultivate moral concern. I am not definitively solving the problem of the relationship between samatha and vipassanā, but rather I show that by making Keown's theory of the cultivation of moral concern more plausible we have more reasons to accept his larger theory of the importance of both samatha and vipassanā.

1. Introduction: Meditations in Tension?

There are two main branches of Buddhist meditation techniques: insight meditation (vipassanā-bhāvanā) and tranquility meditation (samatha-bhāvanā). (1) Insight meditation is aimed at cultivating paññā (most often translated as "wisdom"); tranquility meditation is aimed a cultivating samatha ("calmness, tranquility"). Tradition generally considers the first to have been a new form of meditation invented by the historical Buddha and the second to have been highly developed by Indian practitioners by the time of the Buddha's life. The most common story is that the Buddha learned all that his meditation teachers had to offer and, still unsatisfied, developed his own type of meditation: vipassanā-bhāvanā. After he developed this insight meditation, he achieved nirvana and transcended suffering (dukkha).

I find it useful to categorize scholars who have written on the relationship between vipassanā and samatha into two groups: one group that considers vipassanā to be essential and samatha to be inessential in the pursuit of nirvana, and a second group that views both samatha and vipassanā to be essential for Buddhist soteriology. For the sake of perspicuity, I will refer to the first group as "samatha-inessentialists" and the second group as "samatha-essentialists." My goal in this essay is to approach an answer to the question of which group is correct.

I will make this approach in two steps: (1), I will outline the disagreement between Paul Griffiths (as a representative of the samatha-inessentialists) and Damien Keown (as a representative of the samatha-essentialists) on the issue; and (2), in order to come closer to deciding which of the two has the better answer, I shall augment a small part of Keown's theory, namely his assertion that samatha can cultivate moral concern, with some ideas of my own. I will not definitively solve the general problem of the relationship between samatha and vipassanā, but rather I hope to show that by making Keown's theory of the cultivation of moral concern more plausible we have more reasons to accept his larger theory of the importance of both samatha and vipassanā. As Griffiths and Keown concentrate on Theravāda sources in this context, I will limit the scope of my essay to such sources, predominantly the Nikāyas and Buddhaghosa's Visuddhimagga (Path of Purification).(2) I am making no claims about Mahāyāna or other Buddhist meditation traditions; however, I suspect that there will be much overlapping in the form and content of the problem within other Buddhist traditions.

Most scholars have considered insight meditation to be the more important of the two (for instance, Bodhi, Gunaratna, King, Rahula, and Solé-Leris). Walpola Rahula has written that the states created by samatha meditation are "...mind-created, mind-produced, conditioned.... They have nothing to do with Reality, Truth, Nirvana" (Rahula, p. 68). These sentiments are shared by Bhikkhu Bodhi: "The role of serenity is subordinated to that of insight because the latter is the crucial instrument needed to uproot the ignorance at the bottom of the samsaric bondage" (Bodhi, p. 38). To reach nirvana, one must have the insight into the true nature of reality created by vipassanā-bhāvanā. Thus, samatha is considered to be inessential for nirvana.

If samatha is inessential, why is it included in Buddhist meditation traditions? The answer, for Rahula and most scholars of this group, is that samatha techniques can sometimes help develop qualities useful in vipassanā meditation. Nonetheless, samatha techniques are not as inherently valuable as vipassanā techniques. Of course I am not claiming that these scholars have the same theory. My claim is simply that they share the basic similarity of considering vipassanā the superior form of meditation and samatha to be ultimately unnecessary—however helpful it might be—in the pursuit of nirvana. The work of Paul Griffiths, which falls into this category of vipassanā enthusiasts, will be discussed in section 1.1.

Another smaller group of scholars has sought to portray samatha and vipassanā in more equal terms; Edward Conze(3) and Robert M. Gimello are two examples. Conze does not claim that samatha can lead one to nirvana without vipassanā, but neither does he claim that vipassanā can lead one to nirvana without samatha. Samatha creates a one-pointedness of mind and "A mind of single intent is capable of doing more effectively whatever it does, be it good or bad" (Conze, p. 19). He still considers the wisdom gained by vipassanā to be the highest good because even when the greatest concentration is developed, insight is needed to reach nirvana. However, we can never gain this wisdom without samatha either. He sums up this idea as follows:

Trance, as it is developed, approaches a condition of rapt attention to an objectless inwardness (anarambana); the more wisdom develops, the clearer the intuition of emptiness (śunyatā). These are the two terminal points at which the world is on the verge of extinction. The combination of the two leads to final emancipation (Conze, p. 17, emphasis added).

Gimello very much agrees: "While it is true that discernment is not to be attained without some degree of calming as a precondition, it is no less true that calming itself, without discernment, is of no soteric avail whatsoever" (Gimello, p. 185).

Both vipassanā and samatha are necessary parts of the path toward Buddhism's soteriological goal. Although he explicitly affirms that both are necessary in a quest for nirvana, Conze admits that there "is even some tension between the two modes of approach" (Conze, p. 17). Griffiths develops this tension and Damien Keown offers an interesting solution to it, as I will discuss in section 1.2.

1.1. Griffiths: problematic yoking

In context of the treatment of the state of cessation (nirodhasamāpatti) in his book On Being Mindless, Griffiths specifically addresses the relationship between samatha and vipassanā. He characterizes samatha as enstatic and vipassanā as analytic:

Such analytical meditations are designed, then, to remove standard cognitive and perceptual habit-patterns and to replace them with new ones. Furthermore, these techniques are designed to teach the practitioner something new about the way things are, to inculcate in his consciousness a whole series of knowledges that such-and-such is the case. In contrast, the enstatic meditations are designed to reduce the contents of consciousness, to focus awareness upon a single point and ultimately to bring all mental activity to a halt (Griffiths, p. 13).

