Praise and Blame

Joel Feinberg observed that “moral responsibility… is a subject about which we are all confused” (1970: 37). Perhaps nowhere is this confusion more evident than in our understandings of praise and blame. This entry will contrast three influential philosophical accounts of our everyday practices of praise and blame, in terms of how they might be justified. On the one hand, a broadly Kantian approach sees responsibility for actions as relying on forms of self-control that point back to the idea of free will. On this account praise and blame are justified because a person freely chooses her actions. Praise and blame respond to the person as the chooser of her deed; they recognize her dignity as a rational agent, as Kantians tend to put it. This approach sharply contrasts with two further ways of thinking about the issues. One is utilitarian, where praise and blame are justified in terms of their social benefits. Another, more complex approach is roughly Aristotelian. This approach situates practices of praise and blame in terms of our on-going relationships with one another. This approach stresses the importance of mutual accountability, moral education, and assessments of character in terms of the many vices and virtues.

1. Introduction

This article will not try to convey the exact details of these accounts, but to show how these ways of looking at mutual accountability capture important parts of our everyday commonsense. One modern commentator claimed that, in our attitudes to moral responsibility, “we are all Kantians now” – by “we” meaning not just philosophers but all Western persons (Adkins, 1960: 2). Another central figure in this debate, Bernard Williams, agrees that Kant captured a widespread tendency of modern moral thinking, but also claims that there exist important counter-tendencies in our actual practices of responsibility. For Williams, ancient Greek understandings are actually more realistic and helpful than the Kantian one. So far as our modern praising and blaming actually make sense, he claims, they are better captured by a (roughly) Aristotelian account.

There are some important differences between praise and blame that will not be central to this entry; in fact, blame will get the greatest attention here. This is partly because praise seems less problematic: misplaced blame is felt as deeply unfair, not least because being exposed to blame is unpleasant and costly in a way that being praised is not. But it is principally because blame has a closer connection than praise to matters of intense philosophical interest, including freedom, responsibility and desert. We often praise inanimate objects (such as art works or buildings) and animals (a loyal pet, for example), although we could not blame such entities, however deeply dissatisfied we felt with them. The focus of this article, however, will be upon entities that are clearly open to blame as well as praise: human beings.

What is blame, such that only human beings can be blamed? We are all familiar with resentment, reproach and accusation regarding a person’s past actions; likewise, we all know the sense of guilt, shame or indignation they can elicit. Philosophers differ on how far certain emotions may be central to blame (this relates to a wider dispute, regarding which emotions, if any, constitute a proper basis for moral action). What is clear is that blame suggests both responsibility and culpability. Here, responsibility only implies that the act can be identified with a person, such that she can reasonably be expected to respond for it in some way. That is, it does not necessarily imply fault, or culpability. This is the idea that the person is “in the wrong,” that fault somehow attaches to them so that they deserve blame. (Philosophers tend to describe this as “blameworthiness.”) What sense we should give to these ideas of culpability or desert, and what is necessary for us to think of a person as responsible: these are central issues for this entry.

For further aspects of responsibility, see the sister entry to this article, responsibility. Another article also examines the topic of free will in depth. Nonetheless, since Kant’s account begins with the question of free will, it is also necessary to say something about this straightaway. The entry will then set out the utilitarian and Aristotelian accounts, before returning to Kant’s theory. It concludes by discussing ideas of moral worth and desert that make Kant’s account so appealing.
2. The Problem of Free Will

The free will debate has become an old chestnut of modern philosophy. It is an intuitively plausible way of approaching the issues – familiar to many even before they encounter philosophical texts. It is perhaps surprising, then, that this debate is actually a rather modern one.

The basic gist is this: if I am to be responsible (really responsible) for my conduct, then it must be within my control. However, if it is true that every event in the universe is determined by causal laws, then this must be true of the events that constitute my actions. Therefore, my conduct cannot really be within my control; therefore, I am not really responsible for my conduct. Two conclusions immediately suggest themselves. One is that it is incoherent to praise or blame me – and everyone else – for our actions, because it is so difficult to doubt the causal well-orderedness of the universe. The alternative conclusion, scarcely more appealing, is that the human will somehow sits outside this causal framework – ie, we have free will – because it is unthinkable that our moral ideas be so desperately incoherent.

Both lines of thought are incompatibilist; that is, they see the ideas of responsibility involved in praise and blame as incompatible with the causal well-orderedness of the universe. But while both attract some limited support among philosophers, the overwhelming consensus now lies with compatibilism. This is simply the thesis that responsibility and causal order are compatible. Most philosophers agree that the alleged incompatibility results from some important confusions, although there is much less consensus about what these may be. At least one area of confusion is clear, however, and forms the central issue of this article: what sort of responsibility for conduct is involved in praise and blame? Several familiar points in the free will debate are helpful for approaching this.

In the first place, it is well-known that this debate does not turn on the truth of determinism as such. Determinism is the idea that every event is determined by fixed causal laws. Yet it may well be that every event is somehow random in origin. One interpretation of quantum physics claims that causal laws are the product of statistical regularities, while these regularities stem from a near infinite number of random events. So far as the human will is concerned, this makes no difference. If my conduct is the product of chance, this makes me no more responsible for it than does its being generated by causal laws. The point is that if I am to be blamed or praised, then I must control my conduct – not causal laws, nor mere chance, nor some particular combination of the two.

Second, the free will debate bears a disquieting similarity to an older controversy. In medieval philosophy it used to be asked how God’s omniscience – his knowledge of everything that has happened and will happen – could be reconciled with our being subject to his moral judgment (that is, being sent to heaven or to hell). If God knows what we will do then this seems to imply that it is already decided whether we will act well or badly. And this, in turn, suggests that it makes no sense to punish or reward us. Theologians developed various doctrines to overcome this difficulty, but few sound convincing to modern ears – perhaps because the problem itself is no longer a live one, even for most believers. However that may be, it is interesting that many modern versions of the debate seem to take at least one of the planks of Christian theology for granted: that individuals have wills that can be bad or good, usually now expressed in the terms of people’s “blameworthiness” or (less often) “praiseworthiness.”

In this way, the modern American philosopher Joel Feinberg ironically referred to “a moral bank account” that we carry through life, which sums up our moral credits and debits in a single sum (1970: 20). Whether or not such an “account” makes sense, it is at least clear that the idea of “the will” is by no means self-explanatory. For Kant, as we shall see, it was obvious that all my choices can be summed up in a single moral evaluation, whether I have a “good” or “bad” will. Kant is equivocal, however, as to whether only God might make this evaluation, or whether human beings might also form reasonable opinions on the matter. But especially if we take the point of view of mutual, human accountability, it is not obvious why we should believe any such single evaluation to be possible, or what role this evaluation might play in our individual or collective lives. Certainly, we usually praise and blame in terms of particular actions and particular vices and virtues – not a good or bad will.

Third, this way of framing the issues creates a problematic gulf between normal moral agents (adult human beings of sound mind) and other creatures – animals and children. At some stage of evolution, and at some stage toward maturity, certain animals become “free,” when before they had all been determined in their conduct. Although it is grossly implausible that there are no relevant moral differences between the other animals, children, and human adults, it is no more plausible that the free will simply pops into existence at a certain stage of human development. (Within a Christian framework this issue was less problematic: human beings, and only human beings, have souls.) Nonetheless, we tend to think there is something sufficiently distinctive about human action, so that many non-religious people find the idea of free will plausible, and almost everyone assumes that blame (if not praise) only makes sense with regard to (mature?) human beings.

Taking the last three points together generates a further point. If the idea of the will is complex, and there is no straightforward moral dividing line between children and adults, between humans and other animals – together, these ideas suggest that a “will” is not something we all straightforwardly “have.” In other words: it is implausible that all adult humans have the same capacities, all to the same extent, that are involved in controlling action. One way of retaining the idea of the will might be to think of it as the bundle of capacities that are needed to control action in the light of moral concerns, these capacities being set only at such a level that all adult human beings of sound mind really seem to possess them. But two points need to be kept in mind about such a strategy. First, it remains the case that people will vary in how far they possess such capacities, and this variation will largely be a product of upbringing and natural qualities – that is, not something within an individual’s own control. Second, the sort of ultimate control over one’s moral character supposed in Kant’s or similar “free will” accounts is unlikely to be vindicated in this way.
3. Two Contrasting Approaches

Two influential lines of thought oppose the idea that praise and blame relate to “free will,” the metaphysical idea that we are responsible for our action because they are controlled by us and not (simply) caused by the world around us. For the utilitarian, praise and blame, like all our other practices, can only be justified in terms of their social consequences. A more complex account was given by Aristotle, who shares the utilitarian’s sense that praise and blame have important social consequences, but also offers an extended account of how they relate to the capacities needed for moral action.
a. The Utilitarian Account

The utilitarian case is straightforward. Blame and praise encourage us to perform socially valuable actions and to avoid socially costly actions. If we know we will be blamed for greed or cruelty, for example, then we have powerful motives to avoid these. Praise and blame also involve us in making assessments of people’s strengths and weaknesses, which is important when it comes to deciding who should be entrusted with which tasks and responsibilities. The stingy person might make a good banker, but a bad organizer of social occasions.

This approach does seem to capture important truths: we want to encourage and discourage different sorts of activity, and we need to have a sense of what different people are good at. It also makes sense of why we don’t blame some actions, even if they had bad outcomes (even though, in principle, only outcomes matter to the utilitarian). If the bad outcome was not chosen by the person (for example, she was forced to act that way by someone else), then there is nothing to be gained from blaming them (much better to blame the person who forced her). Thus the utilitarian can accommodate the important fact that praise and blame relate to free action: but this need not be thought of in terms of metaphysical “free will,” but instead the compatibilist freedom involved in choosing one’s actions independently of others’ interference.

But the utilitarian account faces a simple objection: does it really provide for responsibility, still more culpability? For example, if we know that someone does not respond well to criticism, it seems that the utilitarian case for blame is undermined. We would do much better to flatter and cajole them into acting differently. Of course, the utilitarian might reply that this is often what we in fact do with such people. Further, he might add that we do still blame such people when we discuss their characters behind their backs, perhaps describing them as self-righteous or stubborn. What seems to be missing in this response, however, is the idea that the person deserves blame. They seem to deserve criticism in just the same way that a faulty machine or a cracked mug deserve criticism: it’s useful that everyone knows they’re faulty, but they can hardly be described as blameworthy. Especially when we move from blame to the question of sanctions or punishment, this lack of desert seems to present a real problem for the utilitarian account.

