According to H.A. Prichard, the practice of moral philosophy rests on a misguided attempt to provide general, theoretical validation for principles of moral obligation. This is a mistake because moral obligations (that is, the facts that we ought to act in certain ways) are, in reality, self-evident.(1)
These claims are, no doubt, at least slightly exaggerated--not all moral obligations are self-evident, and it is not always a mistake to try to provide arguments for moral judgements; nor have all moral philosophers devoted themselves to validating morality in the manner Prichard describes. Nevertheless, there is considerable truth in Prichard's observation, and with the help of some distinctions and qualifications that Prichard would have accepted, we can render his doctrine into a form that is true.
To begin with, we need to identify what class of principles Prichard is concerned with that other philosophers have mistakenly tried to validate. He is concerned specifically with principles of obligation--principles identifying the right--and not with evaluative propositions in general. A principle of obligation is a principle that could be stated in the form "x ought to do A" (or "A is obligatory" or "A is a duty") Evaluative propositions include principles of obligation, but they also include some other propositions, most notably those of the forms, "x is good" and "x is a virtue". So "Pleasure is good" would be an evaluative statement, since it uses the predicate "good", but it is not a statement of obligation, because it does not identify a course of action that some agent is required to enact. Prichard's claim, then, is specifically this: Other moral philosophers have tried to derive principles of obligation from other facts which were not principles of obligation--sometimes these other facts have been descriptive, and sometimes they have been evaluative, but in any case the attempts have been misguided. Henceforth, let us use the term "morality" to refer to principles of obligation.
Second, we must distinguish what we might call common sense morality from philosophical morality. Common sense morality is the body of principles of obligation that are widely and firmly accepted (allowing for exceptions on the part of some sociopaths and some philosophers), implicitly or explicitly, prior to argument and philosophical training. For example, one ought not to tell lies, one ought to keep agreements, one ought not to attack others without provocation--these are all elements of common sense morality. "Philosophical morality", on the other hand, refers to theories and systems devised by philosophers to explain the nature of morality and/or to compute, in any given case, what one ought to do. Philosophical morality generally results from more or less involved argumentation and analysis. Prichard's complaint is directed against philosophical morality. Common sense morality does not rest on any sort of mistake; common sense morality is a body of perfectly good, rational knowledge. It is philosophical morality that rests on a mistake, the mistake of trying to validate common sense morality through something else.
Third, we need to distinguish two possible ways of trying to validate moral principles. On the one hand, one might try to validate a moral principle by deriving it from other principles of obligation and/or from 'ought' judgements about hypothetical cases (thought experiments). For instance, suppose someone argues that people ought not to have abortions, because (a) to have an abortion is to kill a person, and (b) one ought not to kill people. This sort of argument, in general, is no mistake. What would be a mistake is to try to validate moral principles in the sense of deriving them from non-moral principles. That is, if someone tried to show that we ought not to kill people, and he relied on some argument in which the premises did not contain any principles of obligation, this would be an error on his part, and his argument would fail. And that is what moral philosophers throughout history have very often done.
According to Prichard, the reason philosophers have sought this kind of theoretical validation of morality lies in the aversion that people often have to performing their obligations. Initially, we become aware of various obligations intuitively. But we frequently find these obligations irksome, and so we are inclined to ask, "Why must I do these things?" Indeed, we may feel tempted to reject morality. Moral philosophy then arises to answer the question, and to prove to the moral skeptic that he really should do these things that we normally consider obligations.
The reason this is a mistake is, in the first place, that the project is impossible to carry out; it is not possible to derive an 'ought' judgement from non-moral premises. And, in the second place, it is not necessary to derive 'ought' statements from non-moral premises, because common sense morality is itself a body of foundational knowledge. Our reason is able to directly perceive the truth of certain principles of obligation, just as we can directly perceive the truth of the law of identity, or of Euclid's axiom, "The shortest path between any two points is a straight line".