This characterization of the different methods leads Griffiths to an understanding of different soteriological goals for each method. As he says, "Those who follow and advocate the analytic techniques tend to perceive the basic human problem as one of ignorance, and inaccurate understanding of the way things are" (Griffiths, p. 14). This makes sense. If the disease is ignorance, the cure must be knowledge. Compare this with what he says about those who engage in samatha: "In drastic contrast, the practitioners of the enstatic techniques aimed at tranquility tend to perceive the basic human error as one of attitude rather than cognition; the key Buddhist term here is 'thirst' (taṇhā), a term that denotes all types of passionate desire and attachment" (Griffiths, p. 14). This also makes sense—if a certain attitude causes suffering, then we must change that attitude to one that does not cause suffering.

It is not difficult to see the tension here. It would seem that there are two completely separate goals with accompanying differing methods all jostling about within the same tradition. This has created a great many philosophical problems according to Griffiths: "Throughout Buddhist history, intellectuals have attempted to reconcile thought-systems which are on the face of it, irreconcilable" (Griffiths, p. 16). Specifically, Griffiths attributes many of the problems surrounding the attainment of cessation to the tension between samatha and vipassanā. The attainment of cessation is the stage of enstatic meditation following the attainment of the four jhānas of form and the four formless jhānas.(4) However, according to Griffiths, Buddhist theorists have attempted to describe the attainment of cessation using the soteriological framework of analytic meditation, some even going so far as to identify the attainment of cessation with the eradication of ignorance and even as nirvana itself.

If each kind of meditation has different soteriological goals, trying to piece them together is a confusing business at best. Griffiths explains this uneasy alliance with a historical hypothesis: "it quickly became orthodoxy for Indian Buddhist intellectuals that salvation must involve some degree of intellectual appropriation of doctrine, and any canonical material which appeared to present a self-consistent and coherent set of soteriological practices which involved no such intellectual activity therefore needed to be amended to accord with such orthodoxy" (Griffiths, p. 23). The examples given are of Buddhaghosa and Dhammaphala's insistence that insight must be present to reach the attainment of cessation. Buddhaghosa talks about "yoking together" insight and tranquility in order to achieve successively higher jhānas up to the attainment of cessation (Vism. 23.43). Griffiths finds this yoking process insufficiently explained and confusing given his ideas about the relations between method and soteriology.

This is his main point: it is difficult at best and impossible at worst to try to yoke together two methods into the service of one goal when the two methods are themselves designed to reach radically different goals.

1.2. Keown: Let the yoking commence

Keown's treatment of the relation between samatha and vipassanā comes in the context of a chapter on ethics and psychology in his book, The Nature of Buddhist Ethics. It is important to note that the overarching thesis of the chapter is that sīla (morality) and paññā (insight, knowledge) are two complimentary and equally necessary strands of Buddhist soteriology. It is also helpful to note that the other subtheses of the chapter can be characterized as follows:

1. Citta (mind, psyche) is an aggregate of both rational and emotional elements (dharmas).

2. The moral and the rational are fundamentally interconnected as shown in the linking of the intellectual vice of moha (delusion) and the moral vices of lobha (greed) and dosa (hatred) in a "triangle of taṇhā (clinging)."

3. The Buddha's compassion was based on and his enlightenment was preceded and followed by noncognitive, nonrational moral sentiment; the Buddha provides the paradigm case of ethical motivation.

His thesis about samatha and vipassanā can be stated briefly as follows:

Samatha-bhāvanā (calming meditation) and vipassanā-bhāvanā (insight meditation) are equally necessary, interdependent methods of attaining enlightenment by purging one's self of both moral and intellectual vices.

Thus, Keown's theory about the relationship between samatha and vipassanā is meant to supplement his larger theory on the equal importance of morality (sīla) and knowledge (paññā), which in turn is deployed to support the main argument of his book that Buddhist ethics can be best characterized as a type of virtue ethics. This context is important because it shows us that Keown means for his theory about samatha and vipassanā to solve the tension of Griffiths and at the same time advance his theory that sīla (morality) is important throughout Buddhist soteriological practices.(5)

Keown agrees with Griffiths that samatha and vipassanā aim at different outcomes. However, he disagrees about the nature of these outcomes:

Since progress in the religious life is made on two fronts, there exist two kinds of meditation techniques. I wish to suggest that "calming meditation" (samatha-bhāvanā) cultivates moral virtue and "insight meditation" (vipassanā-bhāvanā) develops knowledge or insight (Keown, p. 77).

So far this is not substantially different than Griffiths's claim that samatha cultivates an attitude and vipassanā cultivates knowledge or wisdom. This is more or less unambiguous in the tradition.

Keown's shift in interpretation of the facts of the different goals of samatha and vipassanā is subtle yet extremely important:

For Griffiths these facts are problematic but in terms of the thesis set out here they are not. Indeed, they are exactly as we should expect. Griffiths's difficulty arises from the suppressed premise of his argument that the unique soteriological objective of Buddhism is knowledge (paññā). Any soteriological technique which does not issue in paññā is therefore redundant and its experience puzzling. If nibbana is defined exclusively in terms of paññā then vipassanā will quite naturally appear to be essential while samatha remains a curious anomaly (Keown, p. 77).

For Keown, reaching the soteriological goal of Buddhism requires both knowledge and moral virtue, so it makes sense there exist meditation techniques to address each of these needs. Griffiths and those like him who view paññā as the unique goal of Buddhist meditation techniques will of course be confused by the presence of a technique that cultivates anything else.(6)

Keown explains that a mistranslation of paññā has encouraged this misinterpretation. He says, "Paññā is essentially the knowledge of facts, but wisdom means something more than just knowledge" (Keown, p. 80). Paññā is knowledge in the sense of knowing that 7 + 5 = 12, all dharmas are momentary, etc. It is the knowledge of facts about the universe. This type of knowledge is vitally important for Buddhism; the lack of it is ignorance, which is one cause of suffering. However, it is not the only cause, and neither is paññā the only cure:

The goal of the Eightfold Path is indeed wisdom, but wisdom is much more than paññā. This is why two meditative techniques are required for the eradication of the roots of evil and the attainment of the ethical and intellectual perfection which is nibbana (Keown, p. 80).