Utilitarians face a more complex criticism, which goes beyond the scope of this entry. Historically more concerned with the actions of government than individuals, utilitarianism never developed a realistic moral psychology – that is, very roughly, an account of what makes the decent person tick. This lack of attention has permitted some of the most devastating critique of utilitarianism, such as Bernard Williams’s and Susan Wolf’s. But if we want to understand responsibility, our capacity to accept praise and blame as well as our tendency to dole them out, then we need to have a fairly good picture of moral agency.
b. The Aristotelian Account

This is where Aristotle’s more complex account enters the story. The most famous discussion of when people can be praised and blamed for their actions remains Aristotle’s. As with the utilitarians, Aristotle saw no need to talk about praise and blame in terms of free will. Aristotle speaks of whether acts are voluntary, and whether we attribute them to a person or to other factors. Some have ascribed this way of framing the issues to a lack of moral or scientific sophistication on the part of the ancient Greeks. However, a number of modern philosophers, most prominently Bernard Williams and Martha Nussbaum, have suggested that an Aristotelian account is actually more coherent and sophisticated than those typical of modern philosophy – and, indeed, more coherent than our modern, “common sense” intuitions about moral responsibility.

At first glance, it looks as if Aristotle takes it for granted that we are responsible for our actions, so that others can reasonably praise or blame or punish us. What he does is to highlight various conditions that lessen or cancel our responsibility. He discusses force of events, threats and coercion, ignorance, intoxication and bad character. Yet, taken together, his account shows us the basic elements involved in being a person who can reasonably be praised or blamed.

The first limitation upon voluntary action that Aristotle discusses is force of circumstances. His well-known example concerns a ship caught in a storm; the sailors must throw goods overboard if the ship is not to sink (NE 1110a). In this case the action is not fully voluntary, and we would not blame the sailors for their actions. (Nor, of course, would we blame the storm: the undesirable consequence, the loss of the goods, must be chalked off as the product of natural causes, for which no one can be blamed.) Note that such cases are extreme examples of the force of necessity under which we always live – we are always constrained in our actions by circumstances, although we only tend to notice this when the constraint is sudden or unexpected. (If blame were to arise in such a situation, it would be where the sailors failed to take account of necessity, so that the ship and many aboard perished.)

In fact, it tends to be the interference of other people that causes us the most grief – and which really causes problems for responsibility attributions. Such interference can take many forms, but its paradigmatic forms are coercion and manipulation. Regarding coercion, Aristotle’s judgment is balanced. It depends on what action my coercer is demanding of me, and what threats he makes. Some actions are so heinous that we should be blamed for doing them, whatever we are threatened with (and whatever blame also attaches to our coercer) – thus Aristotle dismisses the idea that a man might be “compelled” to kill his mother (NE 1110a). This makes it clear that a central issue at stake in attributions of responsibility is the expectations that people have of one another. There are some forms of coercion we do not usually expect people to resist, but there are also some sorts of action that we think people should never undertake, regardless of such factors. In such cases praise and blame are clearly working to clarify and reinforce these expectations – in other words, they provide for a form of moral education.

Aristotle does not comment on manipulation, where other people lead us to a false view of our circumstances. But he does discuss ignorance of these circumstances, and how it undermines our responsibility. If we are ignorant of who someone is, for example – as was Oedipus, who did not know that the old man obstructing him was actually his father – we may commit acts we would otherwise abhor – thus Oedipus committed patricide, killing his own father. For Aristotle, such actions are not to be blamed (with the important provisos that the ignorance is not itself culpable and the action was otherwise justified). What decides good or bad character is how a person reacts when he finds out the truth – if we fail to regret our deeds, then we can certainly be blamed, even if the original choice was justifiable. Our regret about the deed shows that we want to disown it, and prepares us to make up for it as best we can. A lack of regret shows we are happy for the deed to have been done anyhow, even though we are now aware of facts that others think should have prevented us from acting that way.

This argument hints at an important point. For Aristotle, the moral judgment of the self may be quite different from the judgments of others. The actor should regret his action deeply but, as long as he does so, on-lookers should not blame, but rather pity or perhaps console him. If we suppose that both actor and on-looker are making a judgment about the actor’s moral worth this seems puzzlingly inconsistent. Yet Aristotle’s account has a different logic: The actor’s regret reveals his determination not to be associated with such an action. The on-lookers’ pity relates to their awareness that this “self-blame” is proper yet not earned; it is something that could fall upon anyone in the wrong circumstances. Simplifying, we could say that on-lookers make a positive judgment of the actor, based on his preparedness to make a negative judgment of himself. But this is not so paradoxical if we think of these judgments, not as relating to moral worth, but as preparations for action. Something has gone wrong, after all, and those affected seem to deserve some recompense. In such a situation, the actor will feel duty-bound to help put things right (perhaps to compensate, at any rate to apologise or show remorse). On-lookers, pitying rather than blaming, try to make his task easier, since the responsibility, in such a case, was not earned by the actor.

We have just discussed actions done in ignorance of the facts. But not every form of ignorance excuses; factual knowledge is very different from moral knowledge. What if a man did not know murder was wrong? Would this make his murders morally innocent? Aristotle says not: there are certain things we can and do expect people to know – above all, basic moral truths such as the wrongness of murder. But this knowledge is not as straightforward as it might appear: it must include a fairly good capacity to judge which sorts of killing count as murder. Nazi bureaucrat Adolf Eichmann organized the killing of thousands, without a sense of its wrongness. Aristotle is clear: such moral ignorance, an inability or failure to judge, excuses no adult. Eichmann should be held responsible for murder. But why should moral ignorance not excuse, when factual ignorance does? We must recognize that moral knowledge is actually rather different from factual knowledge. If a person is morally ignorant it is his whole character, his lasting ability to judge and act well, that is impaired – and presumably very difficult to set right. Isolated errors in factual knowledge, on the other hand, can be easily corrected. So long as we subsequently recognize and regret what we have done, factual mistakes involve no lasting corruption of character.

Still, if a person is morally ignorant it follows that they are unable to choose well. Aristotle agrees, arguing that those of settled bad character – be they morally ignorant or otherwise – are unable to make decent moral judgments. Does this mean that blame is incoherent or misplaced? He claims not. Even if the vicious person cannot now choose to act otherwise, there was a time when her vices were not fixed, when she could have chosen not to be vicious. Therefore, Aristotle says, she can be blamed. This is neat but rather unconvincing. Aristotle is famous for emphasising the importance of good upbringing and habituation, and presumably many vices are formed in childhood, before people have formed capacities for deliberating reasonably. Indeed, many vices undercut the capacity for rational deliberation. So it is a clear implication of Aristotle’s own account that the badly brought up person may never be in a position to choose not to be vicious. Note, further, that this move represents Aristotle at his most Kantian: blame is justified by reference to control, to a “could have done otherwise” – even when his own account of character formation suggests that such control probably never existed.

What are we to say, then, when a person seems unlikely to change: she appears quite settled in some particular vice, either because she cannot understand the criticism or because she is unable to alter her character or habits? Such cases are very common, and – unless we suppose that they are not morally deplorable – seem to undermine the modern assumption that blame must relate only to conduct under our control. (The same sort of argument can also be made with praise: a virtuous person might be quite unable to do certain things – commit cruelty, for example.) Clearly, if we think a character trait is really beyond alteration, by us or by the person concerned, our blaming won’t involve an attempt to reason with the person we condemn. But our condemnation might have another rationale: for example, to clarify what sort of standards we expect of others, or to signal our fellow-feeling with those who have been adversely affected by someone’s vices.

In sum, Aristotle’s account is not entirely self-consistent. Generally his focus is two-fold: upon the qualities of character revealed by acts, in terms of our overall moral expectations; and upon the responsibilities that must be born, given the effects of an action. For most of the time, his account proceeds without much reference to desert, and it is this neglect that seems to pose the chief difficulty for the Aristotelian story. It is interesting, then, that Aristotle himself sometimes suggests that bad qualities are to be blamed because they were originally subject to choice, even though this quasi-Kantian claim is not (on his own account of character formation) really supportable. Whether or not Aristotle should have made this argument, it does show how powerful is the thought that blame must be justified in terms of what the person herself chose – however long ago that choice supposedly was made.

Despite this, philosophers have returned to Aristotle’s account again and again to illuminate key ingredients of responsible agency.

* The capacity to respond to others’ censure and encouragement, whether expressed emotionally (eg, as resentment) or in the more articulated forms of praise and blame.
* A reasonable grasp of how actions are understood by people around us and how they affect others, including the need to share out responsibilities for “patching things up” where something has gone wrong. (That we praise and blame children, however, emphasises the educative and encouraging role that praise and blame play in developing such knowledge.)
* Together with our own ability to express judgments of others, these capacities allow us to participate in forms of mutual accountability, whereby we inculcate and to some extent enforce shared standards of action.

This list is not comprehensive, but it serves to illustrate the underlying point of an Aristotelian account: our praising and blaming of one another rest on these sort of fairly basic capacities, which do not seem to demand any strong metaphysical elaboration. Indeed, if we approach the matter this way, the puzzle seems to be inverted. Not, “how might free will and determinism be reconciled?;” rather, “why should we feel there is a metaphysical issue at all?”
4. The Kantian Account and Moral Worth

We have seen that the Aristotelian and utilitarian accounts face a common criticism. Illuminating as they may be, they seem to pay too little attention to the question of desert, or culpability. Is the vicious person blameworthy? Does the person of good will, however much she is hindered by bad luck and hard circumstances, not deserve moral recognition? Our intuitions tend to answer such questions affirmatively. And the most usual justification is that the bad person has less moral worth than the person of good will, and therefore deserves blame and perhaps even punishment. A utitilitarian such as JJC Smart sees such justifications as “pharisaical” – that is, as hypocritically self-righteous, and encouraging of excessively moralistic forms of blame and retribution. But there is no denying the power and influence of such justifications.

The reason why so many people – within and without academic philosophy – feel the pull of the free will debate lies in the idea of moral worth we often associate with responsibility attributions such as blame. Galen Strawson expresses the core idea as follows: “if we have [true responsibility], then it makes sense, at least, to suppose that it might be just to punish some with eternal torment in hell, and reward others with eternal bliss in heaven” (1991: viii). Any such “ultimate” merit or demerit clearly has to be a matter of strictly individual desert. If it were merely a matter of chance who went to heaven or hell – or who would do so, if those fates really existed – this would plainly be a matter of mere fortune. Such intense good or bad luck would make the world even more morally arbitrary than it already is. If such merit is to be fairly allocated, therefore, it needs to be seen as something that lies within individuals’ own control. This line of thought, in turn, is based on what John Skorupski calls an “ideal of pure egalitarian desert” (1999: 156). Modern morality regards each person as equal in moral standing, as having an intrinsic dignity and deserving of equal respect. The thought is that we all equally possess control over our will, so that it makes sense to imagine everybody reaping an equally fair return on how well we exercise that control. (Clearly, this line of thought goes against the idea of the will referred to above, as a “bundle” of capacities unequally distributed among human beings.)

The thinker who grapples most systematically with these questions is Kant. He sees us all as equal in our capacity to strive for morality. But he knows that we don’t all do this, and claims that only some are worthy of happiness.