Plato provides one of the clearest positive instances of Prichard's thesis about the nature of moral philosophy. The Republic is, in fact, one long attempt to answer the moral skeptic. Thrasymachus' initial challenge, embodied in the insistence that justice is merely 'the interest of the stronger' and that to be unjust is better and wiser if one can get away with it, is more of a challenge to the concept of justice than an attempt really to define it. That is, although Thrasymachus' thesis is phrased as a definition of "justice", it strays so far from the ordinary meaning of the term that it is clear we are to understand him as cynically declaring: "'Justice' is a fiction invented by rulers, and morality a kind of hoax they put over on their subjects in order to advance their (the rulers') interests."
Glaucon and Adeimantus add to this challenge the suggestion that it is not justice per se that is desirable but only the reputation for justice. So it is clear that the answer to the immoralist or amoralist is Socrates' project. But the project of showing that justice is desirable for its own sake gets blended together without Plato's noticing it with the project of showing that justice is desirable for the just man; and so the ensuing dialectic aims at showing that the just man is always better off than the unjust man; and this is the whole of the reason why we should be just, as far as Plato can tell.
Now this line of reasoning fails for two reasons: first, because the proof that the just man is always better off than the unjust man is inevitably strained. It essentially involves Plato's interpreting justice as a kind of inner harmony of the soul, which state is then said to be intrinsically desirable; but the connection of this state with the form of behavior that we normally think of as just action is then rendered dubious and, in fact, never demonstrated. Plato never undertakes to show that the man who is governed by reason and has inner harmony does not lie, steal, and harm other people; he takes this for granted, perhaps because he fails to distinguish 'justice' as he defines it and 'justice' in the sense of the specific kinds of action that Thrasymachus, Glaucon, and Adeimantus wanted him to prove to be obligatory.
But the deeper reason Plato's line of reasoning, or any reasoning similar to it, can not succeed is that even from the supposition that the just life always profits a man it would not follow that a man ought, that is, is obliged to, act justly. This is easily seen by an analogy: it can perhaps be shown that a man will be better off if he sleeps for eight hours each night. But it cannot be inferred from this that a man has a moral duty to sleep for eight hours. The latter conclusion could be derived only if we were given the additional assumption that a man has a duty to secure his welfare. And on this rock all attempts to derive morality from hypothetical imperatives must founder. From the premises
(A) that I wish to destroy Davison Hall,
and
(B) that dynamite is the necessary and sufficient means to destroying Davison Hall,
it plainly does not follow that I ought to dynamite Davison Hall. We would have, to arrive at this conclusion, to include the assumption
(C) that I ought to satisfy my wish,
for if it be supposed for any reason that I should not satisfy my desire to destroy Davison Hall, then it will as clearly follow that I ought not to dynamite Davison Hall as it follows from A + B + C that I ought. Now (C) is a moral judgement (if not a particularly plausible one), so the 'ought' conclusion can not be derived without a moral judgement among the premises.
Epicurus adopts roughly the same strategy of trying to derive morality from self-interest, though unlike Plato he defends justice on grounds of its beneficial effects (a difference which, however, makes no difference). Epicurus claims that the rules of justice are established for the mutual advantage of human beings and that they lose their validity as soon as they cease to be advantageous. Furthermore, he argues that each person ought to abide by this compact because if he does not he will live in constant fear of being caught and punished. This fear would disturb his repose and so render his life unpleasant, even if he never was in fact caught.
Now this argument fails for the same reasons as Plato's: first, because the intermediate conclusion is shaky. Since the fear Epicurus wants to avoid is really the fear of being thought to be unjust by others (which can eventuate even if one is not in fact unjust), he must argue that the probability of being thought unjust is always greater when one performs an actually unjust action than it is when one acts justly, if he would argue that justice always profits a man more than injustice.
Second, even if the just life is always more pleasant than the unjust life, it does not follow from this that we are obliged to act justly, nor, certainly, is it this that forms the basis of our ordinary sense of the obligation of justice.
Hobbes, however much he may differ from Plato in other respects, rounds out our list of philosophers who tried to validate morality on an egoistic basis. He claims to derive various moral principles, including the obligation of justice, from self-interest. In brief, the argument is that all men desire peace, because conflict with other men causes widespread destruction, precludes economic productivity, and puts men in constant fear for their physical safety; and the principles of justice (or rather, people's obedience of them) are the necessary conditions for attaining peace; therefore, we ought to obey those principles.