As Keown says elsewhere, "there exist two techniques of meditation precisely because the obstacles to enlightenment are themselves twofold, both moral and intellectual" (Keown, p. 79). It is the twofold nature of the causes of the disease and the complimentary twofold cure that Griffiths does not see in his interpretation.

How does this relate to Keown's first subthesis in this chapter that citta (mind, psyche) includes both rational (saññā) and emotional (vedanā) elements? The rational and emotional aspects of the psyche "may be logically distinguished but do not correspond to any real division in the structure of the human subject" (Keown, p. 67). Although the rational and emotional elements are both part of the psyche, they are different processes and each has its own virtue:

The virtue of the cognitive aspect (saññā) is to understand and discriminate correctly; its vice is delusion and error. The virtue of the nonrational part of the psyche is to sense, feel, and respond affectively in an appropriate manner; its vice is to swing to the extremes of craving (rāga) and aversion (dosa) (Keown, p. 67).

This makes it clear that if the psyche consists of two elements with their own virtues, and if the goal of meditation is to cultivate virtues of the psyche, then there must be techniques capable of cultivating these different kinds of virtues. Whereas Griffiths and other samatha-inessentialist scholars seem to feel that the development of paññā automatically leads to the development of morality, Keown feels that morality requires its own cultivation separate from the cultivation of paññā.

Keown stresses the moral aspects of samatha more than the strictly enstatic aspects and in fact conscripts the enstatic into moral cultivation. He says that the "technique of samatha meditation exists to enrich and deepen the capacity for human sympathy which exists in all to some degree and which reached its perfection in the personality of the Buddha" (Keown, p. 77).

In his third subtheses of the chapter, Keown argues that the Buddha is portrayed as displaying a noncognitive, nonrational moral sentiment that becomes the example for all Buddhists. Keown explains it succinctly: "The Buddha's moral concern is found in his sympathy (anukampā)(7) for all beings" (Keown, p. 73). Since this moral sentiment is noncognitive, vipassanā meditation (being the realm of the cognitive) is ill suited to the task of its development. We must turn to a kind of meditation that deals with the noncognitive: samatha-bhāvanā.

Reaching attainment of the jhānas is "a specialized technique for gaining access to the non-rational, emotional dimension of the psyche. It is a means of penetrating the deeper layers of consciousness and restructuring them in accordance with virtue rather than vice" (Keown, p. 78).

On this reading, samatha techniques are the means by which the psyche is restructured to have the proper affective state, namely the kind of noncognitive moral concern epitomized by the Buddha. But how is this done? Keown cites De Silva as saying that this restructuring can take place by entering the third jhāna. De Silva himself is not extremely helpful here either: "Knowledge of this stream of consciousness with a conscious and unconscious component is only within the reach of those who develop the practice of meditation ... when a person enters into the third stage of meditation such knowledge is accessible" (De Silva, p. 76). Keown says that a practitioner in the first jhāna can suppress and dissolve the negative tendencies of rāga and dosa—a process that not only frees the practitioner from the negative qualities but also cultivates positive ones.

Another hint about this process is given here:

The passions will not be extirpated in the course of a single samatha session any more than a single session of vipassanā will boost paññā to the point of perfect illumination; both techniques are slow and gradual but each is the most appropriate in its own sphere (Keown, p. 79).

Whatever it is that meditators must do to cultivate moral concern, it will take some time—perhaps even many lifetimes. Keown is not offering his picture as some kind of "get-enlightened-quick" scheme. It is still the same gradual path endorsed by the majority of Theravādins.

The cultivation of moral concern is important because "the fundamental inspiration for the Buddhist moral life is concern for others ... it is a non-rational sentiment which precedes the formulation of moral objectives" (Keown, p. 74). Without this concern there can be no morality, and without morality there can be no nirvana. The link between moral perfection and nirvana is vital to Keown's thesis because it is the fact that moral perfection can only be realized through samatha meditation that gives us a reason to assign an essential importance to samatha within Buddhist soteriology.

As Keown's theory stands, its greatest merit is a tremendous explanatory power about how it is that there are two seemingly independent and contradictory meditation techniques within one tradition. It provides us with a resolution to Griffiths's tension and gives us a reason to think that the Buddhist philosophers and practitioners throughout the ages had good reason to incorporate both samatha and vipassanā techniques into their work.

Given these merits we still must ask whether Keown's brief suggestions about how it is that samatha cultivates moral concern give us enough reasons to endorse his thesis and discount that of Griffiths. It is my feeling that they do not. We are still left with unanswered questions. For instance, if this picture is accurate, why is there so much tension and confusion among scholars of Buddhism? Keown seems to have an internalist theory of motivation, meaning that reasons for moral action are dependent—at least in part—on the moral agent's affective states. However, others—particularly samatha-inessentialist scholars of the first group above—might maintain that Buddhism generally espouses an externalist theory of motivation, by which I mean that reasons for moral action lie entirely outside the affective states of moral agents.(8) If Buddhist ethical theorists do in fact generally support an externalist theory, how can Keown be right? Is there enough textual evidence for Keown's claim to warrant making it a pan-Buddhist (or at least pan-Theravāda) theory? How exactly do samatha techniques cultivate moral concern?