For Kant, our moral worth – the goodness of our will – is gauged by how sincerely and persistently we have sought to do our duty. To do our duty may be much harder for some people, for instance, those who have violent passions or who were brought up with bad habits. But moral worth is not about results; it is about the will. We all have such a will, an ability to choose well, despite the fact that some of us face stronger counter-inclinations or more difficult circumstances. To truly judge a person’s moral worth involves seeing past all the obstacles that their will has faced. Kant argues that this makes moral worth impossible for us to judge with any assurance; only God can see beyond all those things. This lack of knowledge corresponds to Kant’s main concern, which is how we judge ourselves. Our concern should be to do the right thing, and to do it because it is the right thing. To Kant it’s no problem that we’re never sure about others’ wills, and the obstacles or benefits they have faced. The point is that we can never be sure of our own motivations, and must always be attempting to do better in the future.

Moreover, Kant claims we are all equally well able to see what we should do. For Kant “even the most hardened scoundrel” would act morally, were it not for the opposing incentives of his inclinations and desires (Groundwork, 4:454). Kant needs to claim this because otherwise he would not be able to justify condemning people who suppose they are doing the right thing, when in fact their acts are quite wicked – the problem of the self-righteous wrong-doer. Adolf Eichmann, who we mentioned before, seems to have been sincere in thinking his acts were defensible (he even justified his actions with a twisted version of Kant’s moral philosophy!). Yet no one, and certainly not Kant, would doubt that he deserved the gravest condemnation for his crimes. In simplest form, the Kantian thought is that, if only we wanted to, we could all see that certain things are wrong – for example, no one could possibly want a world where everyone committed actions like Eichmann’s. Nonetheless, such examples are problematic for Kant, because it does seem implausible that people are equal in their capacities for moral knowledge. People’s sensitivity to different moral considerations is highly variable, and is clearly shaped by up-bringing and environment.

(By way of contrast, it may be worth noting that from an Aristotelian perspective, the realities of moral ignorance and moral disagreement pose no theoretical problems. In fact, they provide an important justification for praise and blame in terms of mutual accountability – that is, they help with moral learning by communicating when we have met or failed to meet moral standards. But because Kant’s account goes inward, to my scrutiny of my motives and intentions, he says remarkably little about this crucial educative aspect of responsibility attributions.)

Modern Kantian writers differ on how to deal with these two issues, the invisibility of the will and the claim that we share equal access to moral knowledge. One important line of thought is Christine Korsgaard’s. When we blame someone, she claims, we are recognising his capacity to reason about his conduct. Many people have felt that it is “enlightened” not to blame people for bad conduct, and instead to offer explanations that excuse or mitigate – for instance, by taking a person’s anti-social behaviour to have been caused by a bad childhood rather than a bad will. But Kantians insist that this is to deny someone recognition as a rational agent, as someone capable of choosing his action in the light of reasons. This corresponds to the important intuition that there is something patronising about making excuses for people, and not taking their own point of view seriously.

It is not clear whether blame, on this account, need have any link with the idea that someone’s will has proved defective; and it is this which is important if we are to give a place to culpability within the Kantian schema. Modern Kantians usually concede that Kant was too optimistic about our ability always to see the right thing to do. In this case, it is sometimes difficult for us to judge correctly, and so we have to work together at discovering the moral standards applicable in complex situations. Clearly, then, we need to communicate concerning the rights and wrongs of our individual actions. What this seems to omit, however, is the fact that desert is in play when we blame: blame often has an emotional content, and rarely sounds like a disinterested conversation about what would have been the right thing to do. One reason for this, in turn, is that we are identified by our acts, and tend to identify ourselves with them: if our acts are faulty, and none of the standard excusing conditions apply (such as factual ignorance, as discussed by Aristotle), so too must our character be, if blame is to be deserved. (On the other hand, perhaps it is true that we tend to “take things too personally.”)

This points to a real difficulty for Kantians. Moral evaluation is supposed to concern the will, not all the other complicated factors that have formed our character. (Aristotelians, and many others, reject the idea that such a separation can be made, even in theory.) Although Kantians think such a separation is theoretically possible, in practice they concede that we can only guess at the will. This seems to suggest that we should not blame one another, inasmuch as blame implies culpability, an individual failure to will rightly. But this leaves us with two unrealistic alternatives. One is that we explain bad conduct in terms of mitigating factors, which is plainly unattractive, for the very good Kantian reason that it fails to respect people as the choosers of their deeds. Yet the other obvious alternative, that instead of blame we should pursue an enlightened, as well as enlightening, conversation about correct responses to situations, is patently unreal. If people as we know them are going to change, or learn, by and large it will not be unemotional reasoning that alters them, but the many forces that speak to all aspects of character – for instance, resentment, shame, force of opinion. Yet, for all that these characteristic aspects of blame do not operate on the will (as Kantians conceive it), they certainly convey moral disapproval, and can be very effective.
5. The Idea of Moral Worth

The notion of moral worth central to Kant’s account is probably what one writer on ancient Greek ethics – AWH Adkins – had in mind when he said, “We are all Kantians now.” (1960: 2) Kant’s idea attractively reconciles two broad value judgments: (i) the egalitarian idea that all persons are moral equals by virtue of having freedom to choose morally; and (ii) the idea that responsibility relates to desert, so that people can nonetheless be judged very differently – some being condemned for their lives and characters, others praised. Although we have seen serious problems with the idea that people have an equal ability to choose well, most people agree that blame which attaches to parts of our character that we cannot control is deeply unfair. Does this mean, then, that we should accept a Kantian idea of moral worth, where praise and blame are understood as responses to people’s ultimate deserts?

To begin with, contrast Kant with Aristotle. Aristotle makes no claims about a person’s ultimate merit or demerit. People might be vicious or virtuous in various ways, and there might be rare paragons who possess a comprehensive set of virtues (yes, these are philosophers). Naturally we would not want to associate with the vicious, and naturally we will want to condemn their vices in no uncertain terms: It might help them to learn to do better, and it may caution others against them, and it should reinforce our own and other people’s sense of what character traits are desirable. But for Aristotle there is no sense that the vicious are earning a lasting form of discredit that should condemn them in the eyes of an ultimate judge. If the vicious person were to protest to Aristotle that the condemnations he faced were unfair, perhaps because his character had been shaped by his vicious parents, one suspects Aristotle would be rather unmoved. Life isn’t fair, he might say, and we certainly won’t make it fairer by pretending some vices are less real because of their origin in early childhood, let alone because of their fixity within an individual’s character. It may be unpleasant (he might continue) for you to hear this blame and condemnation – indeed, I’m glad that it is, because at least it shows that you are not so vicious that you don’t care about others’ opinions of you – but there are other matters at stake here, above all the standards and expectations which regulate all our lives together.

So Aristotle’s characteristic view is that some people just are better than others, in their abilities to choose rightly as in other regards. Given this “brute fact,” it is all the more important to give attention to mutual moral education and ensuring that people feel the need to take responsibility where things have gone wrong. Yet it does seem true that Aristotle paid too little attention to the question of desert. We can see this by recalling that he is not wholly consistent here. As we saw, he does try to justify our blame of the vicious person in terms of that person’s choice to become vicious, supposing that otherwise our condemnation would be unfair. Nonetheless, the main thrust of his account seems to be that Kant’s egalitarian fairness is not something we can really achieve.

On the other hand, it is difficult to deny the basic, very appealing intuition of Kant’s ethics: that people’s happiness should correspond to their moral worth – to the sincere intentions that are within everyone’s control. Apart from its appeal to fairness, this conception is also plausible because it corresponds well to several features of praise and blame. We do tend to judge the intent behind people’s actions, rather than the often haphazard results of their deeds. We take account of people’s circumstances, and judge less harshly where these place hard or immoral pressures on people. We also, quite often, feel that allowances should be made for the effects on character of abusive or deprived upbringings. In each case, we can interpret these concessions in Kantian terms – as drawing a distinction between the person’s will and the obstacles of circumstance, thus keeping our moral evaluation to what is within a person’s control – and, therefore, what concerns their deserts.

There are, however, reasons to doubt whether this Kantian interpretation is really the best account of these intuitions. The most obvious problem is that we often expect people to take responsibility for things they didn’t intend. This is not only in those cases where we judge that someone should have formed their intentions more carefully. Certainly we judge the negligent driver who causes an accident more harshly than a driver who was careful but nevertheless caused an accident. But even in the latter case, we expect the driver to bear important responsibilities. The problem that many of the things which attract moral culpability are wholly or partly outside of individual control is connected with the problem of moral luck. It is important to realise, however, that this problem is based on the Kantian idea that moral judgments, be it of character or future responsibilities, are deserved because they relate to a person’s “moral worth.”

Aristotle’s account offers a different way of understanding these everyday intuitions about when blame is justified. On his account we are judging the character of the person we are dealing with, based on how they act, how seriously they take their responsibilities, and how they respond to others’ responsibility attributions. To judge such questions we do indeed give a lot of weight to a person’s intentions: obviously, an intended action reveals a person’s character especially clearly. At the same time, we need to appreciate what he knew about the situation he was responding to, what pressures he was under, and special factors affecting his ability to deliberate and choose. Hence Aristotle’s concern with factual ignorance, force of circumstances, and intoxication; and we might note the more modern concern with mental illness. On an Aristotelian line, the point is that these factors alter the extent to which actions reveal the character of the person. That they undermine the person’s “control” is true, but subsidiary. To support this thought, we might consider how certain forms of bad character constitute a lack of control over one’s actions – thus the person who is weak-willed or indecisive, for example. Here weak-willed, indecisive action reveals the person, and her inability to control her actions.

This suggests that we do not need to accept Kant’s will-based view, where blame relates to moral worth. But we might still wonder if the other accounts can explain the culpability aspect of blame, the idea that it relates to desert.

Both utilitarians and Aristotelians can agree that at least one sense of desert clearly applies. A person deserves to be judged accurately, just as the facts deserve to be assessed truly, if they are to be assessed at all. As we need to judge one another, then clearly we deserve to be assessed fairly. But this doesn’t quite take us to the idea that a person has earned blame, for the fact is that a negative judgment of our character is unpleasant and costly. After all, human beings understand such judgments, and feel their effects, in a way that other entities do not.

There is another question of desert: praise raises the possibility of reward, while blame almost automatically suggests we ought to do something to make up for what we have done or how we have been. Moral philosophers continue to dispute whether utilitarians can give a proper account of this sort of responsibility. But we have already seen how Aristotle could respond. On his view responsibility attributions have a practical aspect: they are preparations for action. It is obvious that when something has gone wrong, we need to distribute the resulting responsibilities: who should pay compensation, apologise, or even be punished. If we take the view that there are always duties to be done, including making good when things have gone wrong, then the question is not what the results say about people’s moral worth, but rather how responsibilities for making good can be fairly divvied up.