Again, the first stage of the inference is strained. Granted, I desire peace with others (at least to the extent that they should not attack me); and granted, this can only be achieved if the principles of justice are generally obeyed. But is this any reason why I personally should obey them? The peaceful state of society will not break down if one person chooses a life of injustice; we know this, since there in fact are some such people in our society. And the one who chooses the life of injustice will get the best of both worlds: he will live in a peaceful and productive society, and he will also gain the benefits of predation on his fellow men. Why wouldn't the best plan, then, be to be unjust in a society of just men, provided one takes care not to be caught? (This is exactly Glaucon and Adeimantus' problematic in The Republic; but Hobbes lacks Plato's insight into what would be required to resolve it within the egoistic framework--namely, that justice should be shown to be intrinsically desirable for the just man.)
Second, this is another instance of the way that attempts to derive morality from hypothetical imperatives fail. Even if it be supposed (as Hobbes did) that all men desire nothing but their own advancement, and even if it could be shown that justice always advances your own interests, it still would not follow that we ought to be just. The argument perhaps will make us want to be just, but it cannot reach the distinctively moral conclusion without some such assumption as that we ought to act on our desires.
Bishop Butler's attempt to validate common morality holds more promise, as he makes no implausible pretense of reducing morality to self-interest. He claims instead to derive morality from 'our nature': right action is supposed to be that which accords with our nature in a certain sense, while wrongful actions go against our nature. That 'our nature' is, at least some of the time, being used as a descriptive and non-evaluative concept appears from Butler's analogy of the watch: a watch is supposed to have a nature, a function (namely, telling the time of day), in just the same way that we do. It is important to realize that the reason the watch's nature is to tell time, is not that we observers choose to use it for that purpose. Rather, as Butler has it, the reason is that the parts of the watch in their characteristic watch-like arrangement are adapted to telling time. This is a characteristic inherent in the watch: you can discern it by empirical examination of the watch itself. (233-4)
Butler seems to think that an evaluative conclusion follows from the observation of our nature. (258) But it is far from clear that we ought always to act in accordance with our nature. Suppose it were discovered that I (an aggressive male) am naturally adapted to fighting other people. I presume it will be granted that such a nature is possible--I could, for example, have naturally quick reflexes, physical strength, and an innate bloodlust; I might even have a naturally fearsome visage, suited to intimidating my opponents. Would it follow from this that I ought now to go out and attack people physically? Butler would argue that this is not in fact the nature of men, that, in the actual world, we are all adapted to peaceful, cooperative social life. The point nevertheless remains that this is not the reason for our moral obligations, since even if some person had a different 'nature' in Butler's sense, we would still hold him to the same obligations.
Moreover, Bishop Butler commits a fallacy parallel to that committed by our previous philosophers. In order to conclude, from the premise that A is conformable to our nature, that we ought to do A, it would be necessary to suppose that we ought to conform to nature. And to the question, "Why ought I to act according to my nature?" Butler offers no answer.
Finally, even supposing him to have established that we ought to conform to nature, Butler still can provide no grounding for our (non-trivial) principles of obligation, because all he discovers about our nature is that conscience is naturally suited to rule over inclinations and self-love. Like Plato (with "conscience" substituted for "reason"), Butler must presuppose that the principles conscience dictates are those of common sense morality, before he can provide any reason for adhering to those principles--which is to say, in effect, that he must assume us to have previously apprehended certain moral principles as valid, and only then does his argument provide a reason for acting in accordance with those principles. But, as Prichard would point out, by the time you know that your conscience dictates act A, you already ipso facto know that you ought to do A; if you did not really know that you should do A, then your conscience would not be dictating it. There is therefore no work for Butler's general theory of virtue to do.