It seems to me that the last question is the most preliminary in the sense that we need it to fully understand precisely what Keown's theory is. If Keown's thesis about the cultivation of moral concern via samatha techniques can be made more plausible, then we have more reason to agree with his picture of the relationship between vipassanā and samatha in general. At the very least, we shall be in a better position to evaluate the theory by asking the other questions listed above. In section 2, I offer two possible answers to the question: how could samatha techniques cultivate moral concern? If one of these answers can be shown to be exegetically responsible and philosophically compelling, then Keown's theory will be more complete and can be more easily evaluated as a response to the alleged tension of vipassanā and samatha techniques within the Theravāda tradition.

2. How Could Samatha Cultivate Moral Concern? Two Answers

Both of the arguments discussed below are attempts to answer the question of how samatha techniques cultivate the type of moral concern Keown has described above. I will evaluate each on two points: exegetical responsibility and philosophical merit. By saying an argument exhibits "exegetical responsibility," I mean that it represents the Buddhist texts in a more or less accurate way. This is an incredibly complex notion because the Buddhist texts themselves have been taken to contain a great number of meanings. Even the most respectful scholar can become confused in a maze of what often appear to be contradictory ideas between separate texts and even within one text. Rather than enter this embattled hermeneutical arena, I will simply (and a bit arbitrarily) say that exegetical responsibility means the following things:

1. The argument is grounded somehow in one or more Buddhist texts either verbatim or via analysis,.

2. The argument does not patently distort or go against the core ideas of Buddhist philosophy such as the Four Noble Truths, Dependent Origination, etc.

In this specific context, by saying an argument has "philosophical merit," I mean, (1), that it does some work toward solving a philosophical problem; (2), that it is consistent with Keown's other claims; and (3), that there are compelling philosophical reasons for us to believe it is true. For example, Keown's thesis as it stands has a great deal of philosophical merit in the first sense in that it could solve the puzzle of how samatha and vipassanā are related soteriologically. The second sense follows from my overall goal in the essay. If I am presenting this argument as an augmentation of Keown's general theory in the hope of making one complete and unified theory, it ought to cohere with what he says elsewhere. Whether Keown's general argument about the place of samatha has philosophical merit in the third sense is the question I am attempting to answer in this essay by making it more complete. Whether Keown's general theory has more philosophical merit than Griffiths's thesis is a question I cannot answer here, but I hope the work here will make such a task more profitable in the future.

In the course of the following sections, I have tried as much as possible to explicitly keep separate the issues of exegesis and philosophy. However, these two areas support and interact with each other in much the same way that Keown argues for the interaction of samatha and vipassanā. I see no shame in this; in fact, the interaction of exegesis and philosophy is something I take to be the proper method of Buddhist philosophy.

2.1 The human nature argument

What I call the "human nature argument" can be basically represented as such:

1. Human beings in their most natural state exhibit noncognitive moral concern.

2. Negative mental states such as lobha (greed) and dosa (hatred) cover up this otherwise natural tendency.

3. If humans increasingly master samatha meditation, then these negative mental states are increasingly removed.

4. If these negative mental states are increasingly removed, then humans will increasingly exist in their natural state, i.e., they will exhibit noncognitive moral concern.

C: If humans increasingly master samatha meditation, then humans will increasingly exist in their natural state, i.e., they will exhibit noncognitive moral concern. (by 3, 4)

First, I should make a note on the use of the word "increasingly" in premises three and four. This is meant to capture the gradual developmental nature of meditation techniques. Presumably, when one masters the techniques completely, one will enjoy the full fruit of the technique. In this case, to master samatha techniques would mean that one exists totally within human nature and completely exhibits noncognitive moral concern.

The most important exegetical question here is this: Do Buddhist texts explain what human nature is, and if so, what is it? Many texts say that human beings exist in a soteriologically profitable middle state between the gods and animals. Humans both know enough suffering to be motivated to practice Buddhism (unlike the gods) and are capable of making the changes prescribed by the Buddha (unlike animals). In the Sakkapanha Sutta, for instance, a god tells the Buddha that he will happily seek a human existence in the next life so that he can practice the Buddha's teachings (DN, 21.2.8). Given this conception we could perhaps say that human nature is to be able to practice Buddhism. This is not particularly helpful to the human nature argument. At best, it simply tells us that human nature is to be able to practice samatha. It does not give us premise one.

It seems even more doubtful that the No-Self (anatta) doctrine can help here. Does the theory of the five aggregates or the lack of a substantial self tell us whether humans have a natural propensity toward moral concern? Anatta might allow us to say that the grasping for self is what obscures our true nature. But does this indicate that our nature is one that includes moral concern? The five aggregates (khandhas) do not help us determine whether humans are naturally disposed to soteriologically profitable moral concern any more than the axles and wheels of King Milinda's chariot help him determine what is the essence of the chariot (Miln. 2.1.1).

My own understanding of Buddhist texts does not seem to indicate any theory of human nature that can address the exegetical responsibility of this argument. Perhaps such evidence can be found, but for now I must leave the matter undetermined.

Moving to the first sense of philosophical merit, the human nature argument seems incredibly useful, on first glance, in explaining how it is that samatha techniques cultivate moral concern. However, the argument does not make clear exactly what it is that the meditator does to remove vices; it is only slightly more specific than Keown's original description. If I am looking for something to make Keown's argument more complete, then this argument is not satisfying.

In my second sense of philosophical merit, the human nature argument is attractive as a supplement to what Keown says because he writes about human nature on a number of occasions in his chapter on ethics and psychology. There is even a section entitled "The Buddhist View of Human Nature" in the section of the chapter on moral and intellectual virtue (Keown, p. 66-68). The statement that most evokes premise one of this argument is: "The malfunction of vedanā [feeling] and saññā [cognition], which is taṇhā [craving], is the basic soteriological problem of Buddhism" (Keown, p. 67). It would seem that if feeling and cognition are malfunctioning, then they are aberrant from their natural state. For instance, if I say my car is malfunctioning, I mean that it is not in its natural working state. The natural working state of feeling and cognition is exemplified by the enlightened Buddha.