But whether this is enough to justify the sense of desert that tends to attach to judgments of blame, or whether we tend to be too keen to invest blame with ideas of personal desert – these are questions much beyond the scope of this entry.
6. Conclusion

Praise and blame relate to our sense of people as capable of taking responsibility for their actions. As we saw, ideas about responsibility are usually presented in terms of a contest between two positions, compatibilism and incompatibilism. Incompatibilists accept the dilemma of free will versus determinism: responsibility depends on me controlling my actions, rather than other causal influences that operate around me. Praise, but especially blame, make no sense if determinism is true. Compatibilists, on the other hand, want to insist that the causal well-orderedness of the universe is, precisely, compatible with our responsibility for our actions. But for most philosophers the question is not whether responsibility and causal well-orderedness are compatible, but how. In other words, to adapt Adkins’s adage, “we are all compatibilists now.”

The essential issue for any compatibilist position lies in the conception of responsibility it relies on – an issue much less well-explored by philosophers than the metaphysics of freedom and determinism. This article has contrasted three broad schools of thought on how we put responsibility into practice, by praising and blaming one another. When Adkins claimed that “we are all Kantians now,” he was not referring to Kant’s (incompatibilist) metaphysics but rather to our tendency to feel that responsibility attributions must have depth, that they reflect something about a person’s “real” deserts. Yet this position leads us to claims about control over the self, to the idea of choices that are really ours and not the result of any external influence. In other words, it is more difficult than it may seem to separate Kant’s position from his metaphysical account of freedom and the incompatibilism which he, above all other writers, so strongly articulated.

The roughly Aristotelian alternative discussed here has been most influentially articulated in Bernard Williams’s critique of modern accounts of morality, which he thinks are most clearly expressed in Kant’s philosophy. Williams argues that these ideas neither make sense on their own terms, nor do they make sense of what we actually do when we do engage in attributions of responsibility. As we have seen, Aristotle’s account of praise and blame is based on: (i) how far acts reveal character; (ii) the fair distribution of responsibilities to act; and (iii) the attempt to exchange reasons, share standards, and maintain relationships with those whom we judge – and who judge us in turn.

What both the Aristotelian and utilitarian accounts lack is the deep thirst for equality and fairness which motivate Kant. Aristotle’s account provides no equivalent to the Kantian will – some moral quantity which all human beings possess and which grounds the idea of their equal worth. Nor does it really satisfy the widespread sense that moral judgment should offer fairness – even though the world does not. There is a deeply appealing sense of fairness in Kant’s concern to do justice to each person’s will, by isolating some moral core to the person independent of all formative and environmental factors. Even if wicked people prosper and the innocent suffer, our moral judgment of each constitutes a deep and subtle form of compensation: with regard to what really matters, the one is lacking while the other is undiminished. Even if goodness is made much harder for some, and its results may be correspondingly less, nonetheless we should try to see past those externals, once more, to what really matters.

To this, the Aristotelian and the utilitarian alike may say: to treat praise and blame as reflecting such a pure form of desert is to lose touch with what really matters about them. Praise and blame help us live together in a world where ultimate deserts are impossible to make out, if they exist at all. But just because we cannot make out people’s “moral worth,” it is still true that we need to take responsibility – not least, in our openness to one another’s praise and blame.

  • Digg
  • Del.icio.us
  • StumbleUpon
  • Reddit
  • RSS
Read Comments

Bhedābheda Vedānta

Bhedābheda Vedānta is one of the several traditions of Vedānta philosophy in India. “Bhedābheda” is a Sanskrit word meaning “Difference and Non-Difference.” The characteristic position of all the different Bhedābheda Vedānta schools is that the individual self (jīvātman) is both different and not different from the ultimate reality known as Brahman. Bhedābheda reconciles the positions of two other major schools of Vedānta. The Advaita (Monist) Vedānta that claims the individual self is completely identical to Brahman, and the Dvaita (Dualist) Vedānta that teaches complete difference between the individual self and Brahman. However, each thinker within the Bhedābheda Vedānta tradition has his own particular understanding of the precise meanings of the philosophical terms “difference” and “non-difference.” Bhedābheda Vedāntic ideas can traced to some of the very oldest Vedāntic texts, including quite possibly Bādarāyaṇa’sBrahma Sūtra (app. 4th c. CE). Bhedābheda ideas also had an enormous influence on the devotional (bhakti) schools of India’s medieval period. Among medieval Bhedābheda thinkers are Vallabha (1479-1531 CE), founder of the Puṣṭimārga devotional sect now centered in Nathdwara, Rajasthan, and Caitanya (1485-1533 CE) the founder of the Gaudīya Vaiṣṇava sect based in the northeastern Indian state of West Bengal.

1. Historical Overview

Bhedābheda is often presented as a school of Vedānta. Vedānta, in turn, is sometimes spoken about as a single philosophy, when in reality “Vedānta” has different uses. Its most inclusive use is as a label for a philosophy that purports to be expressed at length in the latter part of the Vedas, that is, the Upaniṣads. It is centrally concerned with the inquiry into the nature of an ultimate entity called “Brahman.” There are many different accounts of this synoptic philosophy, and the various accounts are often considered schools of philosophy themselves. Unlike the well-known schools of Advaita (Non-Dualist) Vedānta, Viśiṣṭādvaita (Non-Dualism of the Qualified) Vedānta, and Dvaita (Dualist) Vedānta, it makes more sense to refer to Bhedābheda Vedānta as a “tradition” or “family” of philosophies rather than as a single “school.” This is because, unlike the three aforementioned schools, Bhedābheda has no single founder who created an institutionalized network of monasteries dedicated to the study, development, and propagation of the founder’s teachings. The history of Bhedābheda in India stretches back at least until the 7th century CE and likely quite earlier, and continues into the present day. Although there are substantial philosophical disagreements among the many Bhedābheda thinkers, their philosophies also show certain characteristic similarities. After a short historical introduction to several major Bhedābhedavādins, I will discuss a few viewpoints that almost all Bhedābheda schools share. These include the understanding of the relation between individual self (jīvātman) and Brahman as one of part and whole; the doctrine that the phenomenal world is a real transformation of Brahman (Pariṇāmavāda); and the doctrine that liberation can only be attained by means of a combination of knowledge and ritual action (Jñānakarmasamuccayavāda), not by knowledge alone.
a. Bādarāyaṇa and Bhartṛprapañca

Numerous scholars have concluded that Bādarāyaṇa’s Brahma Sūtra (circa 4th c. CE), one of the foundational texts common to all Vedānta schools, was written from a Bhedābheda Vedāntic viewpoint (Dasgupta 1922: vol. 2, p. 42; Nakamura 1989: p. 500). While that claim is disputed by other schools, there is little doubt that Bhedābheda predates Śaṅkara’s Advaita Vedānta. In his commentary on the Bṛhadāraṇyaka Upaniṣad, Śaṅkara (8th c.) repeatedly criticizes interpretations by an earlier Vedāntin named Bhartṛprapañca, who characterized the relation between Brahman and individual souls as one of difference and non-difference. One of the central disagreements between the two is that Śaṅkara claims that Brahman’s entire creation is a mere appearance (vivarta), while Bhartṛprapañca maintains that it is real (Hiriyanna 1957: vol. 2, pp. 6-16).
b. Bhāskara

The first Bhedābhedavādin widely recognized as such by the later tradition is Bhāskara (8th-9th c.). He was either a younger contemporary of Śaṅkara or perhaps lived slightly after Śaṅkara. His only extant work is a commentary on the Brahma Sūtra. That work is expressly written in order to defend the earlier claims of Bhedābhedavādins against Śaṅkara’s interpretation of the Brahma Sūtra. Although he never mentions Śaṅkara by name, he makes it clear from the beginning that his primary intention in commenting on the Brahma Sūtra is to oppose some predecessor: “I am writing a commentary on thissūtra in order to obstruct those commentators who have concealed its ideas and replaced them with their own” (Bhāskara 1903: p. 1). Bhāskara is the earliest in a long line of Vedāntic authors concerned to refute Advaita (including Rāmānuja and Madhva, not to mention numerous Bhedābhedavādins). Many of the stock arguments used against the Advaita originated with Bhāskara, if indeed he did not borrow them from an even earlier source. He also seems to have been remembered by the collective Advaita tradition as a thorn in its side. So, for instance, in the 14th century hagiography of Śaṅkara, theŚaṅkaradigvijaya, Mādhava depicts one “Bhaṭṭa Bhāskara” as a haughty and famous Bhedābhedavādin whom Śaṅkara defeats in a lengthy debate.
c. Yādavaprakāśa and Nimbārka

While the Viśiṣṭādvaita philosopher Rāmānuja (11th-12th c.) is widely acknowledged as the most influential Vedāntin after Śaṅkara, Rāmānuja’s complicated relationship with Bhedābhedavāda is rarely discussed. Rāmānuja’s teacher Yādavaprakāśa was a Bhedābhedavādin. Yādavaprakāśa’s works have been lost, and therefore almost all of what we know of his ideas comes from Rāmānuja and one of Rāmānuja’s commentators, Sudarśanasῡri. However, it is possible from these numerous hints to draw a sketch of Yādavaprakāśa’s basic views. Rāmānuja depicts Yādavaprakāśa as an exponent of Svābhāvika Bhedābhedavāda, the view that Brahman is both different and not different than the world in its very nature, and that difference is not simply due to difference of artificial limiting conditions (see Oberhammer 1997: p. 10). Yādavaprakāśa shares this basic viewpoint with Nimbārka (13th c.?), and disagrees with the Aupādhika Bhedābhedavāda of Bhāskara, who maintains that the difference of the world and Brahman is due to limiting conditions. Another characteristic of Yādavaprakāśa’s thought is his repeated insistence that Brahman has the substance of pure existence (sanmātradravya). The relationship between Brahman and the world is not merely one of class and individual, but rather both are existent entities, standing in the relationship of cause and effect (see Oberhammer 1997: p. 14).
d. Vallabha

In the late medieval period, the doctrine of Bhedābheda became increasingly associated with devotional (bhakti) movements in North India. It is largely on the basis of their reputations as the founders of religious sects, and not as philosophers per se, that thinkers such as Vallabha (1479-1531) and Caitanya (1485-1533) became widely known. Among the former’s most influential works are a commentary on the Brahma Sūtraentitled the Anubhāṣya, and his commentary on the Bhāgavata Purāṇa, entitled theSubodhinī. Vallabha founded the Vaiṣṇava sect of the Puṣṭimārga (“path of nourishment”) now based in Nathdwara, Rajasthan. His philosophical system, called Śuddhādvaita (Pure Non-Dualism), takes its name from his view that there is no dualism whatsoever between a real Brahman and an unreal world. Since both are completely real, he denies that there can be any sort of ontological dualism of real and unreal between the two—therefore it is a “pure” non-dualism. Obviously, this refers to the Advaita school’s view that the phenomenal world is not real in an ultimate sense, and is a clever attempt to re-appropriate the valued label “Advaita” for his own school. Yet in this regard all Bhedābhedavādins might claim the name Śuddhādvaita, since they all assert the reality of the phenomenal world.
e. Caitanya