Utilitarianism represents perhaps the most plausible derivation of principles of obligation, because it does not (usually) attempt to derive evaluative conclusions from descriptive premises. It seeks to derive principles of obligation from facts about the good: from the premise that an action produces more good than any other available action, the utilitarian desires us to conclude that the action is right. But this mode of validation of the principles of obligation suffers from the same two problems as all the preceding ones:
In the first place, it gets the wrong answers. Both Hume and Mill go to lengths to try to show that, from the principle of utility, it will follow that we ought to act justly, but both authors are able to make their case only by sliding between act- and rule-utilitarianism--that is, whereas they at first seem to propound act-utilitarianism, when it comes to explaining why we should follow the principles of justice, they resort to the claim that these principles, if adopted as general rules, promote utility. But even this claim does not hold up, for they do not consider all the possible alternative rules that we could adopt in lieu of the common sense principles of justice. We could, for example, adopt a rule such as, "Act justly, except when you can promote greater utility by acting unjustly," and, even for rule-utilitarianism, this would have to be admitted to be the better rule to adopt, in preference to simply "Act justly." As a result, the utilitarian has to say that it would be permissible and even obligatory to break promises, steal, kill, and otherwise violate principles of common sense morality whenever greater overall happiness could be produced thereby; and several well-known thought-experiments can be adduced in which such results strike us as counter-intuitive.
Second, the utilitarian requires a premise, "One ought always to promote the greatest good." Now if he states this as an ethical first principle, then he is at least not committing the mistake Prichard criticizes; his mistake then is, first, in holding promoting the good to be the sole fundamental obligation; and, second, in holding this principle to be adequate to deriving the commonly accepted maxims of justice. Now both Bentham and Mill claim the principle of utility to be the first principle of morality and itself not in need of proof. But if the principle that we ought to promote happiness is acceptable as a first principle--if, that is, it does not require to be proved--then why the principles of justice (we ought not to steal, to break agreements, to punish the innocent, etc.) should not equally be accepted as first principles becomes obscure. Why does the utilitarian feel that just action needs to be justified while benevolent action does not?
Utilitarians have not always been clear-headed enough to recognize the necessity of a distinct principle, "One ought always to maximize the good" (distinct, that is, from the premise that the good is happiness), in order to derive their principle of obligation, and insofar as they have neglected this point they have been guilty of the mistake Prichard criticizes. Bentham's confusion on this matter is revealed in his claim that "right" can only mean "conformable to the principle of utility" (370), which (incorrectly) would make the principle of utility analytic. Even G.E. Moore, amidst a pointed exposition of the naturalistic fallacy, proceeds to identify rightness with the property of producing the greatest good. (§5, 99-101)
I said at the start that there are two reasons why the attempt to derive moral principles (principles of obligation) from non-moral principles--or, as Prichard puts it, the attempt to arrive at moral knowledge through a process of non-moral thinking--is misguided. The first of these reasons we have already discussed sufficiently, as it is illustrated by the case of each of the above philosophers--namely, that such an attempt is necessarily doomed to failure. If we exclude 'ought' judgements from our premises, no 'ought' judgement can be obtained in the conclusion, except by fallacious reasoning.
The second reason is what we have left to elaborate. The attempt to generally validate common sense morality by philosophy is epistemologically ill-conceived because we do not have a firmer basis of moral conviction on which to stand in evaluating common sense morality. It is our moral common sense that must form the basis of philosophical morality, and not the other way around. The reason for this lies in general principles of epistemology:
In other words, foundationalism is true. This is so because (a) justification cannot be circular, and (b) we cannot have an infinite regress of justifications for any actual item of knowledge (I assert both of these as unargued premises). Hence, any actual item of knowledge we have must rest on, or itself be, some principle(s) that is not itself supported by anything further. I understand a 'first principle' to be a proposition that may be known without being justified through any other proposition.
Now if this is the case, then the interesting epistemological question arises: what propositions, or what sort of propositions, ought we to accept as first principles? This we seek to address by the succeeding observations.
By the "initial plausibility" of a proposition, I mean the degree to which it initially seems true to an observer when he attends to it, independent of arguments considered for or against it. That there must be such a thing as initial plausibility follows from (1) (assuming knowledge exists), since otherwise we could never come to an opinion. And consideration of a few examples should satisfy the reader that this characteristic comes in varying degrees. It is, for instance, initially plausible to me that I have two hands (it appears that way to me, without argument); the Axiom of Choice is also initially plausible to me; but the former is more initially plausible, more obvious to me-- I would give up the Axiom of Choice sooner than I would give up "I have two hands." (This example is also interesting because it shows that an empirical proposition can be more subjectively certain than a mathematical principle.)