Moving to my third sense of philosophical merit we might ask this question: Just what do we mean by "natural state"? Can "natural state" be taken to mean "the usual, normal state"? Nyanamoli's translation of Buddhaghosa seems to hint at this definition: "But in the world the nature of such and such beings is called their 'habit (sīla)', of which they say 'This one is of happy habit (sukkha-sīla), this one is of unhappy habit, this one is of quarrelsome habit, this one is of dandified habit'" (Vism., 1.38).(9) There seems to be something extremely odd about describing a state instantiated only by fully enlightened beings as "natural" in the sense of "natural" as "usual" or "habitual." There are a great many more unenlightened beings than enlightened ones—that is the problem.(10)

Perhaps by "natural state," we mean something like "working properly," or "working with least resistance." Certainly it is the case that the Buddha provides a model of working both properly and with the least resistance. After all, could we not describe dukkha as a kind of resistance? Given this definition, premise two above makes sense as well. It is unwholesome vices such as greed and hatred that obscure our natural state. By removing these vices, we will no longer face resistance to working properly.

The problem here is that it is unclear whether the simple removing of vices will automatically lead to the cultivation of virtues. If moral virtues are affective states, we can assume that they have something to do with feeling (vedanā). Could there simply be neutral feelings after the removal of resistance to virtues? Buddhist texts indicate that there are neutral feelings: "There are three kinds of feeling: pleasant feeling, painful feeling, and neither-painful-nor-pleasant-feeling" (MN, 44.22). Just because the resistance to virtues is gone does not necessarily mean that virtues will arise.

The proponent of the human nature argument might simply say that virtues will arise because they are what happen in our natural state. But then there is another problem that speaks to my second sense of exegetical responsibility, namely, the problem of causation. In Buddhism, causality is often formulated in the form of Dependent Origination (paṭicca-samuppāda): "When this exists, that comes to be; with the arising of this, that arises. When this does not exist, that does not come to be; with the cessation of this, that ceases" (MN, 79.7). The human nature argument is mixing the two halves of Dependent Origination. The cause of vice may cease, leading to the cessation of vice; but for virtue to arise, the cause of virtue must arise. Keown argues that the cause of virtue is moral concern.

The human nature argument seems to say that the cessation of vice will bring the arising of virtue. Of course, there could be another step here. Our natural moral concern is always trying to shine through the artifice of vice, but it is only when the blinds are opened that we can see through the window. Perhaps there is a latent moral concern waiting for the chance to escape. But where are the causes of these latent causes? Perhaps these causes could be considered as the ripening of karma from past lives.(11) In that case, some people would presumably have more moral concern coming to them than others. This may be true to some extent; however, the Buddhist texts explicitly discuss the cultivation of virtue as something available to any human willing to undergo the practices. Karma could explain why the practice is easier for some than others: some of us have progressed toward a more natural state in previous lives and thus find it easier to remove the vices that cover our natural virtues.

I believe this shows that the human nature argument is not incompatible with Buddhist causality. However, for the sake of philosophical elegance, it would be better to subscribe to an argument that does not require so much theoretical work to be in tune with Dependent Origination.

I have not been able to show that the human nature argument is satisfactory in the first sense of exegetical responsibility and it is problematic in the second. I found no textual basis for premise one, and the argument has some initial, but not insoluble, problems in aligning itself with Dependent Origination.

Philosophically, the human nature argument offers a solution with enough explanatory power to solve our problem (sense one). It seems to be consistent on its own terms and can be seen to fit with Keown's statements on human nature (sense two). However, the definition of human nature may be difficult to argue for in the third sense of philosophical merit due to problems of defining "natural" and whether the lack of vice necessitates the arising of vice.

2.2. The Divine Abidings argument

What I have called the "Divine Abidings argument" can be characterized as follows:

1. If mettā (loving-kindness), karuṇā (compassion), muditā (gladness), and upekkha (equanimity) are made the subject of certain samatha techniques, then they begin to entrench themselves into the psyche.

2. If mettā, karuṇā, muditā, and upekkha begin to entrench themselves in the psyche, then they have a strong tendency to push the main vices—namely, dosa (hatred) and lobha (greed, attachment)—out of the psyche.

3. If mettā, karuṇā, muditā and upekkha are made the subject of certain samatha techniques, then they have a strong tendency to push the main vices out of the psyche. (by 1,2)

4. If mettā, karuṇā, muditā, and upekkha do not either entrench themselves in the psyche or push their opposite vices out by means of the first techniques, then there exist further, more specific samatha techniques that will eventually accomplish this.

5. If mettā, karuṇā, muditā and upekkha begin to entrench themselves in the psyche, they cultivate anukampā (basic sympathy, moral concern).

6. If mettā, karuṇā, muditā and upekkha are made the subject of certain samatha techniques, then they cultivate anukampā (basic sympathy, moral concern). (by 1,5)

C: Certain samatha techniques cultivate moral concern by entrenching mettā, karuṇā, muditā and upekkha in the psyche (premises 1 and/or 4), pushing dosa and lobha out of the psyche (premises 3 and/or 4), and creating affective states that foster moral concern (premise 6).

Beginning with the first sense of exegesis, there is no difficulty finding the Divine Abidings (brahma-vihāra) in Buddhist texts. These practices are explicitly mentioned many times in the Nikāyas (for instance, MN, 55.6 and 83.6, and DN, 13.76). Chapter nine of the Visuddhimagga is devoted exclusively to the practice of the Divine Abidings. Buddhaghosa says that the Divine Abidings are "loving kindness, compassion, gladness and equanimity" (Vism., 9.1). In Pali, these are respectively, mettā, karuṇā, muditā, and upekkha. They are called brahma-vihāra because "these abidings are the best in being the right attitude towards beings. And just as Brahma Gods abide with immaculate minds, so the meditators who associate themselves with these abidings abide on an equal footing with Brahma Gods" (Vism. 9.106).

