Maverick Philosopher

One idea worth exploring is that there is room for supererogatory acts in Kant's scheme under the rubric of imperfect duties. Kant tells us that ". . . by a perfect duty I here understand a duty which permits no exception in the interest of inclination . . . ." (Grundlegung zur Metaphysik der Sitten, AA 421, fn) A perfect duty can be either towards oneself or towards another. Kant gives the prohibition of suicide as a perfect duty to oneself and the prohibition of deceitful promises as a perfect duty towards others. These are proscriptions that admit of no exception, and I take it that perfect duties are perfect in that they are excepttionless: in no circumstance may one take one's own life, and in no circumstance may one make a deceitful promise. If so, then imperfect duties are prescriptions that admit of exceptions. This interpretation fits the examples Kant gives (423-424). There is the duty to oneself of cultivating one's talents, and the duty to to others of benevolent assistance. Both of these duties are prescriptions and both admit of exceptions. My general duty to cultivate my talents is not a duty to cultivate all my talents, or even any particular talent. And the duty of benevolent assistance cannot be a duty to assist everyone. Paul Guyer's reading is similar:

In the Groundwork, Kant's principle of morality gives rise to a fourfold classification of duties, resulting from the intersection of two divisions: between duties to oneself and to others, and between perfect and imperfect duties. Perfect duties are proscriptions of specific kinds of actions, and violating them is morally blameworthy; imperfect duties are prescriptions of general ends, and fulfilling them by means of performing appropriate particular actions is praiseworthy. The four classes of duty are thus: perfect duties to oneself, such as the prohibition of suicide; perfect duties to others, such as the prohibition of deceitful promises; imperfect duties to oneself, such as the prescription to cultivate one's talents; and imperfect duties to others, such as the prescription of benevolence (4: 422-3, 429-30 ). It is straightforward what a perfect duty prohibits one from doing; it requires judgment to determine when and how the general ends prescribed by imperfect duties should be realized through particular actions.

Note that Guyer states that the fulfilling of imperfect duties is praiseworthy. Now it is the supererogatory that we praise: it is inappropriate to praise people for doing what they are obligated to do. How morally absurd to praise a pater familias for paying the rent and putting food on the table! That is what he is supposed to do, what he morally must do. The failure to do such is blameworthy, but the performance is merely required, hence not praiseworthy. Since the fulfilling of imperfect duties is praiseworthy, it seems we can conclude that in Kant the class of supererogatory acts either is or is a proper subclass of the class of imperfect duties.

Further support for this interpretation comes at Grundlegung 429-430 where Kant speaks of "necessary or obligatory duties to others" and a "contingent (meritorious) duty to oneself." In English, 'obligatory duty' smacks of pleonasm, but that might not be the case for Kant's schuldige Pflicht. If duties divide into the obligatory (schuldigen) and the meritorious (verdienstlichen), then we can say that Kant accommodates supererogatory acts under the rubric of imperfect or meritorious duties.

There is more to it than this, of course, and there is a technical literature on this topic only a small amount of which I have read; but I think I can safely say two things: (i) a case can be made for Kant's being able to accommodate the supererogatory, and (ii) Kantians are more hospitable to the supererogatory than are utilitarians.

For my part, I have always thought it a fundamental error and an impoverishment of moral philosophy to reduce its scope to duties and obligations. Aristotelian and Stoic philosophy acknowledged that we had duties we should fulfill, but this was only one part and not the central part of morality. The central moral question was how we should live our lives in the best way possible. Part of the answer was indeed fulfilling the duties that pertain to the roles we assume (spouse, parent, citizen, teacher, etc).

BV: I think you are right about the impoverishment you mention. An adequate ethics would probably have to synthesize the insights of consequentialist, deontological, and virtue-ethical approaches. It is also worth investigating whether virtue theories will also have trouble accommodating the supererogatory. Suppose one maintains that

An action is right if and only if it is what an agent with a virtuous character would do in the circumstances.

This seems to imply that it is morally required that one act as a virtuous agent would. Now Dorothy Day and Mother Theresa are virtuous agents. But is one who is in their circumstances morally required to engage in their heroic and saintly acts? See here.

I’ve been thinking about how an Aristotelian would react to your example of feeding some street people. The Aristotelians (and Phil) would emphatically reject the claim that you had an obligation or duty to feed them. They were not relatives or anyone else that you could be thought responsible for. You never promised to help them. You were not responsible in any way for their becoming impoverished street people, were you? In short, you did nothing to obligate yourself to feed these strangers.