By a "rationally persuasive argument" I mean an argument which would persuade a man, if he were rational (the rational persuasiveness of an argument is relative to an individual observer at a time). I take it that a person is 'persuaded' by an argument only if it brings him to believe the conclusion, while he did not believe the conclusion initially. Now, suppose a person, P, were persuaded of a conclusion by some argument in which the premise is less initially plausible for him than the conclusion. Let the premise of the argument be A, and let the conclusion be B. Now if P did not accept A, then there would be no reason for him to accept B; he would, in fact, be irrational to accept a conclusion based on a premise if he didn't believe the premise. Therefore, P must initially accept A. But since the argument is to persuade him of B, he must initially not accept B. But we have also said that B is more initially plausible for P than A--thus, in other words, to P, B more clearly seems to be true than A does--and yet he accepts A while not accepting B. Thus, P is irrational.
From this, we see that if a person is rational, he is persuaded by an argument from A to B only if A is more initially plausible than B.
Suppose our observer P to be again persuaded by a (valid) argument from A to B. Assume this time that A is less initially plausible than ¬B. Now, again, for P to be persuaded, if he is rational, he must believe A but not believe B. And since ¬B is more plausible than A and he believes A, he must also believe ¬B (presumably without having yet noticed the inconsistency between A and ¬B). Now, then, P is simply faced with a choice: He could accept the argument "A; therefore B" or he could accept the argument "¬B; therefore ¬A". Both of these arguments are equally valid, and both start from premises that seem to him to be true; but the latter starts from a more plausible premise. P therefore could not rationally prefer the first argument to the latter. He could not, that is, prefer to keep A over ¬B (these being inconsistent beliefs), given that ¬B is the more plausible, for in that case he would be preferring to believe what less appears true rather than what more seems true.
A little reflection will show this to be a sound principle of informal logic.
This may serve for a definition of "common sense". Consider these examples: I have a body; my body has existed for several years past; it has occupied space during all this time; there have been, and still are, many physical objects; the world has existed before I was born and will continue to exist afterwards; I also have a mind and I have some beliefs; etc.(2) These are all common sense beliefs. Now I would claim that such propositions as "One ought to keep promises"; "One ought rather to seek happiness than suffering"; "One ought not to punish the innocent or reward the guilty"; "One ought not to attack other people spontaneously" (with an "other things being equal" clause understood in each case) belong in the above list. They seem to me to have the same sort of pre-theoretical obviousness, and tend to be taken for granted in ordinary life in the same way, as the propositions in Moore's list.
Moreover--and this is the point of import--they seem to me to be more initially plausible than any philosophical theory. Certainly they have greater initial plausibility than any of the suppositions on which previous moral philosophers have sought to found morality (as, e.g., that we must always serve ourselves, or that we must always conform to nature).
This follows from 4 + 5, since if an argument against common sense (a skeptical argument) appeared, we should inevitably believe the denial of its conclusion more than its premises; we therefore would take the argument as a reductio ad absurdum of its own premises.
This follows from 3 + 5, since there are no principles more plausible than common sense from which we could derive them.
For they are initially plausible (5), they can never be overturned (6), and they can not be based on anything else (7).
Thus, in the words of Thomas Reid:
Common Sense holds nothing of philosophy, nor needs her aid. But, on the other hand, Philosophy (if I may be permitted to change the metaphor) has no other root but the principles of Common Sense; it grows out of them, and draws its nourishment from them. Severed from this root, its honours wither, its sap is dried up, it dies and rots.(3)
To apply the argument to moral philosophy: It is a mistake to attempt to validate common sense morality through some premises and arguments external to it, for those premises will always strike us as less obvious than that which they seek to validate. The principles of obligation that common sense accepts stand in no need of validation; certainly they stand less in need of justification than the sort of theories moral philosophers have tried to adduce in their support. Ethical theories, if there are any, ought rather to start from our intuitive principles of obligation and attempt to draw out their consequences.
1. H.A. Prichard, "Does Moral Philosophy Rest on a Mistake?" in Moral Obligation (Oxford: Clarendon, 1949).
2. Cf. G.E. Moore's list in "A Defense of Common Sense", Philosophical Papers II (p. 33).
3. Inquiry and Essays, Lehrer & Beanblossom, ed. (Indianapolis: Bobbs-Merrill, 1975) p. 7.