Chapter nine of Visuddhimagga consists entirely of Buddhaghosa's intricate meditation instructions for the Divine Abidings. I will only briefly summarize them here. As all four Divine Abidings follow a similar basic model, I will mainly discuss the process of mettā cultivation. The four Divine Abidings are meant to be practiced in the order in which they have been listed, e.g., after one has sufficiently developed loving kindness, one moves on to compassion and so forth.

The meditator should begin with loving kindness, concentrating on a human subject. The proper choice for the first subject is explained: "loving kindness should not be developed at first towards the following four kinds of persons: an antipathetic person, a very dearly loved friend, a neutral person, and a hostile person. Also it should not be developed specifically toward the opposite sex, or toward a dead person" (Vism. 9.4).(12)

The proper first subject is one's self, as Buddhaghosa explains: "First of all is should be developed only towards oneself, doing it repeatedly thus: 'May I be happy and free from suffering' or 'May I keep myself free from enmity, affliction and anxiety and live happily'" (Vism. 9.8).

After this process causes practitioners to fill themselves with loving kindness, they should move on to a teacher or similar relation. After that they should develop loving kindness in the following sequence:

He should next after that, develop loving kindness towards a very dearly loved friend, then towards a neutral person as a very dearly loved friend, then towards a hostile person as neutral. And while he does so, he should make his mind malleable and wieldy in each instance before passing on to the next (Vism. 9.12).

If these techniques are not successful at filling the mind with mettā toward all of these subjects, there is advice for several further practices. For example,

If resentment arises in him when he applies his mind to a hostile person because he remembers wrongs done by that person, he should get rid of the resentment by entering repeatedly into loving kindness (jhāna) towards any of the first-mentioned persons and then, after he has emerged each time, directing loving kindness towards that person (Vism. 9.14).

There are other techniques toward the full cultivation of loving kindness as well, such as use of the Buddha's Jataka stories as moral examples (Vism. 9.26-35). Also, there is the realization that, due to the infinity of previous rebirths, everyone has been everyone's mother, brother, sister, etc. in the past (Vism. 9.36) or dissolution of the angry person into elements:

What is it you are angry with? Is it head hairs you are angry with? Or body hairs? Or nails? ... is it the earth element in the head hairs, etc.? ... Or among the five aggregates or the twelve bases ... For when he tries the resolution into elements, his anger finds no foothold (Vism. 9.38).

These practices support premise four above. They are intended to be gradual and practiced repeatedly (Vism. 9.40).

If practitioners are also adept in the first four jhāna techniques (jhānas of form), these practices will lead to various absorptions in which mettā pervades the psyche: "And here, may all beings be free from enmity is one absorption; free from affliction is one absorption ... free from anxiety is one absorption" (Vism. 9.56).

Compare this with a phrase from the Dīgha Nikāya: "so above, below, around, and everywhere, and to all as to himself, he abides with loving-kindness, abundant, exalted, immeasurable, without hostility and without ill will" (DN, 40.9). These quotes show that premise one has textual basis by establishing that the Divine Abidings techniques are both samatha techniques and serve to pervade the mind with virtue.

Premise two can be shown to follow from the following:

For it is not possible to practice loving kindness and feel anger simultaneously (Vism. 9.98).

For it is not possible to practice compassion and be cruel to breathing things simultaneously (Vism. 9.99).

For it is not possible to practice gladness and be discontented with remote abodes and things connected with the higher profitableness simultaneously (Vism. 9.100).

For it is not possible to look on with equanimity and be inflamed with greed or be resentful simultaneously (Vism. 9.101).

Keown indicates that the main vices are hatred and greed (Keown, p. 64-65). Greed is specifically mentioned as wholly incompatible with the Divine Abiding of equanimity. Hatred, as either the same or a close relative of anger and cruelty, can be done away with probably either with loving kindness, compassion, or both.

Is there any evidence for premise five? Perhaps the same passage from the Dīgha Nikāya used above will work: "he abides with loving-kindness, abundant, exalted, immeasurable, without hostility and without ill will" (DN, 40.9). If one is "abundant" and "immeasurable" in one's abiding in loving kindness, it would seem to be the case that one would also have at least a tendency toward sympathy or moral concern. Perhaps this can work. However, the texts are not particularly helpful here so I will address this question as a matter of philosophical concern later in this section.

Before moving on to philosophical considerations, I should say a word about the second sense of exegetical responsibility in regard to Dependent Origination. The Divine Abidings argument gives us an account of the arising of virtue (premise one), passing away of vice (premise two), and the arising of moral concern (premise five). In this way it is superior to the human nature argument because it is a better expression of Dependent Origination and requires none of the theoretical work needed to reconcile the human nature argument with Buddhist causality.

In terms of my first sense of philosophical merit, the Divine Abidings argument presents us with a theory that gives us a detailed account of how moral concern can be cultivated by samatha meditation. Not only does it have more explanatory power than the human nature thesis, but it does this explaining in greater detail.

The Divine Abidings argument is superior in the second sense as well. In fact, Keown specifically mentions the Divine Abidings in the context of his subthesis on the Buddha's moral concern. It appears, conveniently enough, in a section entitled "The Cultivation of Moral Concern":

The states or dispositions cultivated through the Divine Abidings in samatha meditation also occur in waking consciousness in the course of daily life—they are not exclusive to meditation or to the meditator. The technique of transic meditation (jhāna), however, is a powerful device for accelerating their cultivation and pervasion of the psyche (Keown, p. 76).

It may be objected here that, because Keown already connected the Divine Abidings with the cultivation of moral concern, that the Divine Abidings argument is neither original, nor particularly illuminating. My response is that Keown may have mentioned the Divine Abidings as an example of the cultivation of moral concern, but he did not take this connection far enough. He certainly did not make an argument as precise as the Divine Abidings argument, nor did he argue for it as thoroughly. He also did not mention the Divine Abidings in the context of his disagreement with Griffiths. The Divine Abidings argument is what I would like to think Keown should have said in this context.