BV: Right, I did nothing to obligate myself to feed those strangers. But not all duties (obligations) are contractually acquired, or acquired by some act or ommission. Do you deny this? Arguably, one has some obligations just in virtue of being a moral agent. Surely I am morally obliged not to harm other persons. I have that obligation just in virtue of being a moral agent. It is not so clear that I have a positive duty to help other people, or that they have the positive right to be helped. But suppose I see a stranger having a heart attack. It would strike me as morally dubious in the extreme were you to maintain that I am under no obligation to pull my cell phone out of my pocket and dial 911. And don't forget we are talking morality here, not the positive law.

What you did in feeding them was perhaps a generous act, and if so, virtuous as such. I say perhaps, because we first need to consider whether your act observed the classical modalities of giving the right thing to the right people at the right time in the right amount for the right reason, ETC. I think it could be argued that giving money to people begging in this way is not a wise or proper choice. This is certainly debatable, but is rewarding apparently able-bodied people for begging in this fashion a good idea? I would not do it if accosted in an outdoor café by street people. I would see that they were promptly removed from the restaurant! That kind of rude, intrusive behaviour is not acceptable. This is not heartlessness, but a genuine concern to show them that aggressive begging is not an acceptable way to get by. Unless they are totally incapacitated, they need to be offering some service or entertainment (Ockham’s case) for which I might be willing to reward them. But begging or demanding money of people trying to eat their dinner is not a service.

It depends on the exact circumstances. In my example I described the bums as harmless (and thus no threat to other people in the cafe) and as truly needy. I never give money or anything to the panhandlers I run into on Mill Ave. in Tempe for they are young, able-bodied, shiftless, and not truly needy. Indeed, one is in some circumstances positively obliged to not help them.

Your point, I take it, is that in no conceivable circumstances is anyone under a moral obligation to help a stranger unless one has contracted to help him or done something (inadvertently damaged his property, say) that gives him a claim on one. If that is your point, then it needs defense, and I would say it is more in need of defense than its negation.

To avoid unprincipled appeals to moral intuition, though, one needs a systematic theory, and that is hard to come by.

Clear examples of precepts are the Ten Commandments. Poverty, (complete) chastity, and obedience, on the other hand, are counsels:

As there are three great hindrances to the higher life, so also the counsels are three, one to oppose each. The love of riches is opposed by the counsel of poverty; the pleasures of the flesh, even the lawful pleasures of holy matrimony, are excluded by the counsel of chastity; while the desire for worldly power and honour is met by the counsel of holy obedience. Abstinence from unlawful indulgence in any of these directions is forbidden to all Christians as a matter of precept. The further voluntary abstinence from what is in itself lawful is the subject of the counsels . . . . Vide.

I am at a loss to explain how any viable ethical scheme, whether 'humanist' or 'theological,' could fail to admit a category of actions called 'supererogatory.' And this quite independently of Catholic moral doctrine which to me qua philosopher is just another topic to be examined. Act utilitarianism as standardly formulated implies that that there cannot be a category of supererogatory actions, but then I have never quite understood the appeal of utilitarianism, and am tempted to respond, 'So much the worse for act utilitarianism.' But this remark calls for a bit of discussion which, of course, can only begin to scratch the surface of the surface.

Roughly, what the act-utilitarian is saying is that an action is right iff it maximizes utility. If our utiliarian is a hendonist, then utility is cashed out in terms of pleasure and we get something like:

An action A is right iff A leads to the greatest pleasure of the greatest number.

Now a right action is presumably not merely one that is permitted, but rather one that is obligatory. The right thing to do is what one must do. If so, then the above biconditional, read right-to-left, says that acts that are pleasure-maximizing are obligatory. But this is wildly counterintuitive.

Suppose I have ordered an expensive dinner in an upscale outdoor cafe for you and me, and three hungry but harmless bums approach our table. If they are truly needy, then I am inclined to say that it is obligatory that we feed them by, say, taking them to across the street to McDonalds and buying them burgers and fries. But that action would not maximize the pleasure of all concerned. There would be a greater total pleasure if I were to buy the bums meals equal in excellence to my own. Suppose I could easily do so, money being no object, and my doing so would not anger any of the other patrons of the restaurant or cause the place to becomne flea-infested. Still, it is obvious to me that I am under no obligation to do any such thing. At the very most, I am morally obliged to see to it that the bums get something to eat. Treating bums to pollo alla marsala is surely in excess of anything that could be morally demanded, even if they have the palate for such fare.

No doubt a sophisticated act-utilitarian would have ways of responding to this. But my present purpose is merely to bring out in schematic fashion how the topic of supererogation bears upon various ethical doctrines.