Caitanya was another medieval Vaiṣṇava philosopher/theologian, famous for a school of thought known as Acintya Bhedābhedavāda (Inconceivable Difference and Non-difference). Although Caitanya never wrote down his teachings, numerous followers authored works based on his philosophy, such as Jīva Gosvāmin, author of a well-known commentary on the Bhāgavata Purāṇa. This system’s notion of “inconceivability” (acintyatva) is a central concept used to reconcile apparently contradictory notions, such as the simultaneous oneness and multiplicity of Brahman, or the difference and non-difference of God and his powers. The tradition of Acintya-Bhedābheda, also commonly known as Gaudīya Vaiṣṇavism, thrives to this day in the Indian state of West Bengal. Perhaps the most famous offshoot of this devotional tradition is the International Society for Krishna Consciousness (ISKON), more popularly known in the West by the name the “Hari Krishnas.”
f. Vijñānabhikṣu

The last major Bhedābheda thinker in pre-modern India, Vijñānabhikṣu (16th c.), did not follow the path of bhakti. Vijñānabhikṣu sought to show the ultimate unity of the schools of Vedānta, Sāṅkhya, Yoga, and Nyāya, and is most well known today for commentaries on Sāṅkhya and Yoga texts. In his innovative sub-commentary on Patanjali’s Yoga Sūtra, Vijñānabhikṣu argues that yoga is the most effective means to liberation, although he never repudiates the Bhedābheda metaphysical framework of his earliest writings (Nicholson 2005). Vijñānabhikṣu was a theist who considered Viṣṇu the supreme God. In his commentary on the Sāṅkhya Sūtra, he argues that the Sāṅkhya school requires an omnipotent God in order to cause the union of its two fundamental principles, primordial nature (prakṛti) and pure consciousness (puruṣa). Vijñānabhikṣu grounds his reinterpretations of fundamental concepts in Sāṅkhya-Yoga in Bhedābheda metaphysics. In his earliest works, such as his Bhedābheda Vedāntic commentary on the Brahma Sūtras, he understands the concepts of difference and non-difference in terms of separation and non-separation (Ram 1995). Although for him the fundamental relation of the individual self and Brahman is one of non-separation, the Sāṅkhya-Yoga analysis of the individual selves as multiple and separate from one another is correct, as long as it is understood that this state of separation is temporary and adventitious. While Vijñānabhikṣu’s acceptance of Sāṅkhya-Yoga philosophical truths puts him at odds with some earlier Bhedābhedavādins, he continues in the tradition of Bhāskara in his trenchant criticism of the Advaita Vedāntins, whom he decries as “crypto-Buddhists” (pracchannabauddha) and “Vedāntins in name alone” (vedāntibruva).
2. Ontology

One of the most notable differences between Bhedābheda Vedānta and Advaita (Monistic) Vedānta is their views on the existence of the phenomenal world. While Advaita holds that the phenomenal world is ultimately unreal (mithyā) and that only Brahman truly exists, Bhedābheda thinkers insist that the phenomenal world is real, and not at all illusory. In this basic assertion they are in line with the majority of Indian philosophical schools, including the schools of Qualified Non-Dualist (Viśiṣṭādvaita) Vedānta, Dualist (Dvaita) Vedānta, Nyāya, Sāṅkhya, and Mīmāṃsā. Although Advaitins cite certain passages from the Upaniṣads as supporting the notion that the world is akin to a mirage or a magical trick, Bhedābhedavādins accuse Advaitins of borrowing this idea from the Mind-only (Cittamātra) school of Buddhism, and frequently employ the epithet of “crypto-Buddhist” (pracchannabauddha) to refer to Advaitins.
a. Part and Whole

Bhedābhedavādins understand the relation between Brahman and the individual souls to be a relation between a whole and its parts. They frequently employ stock examples to illustrate this relation. Some of the most common are a fire and its sparks, the sun and its rays, a father and his son, and the ocean and its waves. Each of these is an example of a part-whole relation, which is also a variety of difference and non-difference (Bhedābheda). So, to take one example, the sparks that come off of a fire are both the same as that fire and different from it. They are the same insofar as they came from the fire, and are constituted by the same substance as fire. But they are also distinguishable from the original fire, as occupying a separate point in space. Although these four examples each seem to illustrate a different relation (and it may seem to make no sense at all to understand a son as a “part” of his father), Bhedābhedavādins cite these familiar examples from the physical world in order to shed light on the true metaphysical relation between Brahman and the individual selves. While each might capture some aspect of that relation, inevitably they are mere approximations, requiring further commentary and philosophical analysis.

Advaita Vedāntins object to the characterization of the individual self as a part, and characterize Brahman as partless. All schools of Vedānta understand the Veda as the ultimate epistemic authority, and arguments from scripture play a large part in intra-Vedāntic disputes. Advaitins point out that both the Upaniṣads and the Brahma Sūtras say that Brahman is partless (niravayava, niṣkala). Furthermore, the assertion that Brahman has parts seems to defy logic. It is inconceivable that Brahman could be made up of parts, for things that are made up of parts are dependent on those parts, and impermanent. Advaitins offer their own stock examples to show that Brahman cannot be divided up, and that any such division is purely an artificial limitation on an indivisible entity. For example, Advaitins commonly liken Brahman to the element called “space” (ākāśa). According to traditional science in India, space is an element that is omnipresent in the world, just as all Vedāntins agree that Brahman is omnipresent. Although we can talk about space as being delimited (the space inside a room, the space inside a pot), such limitations of space are purely accidental, not essential to the element itself. It may appear to an observer that the space inside a pot and the space outside the pot are two different entities, but this is a misunderstanding of the fundamental nature of space.

The Bhedābhedavādins can themselves appeal to textual authority for the idea that the relation between Brahman and the individual self is a relation between a whole and its parts. In Brahma Sūtra 2.3.43, The individual self is referred to as a “part” (aṁśa), and Bhedābhedavādins cite this passage whenever they require a textual support for their views. However, Advaitins take this description of the relation as a figurative, and not literal description of the status of the individual self. Otherwise, this passage will conflict with Brahma Sūtra 2.1.26, which says that Brahman is “partless” (niravayava). For Advaita, the world appears as if to be made of parts. But when it is understood correctly, all of the many entities in the world are seen to be false, and only one entity, a single, partless Brahman remains. Bhedābhedavādins, in their assertion of the world’s phenomenal reality, insist that multiplicity is real. Brahman is simultaneously one and many, depending on the perspective from which it is viewed, just as the ocean can be described as one or many, depending on the perspective from which it is described. Bhedābhedavādins maintain that Brahman’s being made up of parts in no way diminishes the perfection of Brahman, just as the existence of waves in the ocean in no way diminishes the amount of water therein.
b. Aupādhika and Svābhāvika Bhedābheda

All Bhedābhedavādins maintain the reality of the phenomenal world and the multiplicity of individual selves. However, some Bhedābheda thinkers edge closer to the Advaita position by arguing that although multiplicity is real, it is in some way less real than the absolute unity of Brahman. The early Bhedābheda thinker Bhāskara (8th-9th c. CE) exemplifies this tendency to reduce the ontological status of the phenomenal world, while still maintaining its reality. Bhāskara’s philosophy is an example of Aupādhika Bhedābhedavāda (“Difference and Non-difference Based on Limiting Conditions”). According to Bhāskara, the one, absolute Brahman becomes finite and multiple by means of limiting conditions (upādhis). Just as a pure diamond appears to be red when it is placed next to a red flower, so too the absolute Brahman appears finite when it is transformed through limiting conditions. This transformation is a real one; the individual self really is finite, subject to ignorance, suffering, and bondage, so long as it is filtered through limiting conditions. Although the individual self is real as differentiated from Brahman, for Bhāskara difference is merely a temporary state. In its natural state, Brahman is one and not many. Although it undergoes limitations to become finite, the ultimate goal of the individual is to realize his or her absolute state. Liberation is precisely the removal of such limiting conditions.

At the other end of the spectrum from Bhāskara’s Aupadhika Bhedābhedavāda is Nimbārka’s philosophy of Svābhāvika Bhedābheda (Natural Difference and Non-Difference). For Nimbārka (13th c.), Brahman is different and non-different not because of artificial limiting conditions, but contains both non-difference and non-difference as its essential nature. Along with Yādavaprakāśa (11th c.), Nimbārka comes closest to upholding both difference and non-difference as equally real states of Brahman. The tendency among most Bhedābhedavādins, however, is to subordinate difference to non-difference. Although difference is a real state that Brahman undergoes as it transforms into multiple individual selves, Bhāskara in essence makes the state of non-difference “more real” than the state of difference. In this way, school of Bhedābheda Vedānta is often closer to the school of Advaita (Monist) Vedānta than it is to the school of Dvaita (Dualist) Vedānta.
3. Causality
a. Pariṇāmavāda (Theory of real transformation)

Closely related to Bhedābheda Vedānta’s ontology is its theory of causality. Bhedābheda Vedāntins subscribe to the theory of Pariṇāmavāda, which states that the phenomenal world is a real transformation (pariṇāma) of the material cause of the world. They share this theory with the Sāṅkhya school of philosophy, as well as with most other schools of Vedānta. The major difference between the Vedāntic theory of Pariṇāmavāda and the Sāṅkhya’s Pariṇāmavāda is the understanding of what constitutes the material cause of the world. For Sāṅkhya, primordial nature (prakṛti) transforms itself into the phenomenal world. The principle of primordial nature is completely insentient, and the process of transformation that creates the world is a blind, automatic process. For Bhedābheda Vedāntins, Brahman is both the material and efficient cause of the universe. Brahman, unlike the Sāṅkhya’s prakṛti, is sentient. Yet both the sentient (individual souls) and insentient (physical things) have their origin in Brahman, according to Bhedābhedavādins. In spite of their apparent proximity to the Sāṅkhya school on the issue of causality, early Bhedābheda thinkers such as Bhāskara took pains to critique the Sāṅkhya notion ofprakṛti, accusing it of being both contrary to scripture and contrary to logic. A few later Bhedābheda thinkers took a softer line on Sāṅkhya. The most notable of these was Vijñānabhikṣu (16th c.), who argued for the ultimate unity of Sāṅkhya and Bhedābheda Vedānta doctrines.
b. Vivartavāda (Theory of unreal manifestation)

Once again, it is useful to contrast the doctrines of the Bhedābheda school with Advaita Vedānta. Advaita Vedānta maintains the doctrine of Vivartavāda, which states that the world is an unreal manifestation (vivarta) of Brahman. Advaita, like other schools of Vedānta, identifies Brahman as both material and efficient cause. But for the Advaitins, Brahman is the cause of an unreal effect. Although the world can be described as conventionally real (paramārthasat), the Advaitins claim that all of Brahman’s effects must ultimately be acknowledged as unreal before the individual self can be liberated. Although the theory of Vivartavāda is traditionally accepted as a theory shared by the entire Advaita school, some recent historians have questioned this, noting passages in the work of Śaṅkara, the founder of the Advaita, that appear to be closer to the theory ofpariṇāma (Hacker 1953: pp. 24ff.; Rao 1996: pp. 265ff.). It is likely that the theory ofVivartavāda is a theory that emerged gradually out of the earlier Vedāntic theory ofPariṇāmavāda, rather than one that sprang fully formed out of the head of Śaṅkara. It also bears repeating that some Bhedābheda Vedāntins come perilously close to the Advaita view of the phenomenal world as only conventionally real, as they often emphasize that multiplicity is an unnatural, temporary state.
c. Satkāryavāda (Theory of pre-existent effect)

Proponents of both Vedāntic theories of causality, Pariṇāmavāda and Vivartavāda, justify each by citing a central passage at Chāndogya Upaniṣad 6.1.4-5. There, the sage Aruni describes the nature of causality to his son, Śvetaketu, using the example of the relation of clay to a pot:

It is like this, son. By means of just one lump of clay one would perceive everything made of clay—the transformation is a verbal handle, a name—while the reality is just this: ‘It’s clay.’It is like this, son. By means of just one copper trinket one would perceive everything made of copper—the transformation is a verbal handle, a name—while the reality is just this: ‘It’s copper.’ (Olivelle 1996: 148)

This passage uses the examples of an everyday material cause, clay or copper, to shed light on the nature of cause and effect. It expresses the doctrine of Satkaryavāda, which says that the effect preexists in its cause. All Vedāntins subscribe to this theory—the doctrines of real tranformation (pariṇāma) and unreal manifestation (Vivartavāda) can be understood as two different versions of the theory of Satkāryavāda. According toSatkāryavāda, the lump of clay does not go out of existence when it is transformed into a pot, a cup, a saucer, or the like, only to be replaced by something entirely new. Although the form of the clay has changed, its essence, its clay-ness, remains. The same logic applies to everything caused by Brahman. The entire world, in all of its many forms, nonetheless shares the same essence, as being Brahman. This view, something like an early Indian theory of the conservation of matter, suggests that nothing ever arrives in the universe completely new, but only as a transformation of some earlier material cause. Nothing can be created ex nihilo. In this belief, the Vedāntins are at odds with the Buddhist and Nyāya schools, who for separate reasons argue that the effect does not preexist in the cause.

Although Bhedābheda Vedāntins and Advaita Vedāntins have the theory of Satkāryavādain common, they part ways when asked to characterize the status of the effect. Is the effect a real transformation (pariṇāma) of the cause, or merely an unreal manifestation (vivarta)? The passage at Chāndogya Upaniṣad 6.1.4-5 has been interpreted in both ways. Advaitins emphasize the apparent nominalism expressed by Aruni in this passage: “the transformation is a verbal handle, a name—while the reality is just this: ‘It’s copper.’” This might suggest that the effect is unreal, and only the cause is truly real. But Bhedābhedavādins see this passage as simply another instantiation of the principle of difference and non-difference, just as it is illustrated by the examples of a fire and its sparks or the sun and its rays. From one perspective, focusing on substance, we can say that all of the various cups, saucers, and plates are one—they are all clay. Yet at the same time, they have been transformed by the pot-maker into different forms, multiple in number, occupying different points in space. From this perspective, the effects are real. Just as the many pots, plates, and saucers are simultaneously different and non-different from the original lump of clay, so too all of the individual selves are both different and non-different from Brahman, the original material cause.
4. Theology and soteriology
a. God in Bhedābhedavāda

In the medieval period, Bhedābheda Vedānta became closely associated with theism in general, and the movement of bhakti devotionalism in particular. There is a reason thatbhaktas such as Vallabha (1479-1531 CE) and Caitanya (1485-1533) built the foundations of their theological systems on centuries-old Bhedābheda concepts. Like the schools of Rāmānuja and Madhva, Bhedābhedavāda is a realist school. Whereas in the Advaita school even God has to be understood as ultimately unreal, since He too is merely Brahman limited by the artificial condition of lordliness, certain types of Bhedābheda philosophy can accommodate a God who is real in his qualified (saguṇa) form. Although on a certain level, an Advaitin can profess a belief in God, he or she knows that ultimately God is merely a crutch, a heuristic to enable human beings to go one step closer to that ultimate Brahman devoid of qualities (nirguṇa). Such a God is ultimately unsatisfying for those whose primary interest is devotion—in any system of Advaita, devotion must occupy a lower position than pure knowledge. On the other hand, many worshippers will also be unsatisfied with the Dvaita school’s uncompromising notion that they themselves are completely separate from God, and that ultimate unification with the Godhead is impossible. Both Bhedābhedavāda and Viśiṣṭādvaita offer the possibility to bridge these two alternatives, by offering the alternative of both a real God possessing qualities and the possibility of personal participation in that Godhead.
b. Knowledge combined with ritual acts leads to liberation

Besides insistence that the phenomenal world is a real transformation (pariṇāma) of Brahman, another view shared by Bhedābhedavādins is the necessity of ritual acts in combination with knowledge (Jñānakarmasamuccayavāda) in order to obtain liberation. Bhāskara devotes much of the beginning of his commentary on the Brahma Sūtra to a critique of Śaṅkara’s radical view that knowledge alone is sufficient for the attainment of Brahman, as long as one has fulfilled one’s ritual requirements at an earlier stage. Although today polemics between Vedāntins are usually depicted in solely philosophical or theological terms, this suggests that above all, Śaṅkara’s new teachings were seen by other Vedāntins in the 8th century as a serious threat to the ritual-social order. Bhāskara’s arguments in favor of Brahmanic ritualism are an important reminder of the continuities between early Vedānta and the Pūrva Mīmāṃsā (Prior Exegesis) school of ritual hermeneutics. The two schools are so close that Sanskrit authors in pre-modern India typically refer to Vedānta by the names “Brahma Mīmāṃsā” (Exegesis of Brahman) or “Uttara Mīmāṃsā” (Later Exegesis), emphasizing the central importance of Vedic interpretation for all Vedāntic thinkers.

The notion of bhakti finds a home in Bhedābhedavāda, since Bhedābheda takes activity in the world (karman) seriously, believing that activities in the world are real, and produce real effects. But it should not be thought that all Bhedābhedavādins were proponents ofbhakti. The early Bhedābheda of Bhāskara was not concerned at all with bhakti. Instead, Bhāskara uses Bhedābheda conceptual terminology as a conservative apologist, to defend the importance of Brahmanical ritual orthodoxy from Śaṅkara, a radical who rejected the ultimate efficacy of Vedic ritual. It is only with Nimbārka, a Bhedābheda thinker heavily influenced by the bhakti system of Rāmānuja, that we first fully see the union of bhaktiworship and Bhedābhedavāda. Even in medieval northern India, where bhakti was influential and widespread, not all Bhedābhedavādins were bhaktas. Vijñānabhikṣu, for example, was more interested in espousing a modified, Bhedābheda Vedāntic form of Patañjali’s Yoga than he was to proselytize for the path of devotion. Such flexibility of the Bhedābheda philosophical apparatus has allowed it to survive as a living tradition for over 1500 years in a number of very different historical contexts. Although in the modern period Bhedābheda Vedānta has been eclipsed in popularity by neo-Vedāntic interpretations of Advaita Vedānta philosophy, its lineage continues today among traditional scholars in Puṣṭimārga and Gaudīya Vaiṣṇava religious communities. And, for the first time in its long history, in the early 21st century Bhedābheda Vedānta is beginning to receive the attention it deserves among historians, philosophers, and theologians outside of India.

  • Digg
  • Del.icio.us
  • StumbleUpon
  • Reddit
  • RSS
Read Comments

Bhartrihari (c.450—510 CE)

Bhartrihari may be considered one of the most original philosophers of language and religion in ancient India. He is known primarily as a grammarian, but his works have great philosophical significance, especially with regard to the connections they posit between grammar, logic, semantics, and ontology. His thought may be characterized as part of the shabdadvaita (word monistic) school of thought, which asserts that cognition and language at an ultimate level are ontologically identical concepts that refer to one supreme reality, Brahman. Bhartrihari interprets the notion of the originary word (shabda) as transcending the bounds of spoken and written language and meaning. Understood as shabda tattva-the “word principle,” this complex idea explains the nature of consciousness, the awareness of all forms of phenomenal appearances, and posits an identity obtains between these, which is none other than Brahman. It is thus language as a fundamentally ontological principle that accounts for how we are able to conceptualize and communicate the awareness of objects. The metaphysical notion of shabda Brahman posits the unity of all existence as the foundation for all linguistically designated individual phenomena.

1. Bhartrihari’s Life and Works

Bhartrihari’s works were so widely known that even the Chinese traveler Yijing (I-Tsing) (635-713 CE) mentions the grammarian-philosopher, mistaking him for a Buddhist. Unfortunately, we do not know much about his personal history and his works do not throw much light on the matter. There are some narratives referring to his background but they are not supported by historical data. In these somewhat dubious accounts, he is said to have been existentially torn between two kinds of life: the path of pleasure and that of the monastic yogi. Although he believed that he should renounce the world of material pleasures (reflected in poetry attributed to him by scholars), it took many attempts to finally achieve the life of dispassion. His hedonism and philosophical acumen led him, according to his legend, to produce works of great breadth, depth and beauty.

Bhartrihari credits some of the theories in his work Vâkyapadîya to his teacher, who was probably one of Candrâcârya’s contemporaries, Vasurata. To be more precise, the noted scholar T.R.V. Murti proposes the following chronology: Vasurata, followed by Bhartrihari (450-510 CE) and Dinnâga (Dignâga) (480-540 CE). Among the major works attributed to Bhartrihari are his main philosophical treatise, Vâkyapadîya (On Sentences and Words) kândas I, II, and III, Mahâbhâshyatîkâ (a commentary on the Mahâbhâshya of Patanjali), Vâkyapadîyavrtti (a commentary on the Vâkyapadîya kândas I and II), and shabdadhâtusamîksha. Since 1884, the Vâkyapadîya, containing approximately 635 verses, has been edited and published several times in English translation.

The first two chapters of the Vâkyapadîya discuss the nature of creation, the relationship of Brahman, world, language, the individual soul (jîva), and the manifestation and comprehension of the meanings of words and sentences. In addition, the literary works attributed by some to Bhartrihari (not mentioned here) have made an impact on the growing popular Hindu devotional (bhakti) movements. More importantly, his philosophical work was recognized and addressed by schools of Hindu scriptural exegesis (Mîmâmsâ), Vedânta (mystical Vedism) and Buddhism.
2. Early Grammarians and Philosophical Semantics

In ancient India, grammarians saw their task as establishing the foundations of the Vedas, but their work often resulted in the development of their own philosophical systems. Patanjali, in his Mahâbhâshya, explains that the study of grammar (vyâkaranam) was meant to maintain the truth of the Vedas, to guide the use of Vedic speech in ritual contexts, and to aid in the clear interpretations of individual human speech. Both Pânini and Patanjali, two major Sanksrit grammarians, were the first to provide a systematic and formal analysis of the grammatical bases of all intended meanings. Pânini (7th century BCE) developed the Ashtâdhyâyî (Eight-Chapters) for the grammarians. This impressive work contains a thorough analysis of the rules of Sanskrit language down into its nominal and verbal components; it contains a science of language, applicable to the Vedas, also comprised of sets of operational rules and meta -rules that interpret the former. Among these “rules for interpretation” of Vedic texts, we are given a “universal grammar. Pânini’s approach is not like the Mîmâmsâ, which focuses on the study of Vedic language. Instead, Pânini deals with spoken and Vedic languages as if they are of the same genre.

Pânini’s Ashtâdhyâyî, its commentaries, and the Vâkyapadîya of Bhartrihari are said to constitute the fundamental texts for the school of Pânini’s grammar, whose object of study was ultimately Vedic. Around 150 BCE, Patanjali wrote the Mahâbhâshya, an interpretation of some of Pânini’s rules written in dialogue form, and it is this work that is the basis for later commentaries on grammar and philosophy. It is of interest to note here that the Dharmashâstras or Treatises on Law, including the well-known Laws of Manu, were composed between 322 and 183 BCE. J.N. Mohanty points out that these treatises can be seen as attempts on the part of orthodox Brahminism to preserve itself against the anti-Vedic philosophies. However, he considers Pânini’s grammar and Patanjali’s commentary to carry greater weight in the Indian philosophical tradition.

With the Vâkyapadîya, Bhartrihari moves grammatical analysis squarely into the realm of philosophy, arguing that grammar can be consider a darshana, a “view,” or an official philosophical school, providing perspective and insight into ultimate reality. The first verse articulates the fundamental view of his newly envisaged school:

The Brahman is without beginning and end, whose essence is the Word, who is the cause of the manifested phonemes, who appears as the objects, from whom the creation of the world proceeds.

It is the project of the Vâkyapadîya to explain this verse, with all of its philosophical, linguistic, and metaphysical implications. At base, we contextualize Bhartrihari’s philosophical inquiry into language as being conditioned by the Indian culture and scriptural tradition, in which this type of intellectual pursuit had a soteriological purpose -the realization of absolute knowledge and the spiritual liberation which ensues; thus, it is a distinctively ontological reflection on language which Bhartrihari added to the thought of earlier grammarians.
3. Brahman, Language, and the World

The Brahminic view of the cosmos put forth in the Vedas is one of constant and cyclical creation and dissolution. At the dissolution of each creative cycle a seed or trace (samskâra) is left behind out of which the next cycle arises. What is significant here is that the nature of the seed from which each cycle of creation bursts forth is expressed as “Divine Word” (Daivi Vâk). If language is of divine origin, it can be conceived as Being Brahman expressing and embodying itself in the plurality of phenomena that is creation.

Bhartrihari considers Brahman, the basis of reality, to be “without beginning and end” (anâdi nidhânam), as a concept that is not subject to the attributes of temporal sequences of events, either externally or in the succession of mental events that form cognitions. The word principle, shabda Brahman, is not defined in terms of the temporal nature of our cognitive states, because it functions as the inherent, primordial ground of all cognitions. Thus, against the Hindu logicians, the Nyâyas, for whom particular forms of human speech may be expressed in conventional terms for practical purposes, language itself is not something which arises or is created in time by God or humans. As B.K. Matilal states, “To talk of an absolute beginning of language is untenable. Language is continuous and co-terminus with human existence or the existence of any sentient being.”

There has been some scholarly debate regarding the meaning of the term “eternal” or “akshara” as Bhartrihari applies it to the word-principle. While some interpret this to refer to an all-pervading entity, existing in opposition to the multiplicity of objects in space and time, others see it as Bhartrihari’s specialized way of referring to phonemes, the minimal units of meaningful sound. It seems that phonemes understood in this way explain how it is the case that Word appears as objects. Eternity is “that which appears as objects, and from whom the creation of the world proceeds.” Phonemes are thus the eternally possible elements that can be combined in inexhaustible ways to manifest the plurality of nature.

This principle accounts for creation on a number of levels: it is the origin of consciousness, of cognition, sensation, language use, cognitive and experiential aspects of the world. In other words, objects of thought and the relations between them are word-determined, regardless of whether they are objects of perception, inference or any other kind of knowledge. When we perceptually apprehend external reality, we always do so in terms of names, for without names objects are neither identifiable nor knowable.

Furthermore, when we consider phenomenal concepts, we see that they do not exist or hold any meaning aside from the words through which they are expressed; we might say that our concepts are “word-loaded” and from this we can infer that the word principle causes the world. Bhartrihari’s causal claim is in keeping with the traditional philosophical discussions on the nature of causality and inference as he applies it to the word-principle:

Just as other thinkers, while explaining causality, saw that the properties of cause continue in the effects….in the same way in the scriptures also, the word in which the power of Enjoyer and Enjoyed are submerged has been declared as the cause of the world.

4. Bhartrihari’s Grammar

In the Vâkyapadîya, kânda I, Bhartrihari defines the scope of his inquiry as the subjects of grammar. Our speech takes the form of the basic structures of language, and grammar deals with this communicatively spoken language. The correct understanding of speech can take us to the limits of our conventional and spiritual capacities, and so language analysis must operate at all the following levels: 1. sentences and words, 2. meanings corresponding to sentences and words, 3. the fitness or compatibility between sound and sense, and 4. the spiritual merit obtained by using the correct language.

In the Sanskrit grammatical tradition, the “elite” are defined as those who use the correct language; we arrive at this standard language by abstracting from communicative language, or “language-in-use.” In his linguistic theory, Bhartrihari distinguishes between two forms of language, the spoken, or “language-in-use” and the analytic. The analytic or formal language emerges from a formal, abstract analysis of communicative language. If we were to gather and compare various sentences and words from different contexts of use, we would logically infer the basic segments (roots, stems, suffixes) that account for a common logical or formal basis of denotation.

This hierarchical conception of language use and language meaning can be understood in the following way, taking off from a representation of Matilal, with the term on the far right of each column understood as the originator of the term in the middle, and the term in the middle being the originator of the term on the left. In other words, Bhartrihari’s conception of utterance and understanding can be grasped with the following schema under the rubric of:
Product Producer Derivative Element
Linguistic Components Language-in-Use Analytic Language
sound
sentences and words word stems, suffixes, etc..
sense sentence meaning and word meaning stem meaning and suffix meaning
sound and sense relations fitness compatibility causality relations
purpose spiritual merit correct knowledge

There is debate about the ontological and epistemological status of relations between these levels of language, and Bhartrihari’s commentary on grammar includes a review of several theories and ultimately he seems to favor the “naturalist view.” In the first chapter of the Vâkyapadîya, Bhartrihari explains the naturalist view. Following the pâdavâdins (those who regard the word as the primary indivisible unit) who consider word-constituents, such as roots and suffixes, to be mere fictitious abstractions from words, so also the vâkyavâdins (those who regard the sentence as the indivisible unit) consider words to be imaginary abstractions from the sentence. The naturalists, such as Pânini, believe that language has an invariant form expressed in grammar. They therefore give epistemic primacy to spoken language; formal language is only an “appearance” and secondary aid to understanding. The conventionalists, on the other hand, hold that the analytic language is primary in that it contains within it all the structural features that may be used to create meaningful speech.
5. The Sphota Theory of Language

Bhartrihari’s theory occupies an interesting place in the ongoing Hindu-Buddhist debates about meaning and reference. For the Buddhists, meaning is a function of social and linguistic convention and reference is ultimately a projection of imaginative consciousness. For the Brahminic Nyâyas or Logicians, words have meaning because they refer to external objects; words can be combined in sentences just like things exist in relation to one another in external reality. With Advaita Vedânta, words mask the meaning of the Absolute Self (Âtman) which is Brahman, so that, when a person predicates categories to their identity such as in the sentence “I am tall,” this predication masks the all-inclusive nature of the eternal Self, which is beyond categorization. Bhartrihari puts forth a theory of language which, rather than starting by taking fundamental ontological, epistemological or social sides in these well-established debates, starts from the question of how meaning happens, how it emerges from the acts of both speaker and audience, and, constructing this theory first, what he believes to be appropriate metaphysical, epistemological and soteriological implications are drawn from it.

For Bhartrihari, linguistic meaning cannot be conveyed or accounted for by the physical utterance and perception of sounds, so he puts forth the sphota theory: the theory which posits the meaning-unit, which for him is the sentence, as a single entity. The term “sphota” dates back to Pânini’s reference to “sphotâyana” in his treatise Ashtâdhyâyî, however it was Patanjali who explicitly discusses sphota in his Mahâbhâshya. According to him sphota signifies spoken language, with the audible sound (dhvani) as its special quality. In Bhartrihari’s treatment of this concept, while the audible noise may vary depending on the speaker’s mode of utterance, sphota as the meaning unit of speech is not subject to such variations. This is so because for Bhartrihari, meaning is conveyed by the sentence. To explicate this theory, Bhartrihari depends on the root of sphota, namely sphut, meaning “to burst forth…” as in the “idea that spews forth” (in an internal mental state) when a meaningful sound, the sentence as a whole, is uttered.

The meaning of the sentence, the speech-unit, is one entire cognitive content (samvit). The sentence is indivisible (akhanda) and owes its cognitive value to the meaning-whole. Thus, its meaning is not reducible to its parts, the individual words which are distinguished only for the purposes of convention or expression. The differentiated word-meanings, which are also ontological categories, are the abstracted “pieces” we produce using imaginative construction, or vikalpa. Sphota entails a kind of mental perception which is described as a moment of recognition, an instantaneous flash (pratibhâ), whereby the hearer is made conscious, through hearing sounds, of the latent meaning unit already present in his consciousness (unconscious). The sentence employs analyzable units to express its meaning, but that meaning emerges out of the particular concatenation of those units, not because those units are meaningful in themselves. We analyze language by splitting it up into words, prefixes, suffixes, etc….but this is indicative of the fact that we “misunderstand” the fundamental oneness of the speech-unit. Words are only abstracted meaning possibilities in this sense, whereas the uttered sentence is the realization of a meaning-whole irreducible to those parts in themselves. This fundamental unity seems to apply, also, to any language taken as a whole. Matilal explains: “it is only those who do not know the language thoroughly who analyze it into words, in order to get a connected meaning.” As this scholar suggests, it is rather remarkable that Bhartrihari’s recognition of the theoretical indivisibility of the sentence resonates with the contemporary linguistic view of learning sentences as wholes (at a later stage of development we build new sentences from learned first sentences through analogical reasoning).

Sphota is therefore the cause of manifested language, which is meant to convey meaning. Sphota is more specifically identified as the underlying totality of linguistic capability, or “potency” and secondarily as the cause of two differentiated aspects of manifested meaning: applied meaning expressed as dhvani, the audible sound patterns of speech and artha-language as meaning-bearing. The grammatical/syntactical parts of the underlying sphota can only be heard and understood through its phonetic elements. Bhartrihari explains that the apparent difference between sphota and dhvani arises as we utter words. Initially, the word exists in the mind of the speaker as a unity but is manifested as a sequence of different sounds-thus giving the appearance of differentiation. dhvanis may be more specifically described as merely the audible possibility of meaning, a necessary but hardly sufficient condition of meaning.

We might think of this unit of linguistic potency, the sphota, as the cognitive/propositional whole content of meaning that can be transposed into different languages, while the actual word-sounds comprise the contents of the “speech-act”. But what holds the act to its ability to convey intended meanings? The words sounded by a plurality of speakers comprise the physical manifestation of vâk or vaikharî-vâk and it is upon this form of vâk that physical objects as objective forms are modeled. The unity that underlies these objective referents and meanings, however, is known as the intuited vâk-pashyati-vâk, which makes possible the unmediated understanding of a complete linguistic expression. This intuitive level of understanding, constitutive of the sphota, is teleological in its nature and structure in that it contains all potential possibilities of meaning-bearing dhvanis and their order of manifestation.

But, what guarantees that the hearer of speech properly comprehends what is uttered? In the second book of the Vâkyapadîya, Bhartrihari states:

Sentence meaning is produced by word meanings but is not constituted by them. Its form is that Intuition, that innate “know how” awareness (pratibhâ) possessed by all beings. It is a cognitive state evident to the hearer…not describable or definable, but all practical activities depend on it directly or through recollection of it.

Pratibhâ intuition can be characterized as shabda, the very same speech principle externalized in the utterances of speakers, as it operates within the hearer, causing her to instantaneously comprehend the meaning of the utterance. However, linguistic convention, shared by speaker and hearer, cannot account for the flash of comprehension. If that were the case, we would not have instances where communication breaks down in spite of the shared language between speaker and hearer. The comprehension of meaning lies in the sphota that is already present in the hearer’s awareness. As she hears the succession of audible phonemes, the latent and undifferentiated language potency within her is brought to “fruition” in the form of grasping the speaker’s meaning. Thus, while the audible words are necessary for such verbal comprehension to occur in the hearer, they are not sufficient. It is her own ability to understand meaning referred to by these words, by virtue of sharing the same sphota with the speaker, which completes the act of cognition.

It is at this point that the philosophy of language has for Bhartrihari religious implications of both ontological and interpretive scope. Just as various sentences might sound different in the mouths of different speakers and yet convey the same meanings, various Vedas may seem different in form and style, but there is a unity carried by the underlying sphota, which ensures that it is the same truth, or dharma that is expressed throughout the texts. Bearing in mind that Brahman is the ultimate referent of all speech forms, this higher reality is manifested in the sacred texts-whose efficacy (ritual, soteriological, epistemological) depends upon our ability to correctly apprehend its meaning. The sphota concept makes such interpretation possible. Again, the sphota expresses a meaning-whole behind individual letters and words. The implication for the truth of Vedic discourse is clear, for that truth is already present in the speaker (the Veda) and is potentially present in the consciousness of the hearers (the practitioners).

According to Bhartrihari’s theory, we can justify this particular philosophical method as revelatory by using the concept of shabdapramâna. The implications of this method are explained in the following section; here, we examine the source of our cognitions. But in order for one to give their assent to a worldview that renders to language the cosmic and salvific roles Bhartrihari does, a theory that posits that language is the medium of ultimate knowledge, one must be convinced that language in general has the capacity to yield ordinary knowledge. Given the way Bhartrihari conceptualizes language, as not primarily referent directed, but instead as referent-constructing, we need to look at how the grammarian thematizes the knowledge-conferring power of language within his own peculiarly unique framework.
6. Phenomenology of Language and the Concept of Shabda-Brahman

Sphota may be characterized as the intersubjective, universal “store-house” of meaning, the ground of all linguistic activity and communication. Sphota is the unifying principle that connects the word, the grammatical form of the word, and the meaning. Furthermore, just as words and sentences represent “pieces” of the meaning extracted from the whole, the objects and states of affairs these pieces represent actually refer to a “whole of objects meant” or an entire reality.

In classical Indian thought, objects are thought to be constituted of substance (dravya), but in Bhartrihari and especially in his first major commentator Helârâja, substance can be distinguished into two kinds, the substance of all things, which is Brahman, and the other individual, empirical substances. The empirical notion of substance here may be derived from the grammatical operation of ekashesha, explained by Pânini as using individual word-tokens to refer to individual objects-substances. Thus, names or singular terms are said by the earliest grammarians to refer to one substance at a time, therefore substance is defined through the relation of reference, and the nature of each substance is so specific that we cannot posit any general properties possessed by all of them. For example, each time we say the word ‘cow’ we refer to a different cow, and each cow is actually a different wholly individual entity.

Bhartrihari defines “actual” or empirical substance as that which we refer to by using indexicals and quantifiers, which refer to anything in our ontological reality: ‘this’ ‘that’ ‘something’ or ‘anything.’ The term ‘this’ points out an existence given to perception, while ‘that’ refers to something whose existence can be validated by some other means of knowledge but which is not available to perception. Bhartrihari also acknowledges the pragmatic and common sense view of “substance” as “a relative concept being dependent on our concept of quality (guna). A substance is that which is said to be distinguished and a quality is that which distinguishes the substance.

But Bhartrihari’s contribution to this debate changes the very notion of substance into a much more inclusive and general concept, since anything we refer to using a name or substantive term, even generic properties and verbs, become substances in that they are distinguished by words, as Matilal illustrates: “Thus, cooking would refer to the fact of cooking and ‘walking’ to the fact of walking as long as the speaker intends to distinguish the act of cooking from the act of walking.” “In the third book of the Vâkyapadîya, he defines the concept of ‘quality’/guna as dependent upon, as arising from substance. He rejects the Vaisheshika view that substances and qualities belong to entirely different categories (padârtha-s), and espouses the revolutionary view that the latter arises from the former. For him, qualities, existing in relation to substances serve to further differentiate those substances by “delimiting their scope.” But how does he account for such a radical revisioning?

Bhartrihari’s contribution of his particular theory of the “imaginative construction” of perceptions and language once again emerges within the context of debates with competing theories of knowledge. The Buddhist idealistic claim also argued that the world of experience or phenomena is at base a product of the human imaginative faculty. The Buddhists claim that our cognitive experiences construct our reality; these are modes of consciousness containing cognitive contents and in the final analysis, do not yield any knowledge about reality as it may be outside of themselves. It is consciousness that posits the (apparent) externality of objects, not the “objects themselves.” This form of phenomenal-idealism is developed as a counterclaim to the Hindu realist position, which affirms the existence of external reality. For the Buddhist, objects are only the external contents of the human imagination. Interestingly enough, Bhartrihari’s sphota theory of language and cognition is sometimes understood as an extension of the Buddhist position; according to the grammarian, cognition is entirely language-dependent in that the structure of our cognitive states is determined by grammar. But Bhartrihari’s theory posits knowledge as a matter of specifically linguistic construction. The concept of vikalpa for him implies the following: the structure of language shapes how we categorize the objects of our experience and our descriptions of reality as a whole. Even at the most immediate levels of awareness), we must conceptualize and therefore interpret the contents of sense perception. Thus, at the level of pure sensation, the sensory core is already saturated, as it were, with the “deep structure” of language. In this respect, Bhartrihari’s position differs from the Buddhist position rather dramatically. The Buddhists clearly distinguish between pure perception (nirvikalpa-pratyaksha), which is pre-conceptual, unverbalizable and correspondent to reality, and constructed perception (savikalpa-pratyakasha) that is conceptual and may therefore be considered a verbalized interpretation of the real. For the Buddhist, the pure sensory core is the real locus of perception. Bhartrihari, as an ontological monist, does not distinguish between a pure perception and a constructed perception such that the former is concept-free and ineffable and the latter concept-loaded and autonomously constructed, because he thinks that perception is inherently verbal. Not only are sense data and linguistic units non-different, but they are expressive of the unitary principle of Brahman-which is differentiated into the plurality of linguistic objects that make up the world.

Bhartrihari’s notion of vikalpa is also directed against the early Nyaiyayikas, who, while agreeing on the correspondence between word and thing, uphold the distinction between language and its object-referents. These Hindu Logicians held that the perceptual apprehension of the object could be distinguished from naming the object. For the Nyâyas, who are ontological pluralists and materialists, words refer to distinct generic properties of and relations between objects. Perception is a two-step process involving the initial apprehension of the object and then the subsequent apperception/awareness that results in mental and syntactic/linguistic representations of the first moment of awareness. Here, linguistic categories originate in the different substances and attributes that exist in the world. Bhartrihari counters them by arguing that the act of perception, rather than acquiring linguistic clothing after the bare particular has already been presented to consciousness, can only be aware of the object before it as a ‘this’ or ‘that’, that is, as an awareness of something only as a particular and as such, identifiable. That is to say, significantly enough, that for Bhartrihari, the word makes the thing an individual, and as one moves further and further along the refined categories of what is conventionally known as denotation, the word makes the thing what it is. For Bhartrihari, the difference the Logicians posit between the ontological and the linguistic would make meanings of all kinds, mundane ones and religious ones, contingent on the circumstances and speaker. But if perception is innately verbal, no perilous bridge need be suspended over some supposed abyss between vision and truth, both in our mundane lives and for the rishis who pronounced the Vedas. The word then makes the thing, and Brahman makes the world, and so it is entirely proper to speak of words as the creator of all things (shabda-Brahman).
7. Bhartrihari and Western Philosophy

Although previous Bhartrihari scholarship has progressed rather slowly due to numerous difficulties, within the last decade or so his work has garnered attention from Western scholars. Bhartrihari’s explorations into the relations between language, thought and reality reflect contemporary philosophical concerns with meaning, language use, and communication, particularly in the work of Chomsky, Wittgenstein, Grice, and Austin. His theory of language recognizes that meaning is conveyed in formalist terms where meaning is organized along syntactical rules. But it makes the leap, not made by modern Western philosophers, that such a view of language does not merely serve our mundane communicative purposes and see to the achievement of practical goals, but leads to paramount metaphysical knowledge, a knowledge carrying with it a palpable salvific value.

  • Digg
  • Del.icio.us
  • StumbleUpon
  • Reddit
  • RSS
Read Comments