The most substantive philosophical issue for the Divine Abidings argument is whether the cultivation of mettā can be said to bring about anukampā (sympathy, moral concern) when anukampā is supposed to be the fundamental level of mettā. This would seem to be circular—how can anukampā both cause mettā and be caused by it?

Let me begin with an examination of the word "anukampā":

Etymologically, "sympathy" (anukampā) can be understood as the condition of "being moved" (kampa) "in accordance with [others]," or "in response to [others]" (anu)... there are definitions in the commentaries—"the preliminary level of love" (mettāyapubbabhāga)... Similarly it is said to be synonymous with "tender care" (anuddaya) ... and "simple compassion" (kāruñña) (Aronson in Keown, p. 73).

Strictly on the basis of this etymology, it would seem that there is no reason to wonder how sympathy can be cultivated by mettā and karuṇā. Buddhaghosa gives an etymology of mettā as being related to the word "mitta" (friend) as in "with respect to a friend" or "behavior towards a friend" (Vism. 9.92). Of karuṇā the following etymology is given: "When there is suffering in others it causes (karoti) good people's hearts to be moved (kampana), thus it is compassion (karuṇā)" (Vism. 9.92).(13) If the mind is filled with loving kindness and compassion, for instance, loving kindness would seem to cause one to be moved in accordance with others (as in friendship) and compassion would seem to cause one to be moved in response to others (particularly if they are suffering). Similarly, "tender care" could be taken as an instance of friendliness brought about by mettā. This covers the first two parts and the last part of the above etymology of anukampā.

The problematic parts are "the preliminary level of love" (mettāyapubbabhāga) and "simple compassion" (kāruñña). It seems that the preliminary level or simple state of something must come before that thing. In the Divine Abidings argument, however, the presence of the Divine Abidings in the psyche is supposed to bring about the "preliminary level of love" and "simple compassion."

The solution is related to two more Pali words: sīla and anusaya. Sīla is usually translated as "morality." Nyanamoli, however, has translated it as "habit" in Vism. 1.38, as quoted in the section on the human nature argument. If the Divine Abidings techniques cause mettā to become an entrenched habit of a person, then that person will have the tendency to treat everyone in a friendly manner. The consequent of the previous conditional is indicative of the dispositional nature of anukampā. To have sympathy and moral concern is to be disposed to treat beings with a certain correct attitude—an attitude that having a mind filled with mettā promotes. This same basic strategy can be applied to habits created by karuṇā meditation for fostering the habits that bring about simple compassion.

It may be objected that this move relies too closely on etymology. Because some translators occasionally see fit to translate sīla, as "habit" does not necessarily make it the case that morality and habit are somehow related in terms of Buddhist meditation practices. Secondly, it does not give us a specific enough account of this process. These objections can be answered by giving another account with enough specific meditation connections to satisfy both objections.

Anusaya means "dormant or latent disposition," according to Padmasiri De Silva.(14) Of these he says, "They are basically dormant passions which become excited into activity by suitable stimuli" (De Silva, p. 73). It could be that the Divine Abiding of mettā creates a disposition (anusaya) to act with anukampā when the appropriate stimuli arise outside of the meditation. Then, anukampā gives rise to mettā in nonmeditative contexts. For example, say I practice a mettā Divine Abiding on Tuesday afternoon. This creates a latent disposition for moral concern to give rise to mettā, say when my sister needs my help with painting her house on Friday.

But where is this disposition? Perhaps it is a causal continuum in my consciousness. Surely, if there can be causal continuums across lifetimes, there can be causal continuums from my meditation on Tuesday to the house painting on Friday.

This analysis of the psychological disposition also helps solve another problem: the problem of how it is that absorptions within the Divine Abidings can break through to our normal daily lives—the parts of our lives where morality really matters. By practicing the Divine Abidings, I can change my habits and dispositions in my nonmeditative life. Perhaps this is what Robert M. Gimello means when he says: "The mystical experience affects the moral life, Buddhists believe, and they therefore take the greatest pains in their meditative disciplines to see to it that its effect is the proper, just, and compassionate one" (Gimello, p. 194).

As should be obvious at this point, I find the Divine Abidings argument eminently preferable to the human nature argument. It is more exegetically responsible, both in being close to the texts and agreeing easily with Dependent Origination. It has more philosophical merit in these ways: (1), It offers more explanatory power and more details; (2), it agrees with Keown and is what, in my opinion, he should have said; and (3), the initially problematic premise five can be shown to be philosophically plausible given the notions of moral habits and of latent dispositions.

3. Conclusion: Moral Virtues and the Cultivation of Keown's Thesis

There are some large issues I have not dealt with. For instance, I have not considered the issues of resilience (i.e., how we can be guaranteed the results of samatha techniques will last after the meditative states are finished),(15) whether samatha states are mind-created, what, if any, purpose there might be for the formless jhānas toward the goal of moral cultivation,(16) how to reconcile Keown's internalist theory of motivation with the externalist theory many other scholars consider Buddhism to have, and lastly, whether the Diving Abidings argument gives us any reason to accept Keown's larger thesis that Buddhist ethics can be characterized as virtue ethics. Now that a fuller version of Keown's theory can be established with the help of the Divine Abidings argument, I would imagine these questions can more easily be answered by scholars advocating the equal importance of samatha and vipassanā.

My hope is that the Divine Abidings argument will make Keown's thesis more complete and more plausible. It now has the ability (or at least the latent tendencies) it needs to be evaluated in context of the larger debate on the role of samatha-bhāvanā in Buddhist soteriology. To anyone willing to take up this task, I offer this essay as an embodiment of my mettā and karuṇā.