Since by definition a supererogatory action is one that is good but not obligatory, the above amounts to an argument for supererogatory actions. The argument is valid and the first premise self-evident. So the soundness of the argument rides on the second premise. Here, I suppose, an appeal to intuition is unavoidable.

So I will simply state that it is morally obtuse to praise someone for doing what he is obligated (whether legally or morally) to do. You don't praise a person for driving in accordance with the traffic laws, you blame him for failing to do so. It is a sign of moral confusion and moral decline to praise people for doing what they ought to do.

Praise and blame attach to the supererogatory and the prohibited, respectively. Neither attach to the obligatory.

While I'm on the topic of moral confusion, I saw an official sign on a pedestrian traffic signal in Tempe, Arizona. It read: Please do not cross against the light. The pertinent code was cited. Now what moral idiot formulated this sign? 'Please' is a word used in making a request, not in issuing a command. Am I being requested to not cross against the signal? If it is a request, then it is permissible for me to decline. "No thank you, I prefer to do as I please regardless of public safety and order." And if it is a request, why cite the traffic code? Or am I being commanded to not cross against the signal? But then 'please' is out of place.

If you cannot see the difference, then you cannot think in moral categories.

1. There are supererogatory actions
2. Whatever is good to do ought to be done
---
3. There are two moral senses of 'ought,' one prescriptive the other commendatory.

Ad (1). I am using 'supererogatory' in a wide sense. Accordingly, an action can be supererogatory (roughly: above and beyond the call of duty) even if it is not heroic or particularly costly (financially or otherwise) to the agent. It can be something as trifling as leaving a 50% tip in a restaurant. If an action is obligatory, then it is good to do, and bad to leave undone. If an action is supererogatory, then it is good to do, but NOT bad to leave undone. Suppose I have lunch in a restaurant in which the food is good and the service adequate. Then

a. I am legally and morally obligated to pay the bill.
b. I am morally but not legally obligated to leave a 15% tip. (Or so I would argue.)
c. I am neither legally nor morally obligated to leave a 50% tip.

Nevertheless it is a good thing to leave a large tip (other things being equal), but not bad if I fail to leave a large tip. So the action is supererogatory in the wide sense here in play.

Ad (2). The principle that whatever is good to do ought to be done is theoretically attractive. It is attractive because it forges a partial link between the axiological and the deontic which are two aspects of the normative. Axiological terms include 'good,' 'evil',' 'value,' 'disvalue,' 'ideal.' Deontic terms include 'obligation,' 'prohibition,' 'permission,' 'right,' 'duty.'

Note the difference between (2) and

4. Whatever is good ought to be done.

(2) is about the good-to-do; (4) is about the good-to-be, which is a wider category. (4) says that everything which would be good were it to exist ought to be brought about by someone. It suffices to refute (4) to adduce a state of affairs that would be good if realized, but is not in the power of any finite agent to realize. Well, it would be good if none of us had to to die; but the immortality of human beings is not a state of affairs that can be brought about by any finite agent or agents.

Much more needs to be said about the principle (2) but let's assume that one accepts it.

Now under what conditions can it be true both that (1) there are supererogatory actions, and (2) whatever is good to do ought to be done? On the face of it, these seem contradictory. The supererogatory is that which it is good to do but not obligatory. But if what is good to do ought to be done, then it seems to follow that what is good to do is obligatory — which contradicts (1).

The contradiction is avoided if we distinguish between two moral senses of 'ought.' In the prescriptive sense, what one ought to do is what one is obligated to do. In the commendatory sense, what one ought to do is good to do but not obligatory.

But are there any sentences that feature moral uses of 'ought' that do not impute obligations? Suppose my wife says to me, 'We ought to give more of our income to charity.' Suppose further that there is an obligatory percentage and that we are already giving that percentage. In saying what she says, my wife is not imputing an obligation to us; she is recommending that we do more than we are obligated to do. Since the context is moral, this seems to be a moral use of 'ought' that does not impute an obligation. This seems to be a clear counterexample to (P) above.

Suppose instead she had said to me, 'We ought to check out the Roy Orbison exhibit at the Tempe Center for the Arts before it is too late.' Visiting such an exhibit is a morally indifferent action: neither prohibited, nor obligatory, nor supererogatory. The use of 'ought' in this example is nonmoral. It like the 'ought' in 'If you want to get to Tempe from Gold Canyon by freeway, then you ought to head East on the Superstition Freeway.'

Robert Merrihew Adams (Finite and Infinite Goods, Oxford 1999, p. 235) offers this example: "The religious leaders of the world ought to launch an appeal to protect the physical environment in which humanity must live." Someone who says this is not saying that the leaders are under an obligation; he is saying that it would be a good idea were they to act as recommended.

To sum up. There are nonmoral and moral uses of 'ought,' but not every moral use is an obligation-imputing use. So the biconditional (P) is false. What is true is the conditional,

P*. If agent A is morally obligated to X, then A ought to X.

Note first that the following is not a counterexample to (P):

1. You really ought to get out for a hike while the wildflowers are still in bloom.

Someone who asserts (1) is not speaking from a moral point of view. He is using 'ought' in a nonmoral sense. He is saying to the addressee that it would be a good idea, not morally but aesthetically, to hike while the wildflowers are still blooming. The sense of 'ought' in (P), however, is moral. So (1) is not relevant to (P) and cannot tell against it. But consider

2. Given the amount of vile and irresponsible speech engaged in by cyberpunks, you ought not allow just anyone to comment at your site.

In (2) the sense of 'ought' is moral just as it is in (P). But someone who asserts (2) does not by so doing imply that the addressee is under a moral obligation to not allow just anyone to comment. (2) expresses a moral recommendation. It is morally commendatory rather than morally imperative. One who asserts (2) is saying that it is a morally good idea to not give a forum to just anyone.

Since (2) and (P) both feature the same sense of 'ought,' a moral sense of 'ought,' and (2) does not impute an obligation, (2) is a counterexample to (P). So it looks as if I ought to replace the biconditional (P) with the conditional

P*. Necessarily, if agent A is morally obligated to X, then A (morally) ought to X but not conversely.

This result seems to tie in nicely with what I said about the obligatory and the supererogatory. The morally obligatory is that which it is morally good to do but morally bad to leave undone, while the morally supererogatory is that which it is morally good to do but NOT morally bad to leave undone. What counterexample (2) shows is that there are actions that are morally supererogatory: morally good but not required.

Thanks to correspondent Phil O. Sophy the TerraVitan for help in sorting this out. Any errors that remain are of course also in part his responsibility!

Pursuing the analogy, we note that the following two alethic modal principles each has a deontic analog, where 'p' ranges over propositions and 'A' over actions:

1A. If p is necessary, then p is possible, but not conversely.
1D. If A is obligatory, then A is permissible, though not conversely.

2A. P is impossible iff ~p is necessary.
2D. A is impermissible iff ~A is obligatory. (e.g. Sexually molesting children is impermissible iff not sexually molesting children is obligatory.)

But of course the following ab esse ad posse principle has no deontic analog:

3A. If p is true, then it is possible that p is true.
3D. If A is done, then A is permissible.

This is not surprising since the deontic is a species of the normative.

Now let's try to characterize (not define) the deontic concepts of the permissible, the impermissible, and the obligatory in terms of the axiological concepts of good and bad:

4. If A is impermissible, then A is bad to do and good to leave undone.
5. If A is obligatory, then A is good to do and bad to leave undone.
6. If A is permissible then A is axiologically indifferent: neither bad to do and good to leave undone, nor good to do and bad to leave undone.

These characterizations seems to leave conceptual room for two other sorts of actions, actions that are good to do but NOT bad to leave undone, and actions that are bad to do but NOT good to leave undone. The former are supererogatory in a wide sense of the term, and the latter are suberogatory.

For example, paying one's bill in a restaurant is (legally and morally) obligatory: it is good to pay the bill and bad to leave without paying. But tipping is (legally, though perhaps not morally) supererogatory: it is good to tip but not legally required, and in this sense not bad to leave no tip. If this is unclear, consider tipping in excess of 20%. Surely this would be supererogatory: good to do (ceteris paribus) but not bad to leave undone. One could plausibly argue that restaurant patrons are morally obligated to tip in the amount of 15-20% of the tab given that waiters and waitresses work for minimum wage and rely on tips for a livable wage; but surely no one is obligated to leave a 50% tip, say. And yet it would be a very nice thing to do for that pretty waitress working her way through school. It would be good to do but not bad to leave undone.

Do you balk at calling such trifling actions supererogatory because they are not grand or heroic like throwing oneself on a grenade to save one's buddies? Well then, I will accommodate your point by distinguishing between wide and narrow senses of 'supererogation.' Or I might just give you the word as long as you grant me a category of actions an example of which I have just provided.

Now for some examples of suberogation. A supererogatory act is a good act that is in excess of what is morally or legally required. A suberogatory act is then one that is bad but not so bad as to be impermissible. The following might be examples: talking too much, public displays of affection, breaking wind in public, picking one's nose in front of others. The practitioner of public rhinotillexis does nothing legally or morally impermissible, but it is something bad, and not just aesthetically. I want to say that it is a morally bad action that not impermissible. So I say it is suberogatory.

More needs to be said. I am not sure that in the end the sub- and supererogatory don't collapse into the permissible thus obviating the need for two further categories of actions. Further reading: My essay On Tipping, and the SEP article, Supererogation.

I don't see that "x is something I ought to do iff x is something I'm morally obligated to do" is a conceptual truth, or even true. [. . .] Non-consequentialist moralities allow room for good deeds that are not obligatory. If helping a stranger is a good deed and you are fully able to perform it without endangering others, then I am quite comfortable recommending to you that you ought to do it. But I am not suggesting you have any duty or an obligation to do so. [. . .] So, you ought to help does not imply you have a duty to help.

I will now demonstrate that you ought to help does indeed imply that you have a duty to help, assuming that one is not equivocating on 'ought' and is using 'ought' as it is used in (P).

I agree that there are good deeds that are not obligatory. Suppose my neighbor is away when an important-looking package is delivered to his door. I take it into my house for safekeeping until he returns. Surely I am under no obligation, moral or legal, to do such a thing. Yet it is a good deed.

But it does not follow that it is a deed that I ought to do, or that I have a duty to do; it is precisely a supererogatory action, one above and beyond the call of duty. (A supererogatory action can be something as trifling as this, and need not be grand or heroic, but more on this in a separate post on supererogation.) If I ought to X, and I omit to X, then I do something wrong. Therefore, if I ought to pick up my neighbor's package, but omit to do this, then I do something wrong. But obviously I do nothing wrong in leaving my neighbor's package where it lies. Hence it is not the case that I ought to pick up my neighbor's package. Nor do I have any duty to pick up my neighbor's package.

I suspect my correspondent is simply playing fast and loose with 'ought,' a word with several meanings in English. Some examples:

a. 'The car ought to start; I installed a new battery.' This looks to be a non-normative use of 'ought,' one with no relevance to moral theory.

b. 'If you want to get to Tucson from Phoenix by interstate highway, you ought to take I-10 East.' This sentence is a hypothetical imperative, and the subject-matter is morally indifferent.

c. 'If you want to be a successful hit man, then you ought to learn how to kill with a .22 caliber gun.' A second hypothetical imperative. Here the subject-matter is not morally indifferent, but the 'ought' has noting to do with a duty.

d. 'If helping a stranger is a good deed, and one wants to be helpful, then one ought to help.' Another hypothetical imperative, and close to what my correspodnent said above. But this use of 'ought' is not the use in principle (P) above.

e. 'You ought to pay your debts.' A categorical imperative, and a morally relevant use of 'ought.' This is the use of 'ought' that is featured in (P) above.

In sum, (P) seems rock-solid and I will continue to adhere to it until someone can instruct me otherwise.

What I just said is by no means obvious, but it is very plausible; suppose you grant it. The question that interests me is this: How far must one go in the prevention/alleviation/elimination of evils? If I am obliged to give the poor devil in the lonesome canyon some of my water, am I obliged to prevent/alleviate/eliminate every evil I know about and can prevent (etc.) involving anyone as long as I don't shirk other duties to myself and others? Or are there some such evils the prevention (etc.) of which by me would be supererogatory, i.e., above and beyond the call of duty? (I assume it to be a conceptual truth that whatever one ought to do one is morally obligated to do, and vice versa.)

Suppose the poor devil in the canyon is not only thirsty but also hungry and blistered. It seems reasonable to say that although I am morally obliged to give him water (to ward off imminent heatstroke), I am not obliged to give him food and moleskin, even though the pains of hunger and blistered feet are evils. If I minister to him in those further ways inessential to his survival, then my deeds are supererogatory.

As I understand the obligatory/supererogatory distinction, an agent's failure to perform a supererogatory act that he can perform in no way diminishes his moral goodness. This suggests a definition:

D1. An action X performed by agent A is supererogatory =df (i) X is permissible but not obligatory; (ii) X exceeds what is morally required; (iii)A's failure or refusal to perform X in no way diminishes A's moral goodness.

Now I have just given two examples of acts that look to be supererogatory: giving the greenhorn food, and giving him moleskin. The guy's feet hurt and that is a low-level evil. Suppose the evil is dysteleological: it serves no good purpose such as that of teaching the bonehead that setting out on a ten mile hike in brand-new boots is a stupid thing to do. (For if he truly is a bonehead, he won't learn from the experience.) Even on the assumption of dysteleologicality, it seems that I am under no moral obligation to provide foot care, and that my failure to provide foot care to the witless biped does not diminish my moral goodness.

Further examples are easily constructed. Now if there are supererogatory actions, then one or both of the following principles would seem to be false:

P1. A morally perfect being prevents/alleviates/eliminates all gratuitous or dysteleological evils that it knows about and can eliminate.

P2. A morally perfect being prevents/alleviates/eliminates all evils that it knows about and can eliminate.

Now suppose that an argument from evil employs (P1) or some equivalent as I think Peter Lupu's LAFE does. Then it may be possible to neutralize such an argument by citing a possible case in which an omni-qualified being does not prevent (etc.) a dysteleological evil, but without prejudice to his being perfectly (completely) good, because the removal of the evil would be a supererogatory act.

Suppose I stub my toe, mildly, twice in a row. Perhaps the first toe-stubbing serves the purpose of warning me to be careful. It leads to a greater good. But the second stubbing is dysteleological, and the pain, which is a low-level evil, is gratuitous or dysteleological. How then can an all-good God, who is all-knowing and all-powerful, allow such an evil to occur? Surely God could have created a world in which Vallicella on 25 March 2008 stubs his toe only once! On some versions of LAFE (and perhaps on Lupu's) just one such instance of evil suffices to prove the nonexistence of God. For on LAFE an all-good God simply cannot exist if even one instance of gratuitous evil exists!

The 'evil' premise in Lupu's LAFE is simply this: Evil exists in the world. (p. 4) Now for evil to exist it suffices that there be just one instance of evil, e.g. the evil of my second toe-stubbing. But how could one instance of evil, even if it is gratuitous, prove the nonexistence of God? That sounds absurd. It is on a level with saying that if God existed, then turkeys would fly around ready-roasted, thereby saving us much bother on Thanksgiving; but they don't, so God does not exist.

The reasonable thing to say, it seems, is that God's failure to stop me from stubbing my toe a second time does not diminish his goodness one bit. And this for the reason that such an intervention would have been supererogatory. So I am wondering whether logical arguments from evil can be neutralized by appeal to the obligatory/supererogatory distinction. Evidential arguments from evil are of course another story.

On my view morality and legality are both 'notionally distinct' and 'extensionally distinct.' Thus I maintain both that the concept of a morally qualified action is distinct from the concept of a legally qualified action, and that there are instances of each concept that are not instances of the other. If the following three principles are defensible, then the notional and extensional distinction between morality and legality is defensible.

P1. Not everything morally impermissible should be legally impermissible.
P2. Not everything morally obligatory should be legally obligatory.
P3. Not everything morally permissible should be legally permissible.

Ad (P1). This could be put idiomatically by saying that not everything morally wrong should be illegal. I hope you will agree with me that drinking oneself into a stupor is morally wrong. It's a Friday night, you are alone, you stay in your house the entire time, but you down a quart of whiskey. I hope you will agree with me that this is morally wrong. (I could give an argument, but it belongs in a separate post.) Though morally wrong, such behavior should not be illegal. If you don't like my example, you are invited to concoct one of your liking that makes the same point.

Why should not everything morally wrong be illegal? The positive law is made by human beings and enforced by state power. But it is notorious that power tends to corrupt its possessors, and this is especially clear in the case of the agents of the state. (Ever have a run-in with a cop, a sheriff, an IRS agent who overstepped the bounds of his legitimate authority?) Given the egregious violations of individual liberty including mass murder that state agents have been known to commit, reasonable people support limited government. One way to limit government is to restrict its sanctions to behavior that impinges upon public welfare. Thus there need to be laws against drunk driving. But laws against private drunkenness (as in my example above) tend to give too much power to the state, besides being unenforceable. Laws that are unenforced, however, tend to breed disrespect for law. Laws should be few in number, rational in content, and enforced with credible sanctions. So not everything morally wrong should be illegal.

Ad (P2). If you morally must do (ought to do, are obligated to do, have a duty to do) X, should the positive law require it of you? I answer in the negative. Consider the philosophical/religious duty to worship no such 'false gods' as money, power, sexual pleasure, Lady Luck, but to worship or devote yourself only to worthy objects such as God, truth, justice, benevolence, and the like. I would argue that it is morally impermissible to worship any such thing as money, sex, etc. and morally obligatory to apportion your highest concern only to worthy objects such as truth and justice. But these views of mine, thought rationally defensible, can be opposed with a show of rationality by sincere people. A hedonist might sincerely and thoughtfully maintain that pleasure alone is the good and that there can be no grounded distinctions among pleasures, and that therefore the pursuit of sexual pleasure among consenting adults is a perfectly legitimate and morally permissible object of pursuit. I would consider it better if people devoted themselves to truth over sexual pleasure, but the abuses that would result from using state power to enforce that view of mine would be worse than the positive outcomes would be good.

For a second example, consider the moral obligation to maintain one's physical health by the usual measures including good diet and vigorous exercise. If you value individual liberty and understand what I wrote above, you will not want laws that mandate these measures. (By the way, this is why 'universal health care,' i.e., government conrol of the delivery of health services ought to be opposed: a government that controls health care will be in a position to demand 'apropriate' behaviors from people including not engaging in activities deemed risky.)

Ad (P3). If an act is morally permissible, should it also be legally permissible? No. Surely there are circumstances in which running a red light is morally permissible. Suppose at three in the morning you slowly and cautiously coast through a red light at a deserted intersection in the proverbial 'middle of nowhere.' Surely that is morally permissible. Yet it ought not be legally allowed: a well-crafted law cannot have a clause that reads, "A motorist shall not run a red light except when it is safe to do so."

For a second example consider that is it morally permissible for some 15 year olds to drive. But it ought not be legally permissible for 15 year olds to drive. The same goes for voting. There are 15 year olds more mature than some 50 year olds and more fit to cast a vote. Yet it ought not be legal for 15 year olds to vote. Finally, it seems obvious that some marijuana use by some people in some circumstances is morally permissible, but it scarcely follows that marijuana use ought to be legalized. What is harmless for some will be dangerous for many. Laws crafted for the common good cannot take into account individual differences.

Consider a mother and her child. The mother has various prima facie obligations with respect to the child. One of these obligations is to protect the child from preventable instances of needless suffering. Among the definitions and principles I assume are these:

P1. One is obligated to do X iff one ought to do X.
P2. 'Ought' implies 'can.'
P3. Prima facie obligations can be overridden.
P4. If X is obligatory, then X is permissible, but not conversely.

Suppose the child will die unless he has a series of painful operations. On this supposition the mother's prima facie obligation to protect the child from preventable instances of suffering is overridden by the 'greater good' of being cured of a terminal disease. Therefore, the mother is morally justified in permitting her child to suffer the pains that are the unavoidable concomitants of the series of operations. We can express this by saying that the mother's allowing of her child's suffering is morally permissible. Now if an act is permissible, it does not follow that it is obligatory. But in the case before us, the mother's allowing of her child's suffering, since it is necessary for her child's cure, is also morally obligatory. That is, the mother ought to allow the suffering; it is not the case that she merely may allow the suffering. For she ought to do what she can to help her child, and among the things she can do is allow the operations; therefore, she ought to allow the operations and the suffering that is unavoidably concomitant with them.

It seems to follow that the suffering associated with the operations ought to occur, and that the mother has the moral obligation both to permit it and to do what she can to bring it about. Now suppose the suffering occurs. Then it both is and ought to be. It is because it occurs. It ought to be because its occurrence is necessary for a greater good. Is this paradoxical or counterintuitive? I think not. It may sound paradoxical to say that the child ought to suffer certain pains; but the paradox is removed when it is realized that this ought to be the case only on condition that without these sufferings the child would be worse off than with them.

Perhaps a paradox can be generated if God is put in place of the mother, and in place of the child's suffering (which is necessary for a greater good) we put every instance of suffering in the actual world. The idea is that all the instances of suffering in the world, although prima facie such that they ought not exist, are morally justified by their roles in contributing to greater goods. It would seem to follow that not only would all of those instances of suffering be permissible, but they would also all be morally obligatory in that without them the maximal amount of good would not be attained.

Now Lupu finds this paradoxical and indeed incoherent:

Therefore, every existing instance of suffering ought to have occurred; and every moral agent, including a perfect God, has the moral obligation to permit it or bring it about. So, according to this theodicy, whatever is the case, ought to be the case.

What an astonishing an counterintuitive result! It threatens to obviate moral distinctions and undermine the very purpose of our moral conceptual system. (p. 26, emphasis in original)

Now it surely sounds absurd to say that, since a world containing the Holocaust is (tenselessly) the case, that it ought to be the case. But the counterintuitivity may be lessened if we bear in mind that a world containing free agents is of greater value than one without them, and that God has no control over the actions of free agents. Absolutely considered, there ought to be no evil. But given free agents and their wrong choces, there is evil and this evil must be permitted by God if there is to be the greater good of finite free agents.

What follows from a higher-goods theodicy is not that whatever is the case ought to be the case, but that whatever is the case ought to be the case GIVEN that free finite agents exist and make wrong choices. Thus God is not morally obligated to actualize a world containing genocide; he is morally obligated to permit the existence of a world containing genocide given that a world containing free agents is better than one without such agents.

As far as I can see, Lupu hasn't made the case that a higher-goods theodicy entails the collapse of the distinction between is and ought.

In developing the argument for atheism based on the existence of evil, it will be useful to focus on some particular evil that our world contains in considerable abundance. Intense human and animal suffering, for example, occurs daily and in great plenitude in our world. Such intense suffering is a clear case of evil. Of course, if the intense suffering leads to some greater good, a good we could not have obtained without undergoing the suffering in question, we might conclude that the suffering is justified, but it remains an evil nevertheless. For we must not confuse the intense suffering in and of itself with the good things to which it sometimes leads or of which it may be a necessary part. Intense human or animal suffering is in itself bad, an evil, even though it may sometimes be justified by virtue of being a part of, or leading to, some good which is unobtainable without it. What is evil in itself may sometimes be good as a means because it leads to something that is good in itself. In such a case, while remaining an evil in itself, the intense human or animal suffering is, nevertheless, an evil which someone might be morally justified in permitting.

Suppose that to be restored to health a child must undergo an extremely painful medical treatment. So the parents of the child allow the treatment to be administered. We will agree that the infliction of the suffering upon the child is morally justified by the fact that the treatment is necessary to prevent a greater evil which we may assume to be the child's death. Now what Rowe is saying above is that in a case like this, the suffering is (morally) justified but evil nevertheless.

I find this difficult to understand. It sounds like a contradiction. For if the infliction of the suffering is morally justified, then the infliction is morally permissible. But if the suffering is morally evil, then its infliction is also morally evil, which is to say that its infliction is morally impermissible. But surely it is a contradiction to affirm of any action A that A is both morally permissible and morally impermissible.

If the suffering is morally justified in that it leads to a good unobtainable without it, then the suffering, though certainly unpleasant, diagreeable, repugnant, awful, excruciating, etc., is not under the conditions specified evil. In the situation we are imagining, it is not only morally permissible but also morally obligatory for the parents to allow the painful treatment to be administered. This implies that the treatment ought to be administered. Therefore, if you say that the child's suffering remains evil despite its leading to a greater good, then you are committed to saying that the infliction of evil upon the child is morally obligatory, something that ought to be done. But this smacks of absurdity since it is hard to understand how any infliction of evil could be morally obligatory. Since in our example the infliction of suffering is morally permissible, I conclude that (intense) suffering is not in every case evil.

What Rowe is saying is that (intense) suffering is intrinsically evil, and that its evilness remains the same whether or not the suffering is instrumentally good. What I am suggesting contra Rowe is that whether or not suffering is evil depends on whether or not it is instrumentally good. For me, suffering that is instrumentally good is not evil. I concede of course that such suffering remains unpleasant, disagreeable, repugnant, awful, excruciating, etc. But I do not understand how suffering in itself, or intrinsically, can be said to be evil.

Perhaps the problem is that there are two senses of 'evil' in play, one nonnormative the other normative, and that Rowe is appealing to the former sense. Accordingly, the nonnormatively evil is that which elicits aversion. In this sense, mental and physical suffering is evil in that beings like us are prone to shun it. The normatively evil, on the other hand, is that which ought not exist. So perhaps my puzzle can be resolved by saying:

1. Every instance of suffering is evil in the nonnormative sense that, as a matter of empirical fact, beings like us are prone to shun it. (Being prone to shun X does not entail the shunning of X.)

2. Some instances of suffering are not evil in the normative sense that it is false that they ought not exist.

3. If an instance of suffering conduces to a good that outweighs it, and the good is unobtainable by any other means, then the instance of suffering ought to exist. Thus the child's suffering in our example ought to exist. Admittedly, this sounds paradoxical. But note that this 'ought' is not categorical but hypothetical or conditional: the child's suffering ought to exist given that, on condition that, the treatment that causes it is the only way to avoid the child's death, which would be an evil worse than the child's suffering from the treatment.

4. (3) is not paradoxical or incoherent.

5. The moral goodness of God is called into question not by the existence of evils in the nonnormative sense, but by the existence of evils in the normative sense. Thus the mere existence of suffering, which is nonnormatively evil, does not by itself cause a problem for the divine moral goodness. For it may well be that all instances of suffering are morally justifiable in the light of a greater good. This does not make these sufferings any less repugnant; but this repugnance is not a moral repugnance but the nonnormative property of thwarting desire.