Buddhist Text Abbreviations

I have used the following abbreviations to indicate Buddhists texts. All are in translation, except the Pali version of the Visuddhimagga edited by Sasrti. Full citations are given under the translator's name in the bibliography below.

DN (Dīgha Nikāya)—Walshe

Miln. (Milindapañha)—Rhys Davids

MN (Majjhima Nikāya)—Nyanamoli and Bodhi

SN (Saṃyutta Nikāya)—Bodhi

Vism. (Visuddhimagga)—Buddhaghosa (author), Nyanamoli (translator)


Bodhi, Bhikkhu (translator). The Connected Discourses of the Buddha: A New Translation of the Sa.myutta Nikāya. Boston: Wisdom Publications, 2000.

Bronkhorst, Johannes. The Two Traditions of Meditation in Ancient India. Stuttgart: Franz Steiner Verlag Wiesbaden GmbH, 1986.

Buddhaghosa. Bhikkhu Nyanamoli (translator). The Path of Purification (Visuddhimagga): A Classic Textbook of Buddhist Psychology. Boulder, Shambhala, 1976.

Buddhaghosa. Visuddhimagga. Edited by Swami Dwarikadas Sastri. Varanasi: Bauddha Bharati, 1977.

Conze, Edward. Buddhist Meditation. London: George Allen and Unwin, Ltd., 1956.

De Silva, Padmasiri. An Introduction to Buddhist Psychology. New York: Barnes and Noble, 1979.

Gimello, Robert M. "Mysticism and Meditation," in: Mysticism and Philosophical Analysis. Edited by Steven T. Katz. New York: Oxford University Press, 1978, p. 170-199.

Griffiths, Paul J. On Being Mindless: Buddhist Meditation and the Mind-Body Problem. La Salle, IL: Open Court, 1986.

Gunaratna, Henepola. A Critical Analysis of the Jhānasin Theravāda Buddhist Meditation. Dissertation, American University, 1980.

Keown, Damien. The Nature of Buddhist Ethics. Chippenham, UK: Palgrave, 2001.

Nyanamoli, Bhikkhu and Bhikkhu Bodhi (translators). The Middle-Length Discourses of the Buddha: A New Translation of the Majjhima Nikāya. Boston: Wisdom Publications, 1995.

Rahula, Walpola. What the Buddha Taught. New York: Grove Weidenfeld, 1959.

Rhys Davids, T. W. (translator). The Questions of King Milinda Translated in Two Parts. New York: Dover Publications, 1963.

Solé-Leris, Amadeo. Tranquility and Insight: An Introduction to the Oldest Form of Buddhist Meditation. London: Rider, 1986.

Walshe, Maurice (translator). The Long Discourses of the Buddha: A Translation of the Dīgha Nikāya. Boston: Wisdom Publications, 1995.


(1) The earlier existence of samatha and the peculiarly Buddhist innovation of vipassanā have been established in a thorough exegetical and philological study by Johannes Bronkhorst in The Two Traditions of Meditation in Ancient India. Return to text.

(2) Citations from Buddhist texts will be given by abbreviations listed at the end of the essay. Return to text.

(3) Conze deals extensively with Mahāyāna in addition to Theravāda, but his respect for Theravāda can be seen in the following quote about Buddhaghosa's Visuddhimagga: "he has composed one of the great spiritual classics of mankind. If I had to choose just one book to take with me on a desert island, this would be my choice,—with perhaps a Horace tucked away out of sight in my pocket" (Conze, p. 25). Return to text.

(4) Most scholars prefer not to translate jhāna. Those who do often translate it as "absorption" or more controversially as "trance." Return to text.

(5) Whether the establishment of this level of importance for Buddhist morality would indeed point to the success of Keown's contention that Buddhist ethics can be characterized as a virtue ethic is an issue I will not address in this essay. Return to text.

(6) Gunaratna seems to share Griffiths's assumption: "Since bondage ultimately springs from ignorance (avijja) the key to liberation, for Buddhism, is found in wisdom (paññā)" (Gunaratna, p. 3). Return to text.

(7) Anukampā is the term Keown uses for "moral concern" as well. It will receive more treatment toward the end of section 2.2. Return to text.

(8) In mapping these debates onto Western philosophy, Keown is closer to philosophers such as Aristotle or Hume and the externalists are more akin to Plato or Kant. Return to text.

(9) It would seem that his translation of sīla as "habit" is taken in a different sense than the other rendering as "morality." Interestingly, the relationship between morality and habits would probably please Keown greatly given his characterization of Buddhist ethics as related more closely to Aristotle's Virtue Ethic than to other Western theories. Return to text.

(10) Of course, some Mahāyāna schools could perhaps answer this objection by saying that all beings have Buddha Nature and already are enlightened in some sense. This option, unfortunately, is not available in Theravāda. Return to text.

(11) De Silva says that the word anusaya can be translated as "dormant or latent disposition" and "The Buddha would consider dormant learnings as persistent traits coming down innumerable lives" (De Silva, p. 74). Return to text.

(12) There are various reasons for these exclusions. For instance, development toward a neutral person is "fatiguing," toward a hostile person it causes anger, and toward the opposite sex it causes lust (Vism. 9.5-6). Return to text.

(13) On the basis of this etymology it would also seem that anukampā and karuṇā are etymologically related. Return to text.

(14) De Silva intends these dispositions to be characteristic of the vices. Here, I take hold of them and direct them toward the service of virtue. Return to text.

(15) I imagine the answer would have something to do with dispositions (anusaya) as discussed near the end of the previous section. Return to text.

(16) My wild conjecture on this question is that after the Divine Abidings have created moral concern, the formless jhānas come in to root out the affective causes of saṃkhāras (complexes) that are responsible for our deepest dispositions toward vice. Return to text.


Source: Journal of Buddhist Ethics, http://jbe.gold.ac.uk 

[Back to English Index]
last updated: