Ethical Egoism, Utilitarianism and the Fallacy of Pragmatic Inconsistency

J O N A T H A N H A R R I S O N

10 Halifax Road Cambridge CB2 3PY, U.K.

How art thou out of breath when thou hast breath To say to me that thou art out of breath?

William Shakespeare, Romeo and Juliet.

But when I tell him he hates flatterers, He says he does, being then most flattered.

William Shakespeare, Julius Caesar.

ABSTRACT: In this paper I shall consider the difficulty for Ethical Egoism, ~ Act Utili- tarianism and later what I shall call Cumulative Effect Utilitarianism, 2 that they both commit the fallacy of pragmatic inconsistency. I shall distinguish various forms of the fallacy of pragmatic inconsistency; in particular I shall distinguish between the fallacy of direct and indirect pragmatic inconsistency, and shall argue that though both Ethical Egoism and Act Utilitarianism probably commit both, Cumulative Effect Utilitarianism does not#

KEY WORDS: Act utilitarianism, cumulative effect utilitarianism, ethical egoism, prag- matically inconsistent, rationality

(I) THE FALLACY OF PRAGMATIC INCONSISTENCY AND ORDINARY INCONSISTENCY

Ber t rand Russe l l once wro te about a l ady who had recen t ly b e c o m e per- suaded o f the truth o f so l ips i sm, 4 and p ro fe s sed herse l f to be surpr i sed that more peop l e bes ides he r se l f d id not be l i eve the same. 5 The lady was com- mi t t ing the fa l lacy o f p r agma t i c incons is tency. 6 W h a t someone asserts when he or she is p r a g m a t i c a l l y incons i s t en t is not incons is ten t . There w o u l d be no incons i s t ency i n v o l v e d in this l a d y ' s be ing the only pe r son in the world . The " i n c o n s i s t e n c y " l ies in her b e h a v i n g as i f wha t she was asser t ing was fa lse , as one w o u l d be i f one p r o c l a i m e d to o thers what one took to be the fact that the un ive r se c o n t a i n e d no one but onesel f . P r a g m a t i c incons i s - t ency c o m m o n l y occurs when one is pu t t ing fo rward a ph i l o soph i ca l v iew wh ich is unneces sa r i l y pa radox ica l . T h o u g h the fact that it is p r a g m a t i c a l l y incons i s t en t to put a v i ew fo rwa rd does not show that i t is wrong , it o f ten ind ica tes that the pe r son pu t t ing it f o rwa rd does not r ea l ly be l i eve what he h i m s e l f says. Be r t r and R u s s e l l ' s l ady d id not r ea l ly be l i eve that she was

Argumentation 9: 595-609, 1995. © 1995 Kluwer Academic Publishers. Printed in the Netherlands.

596 JONATHAN HARRISON

the only person in the world, and was for this reason betrayed into acting as if what she herself was saying was false, which, indeed, it was.

Similarly, though there is nothing inconsistent about one's not existing, it is pragmatically inconsistent to assert that one does not exist; one is doing something that one could not do if what one said were true. And though there is nothing inconsistent in the assertion that all Cretans are liars - someone other than a Cretan could assert this very same thing without any problem - when a Cretan asserts that all Cretans are liars he is behaving in a way in which he could not behave if it were true, i.e., asserting some- thing true. This means that though it itself does not entail its own falsity, this statement, together with the fact that it is asserted by a Cretan, does entail that what the Cretan asserts is false.

There is no reason why a remark should not both be inconsistent and my making it pragmatically inconsistent. Were I to say that I both did and did not exist, what I said would be inconsistent, but my saying it would also be pragmatically inconsistent.

(115 EXAMPLES OF PRAGMATIC INCONSISTENCY

There are, of course, unphilosophical examples of pragmatic inconsistency, such as the original Catch 22, applying to leave the American Air Force on the grounds that one is insane, which is supposed to demonstrate that one is sane. 7 But pragmatic inconsistency is a way of scoring own goals to which philosophers seem especially prone. They assert as truths propositions to the effect that all propositions are false. 8 They accurately reveal the fact that nothing can be accurately revealed (Richard Rorty in Consequences of Pragmatism, University of Minneapolis Press, Minnea- polis, 1982). They attempt to prove that nothing can be proved, and claim to know that there is no such thing as knowledge. 9 They obstinately refuse to revise their opinion that all opinions are subject to revision (Willard van Orman Quine, "Two Dogmas of Empiricism", in From a Logical Point of View, Harvard University Press, Cambridge, Massachusetts, 1953). Many are certain that nothing is certain. 1° Some believe that belief is impossible (Stephen Stich, in From Psychology to Cognitive Science The Case Against Belief, Clarendon Library of Logic and Philosophy, Oxford, 1975); others think that there is no such thing as thought. 11 They authorise the transla- tion of learned philosophical works that claim that translation is impos- sible, ~2 and write with the greatest lucidity words that mean that there is no such thing as meaning) 3 They eloquently describe the ineffable. ~4 Some advance good inductive evidence for thinking that science may proceed counter inductively (Paul Feyerabend in Against Method, outline of an anarchist theory of knowledge, Verso, London, 1079); and others give reasons for thinking that though the past is no guide to the future, there is every reason to suppose that a hypothesis that has once been refuted will

ETHICAL EGOISM 597

go on being refuted in years to come (Karl Popper, op. cit.). They provide inductive proofs that induction is untrustworthy, via the inductive gener- alisations that all inductive generalisations eventually break down. ~5 Some (e.g., I think Bohr) produce causal explanations of the fact that all causal explanation is impossible. Some write books proclaiming that it is absolutely true that all truth is relative. ~6 Others ~7 argue as if it were unequivocally true that there is a third possibility to truth and falsehood. They offer a priori proofs that there is no such thing as a priori proof (John Stuart Mill, in Book II of A System of Logic, Longmans, Green and Co. Ltd., London, 1947). They maintain as an analytic proposition that there is no such thing as analyticity (Quine, in "Two Dogmas of Empiricism", in From a Logical Point of View); and that there is, of necessity, no such thing as necessary truth. One philosopher has raised the question whether "plus" might not mean plus, but quus, doing which is pragmatically incon- sistent in that he would have had to know perfectly well that he was referring to plus by "plus" in order to raise the question whether plus it was that might mean quus (Saul Kripke, in Wittgenstein on Rules and Private Language, Blackwell, Oxford, 1982). They build great edifices on the premise that foundationalism is false. They hold it as a scientific fact that not all facts are scientific, TM and as a truth of philosophy that there is only scientific truth. 19 They propound as a philosophical truth that there are no philosophical truths, but only techniques for dissolving problems (Wittgenstein in Philosophical Investigations, translated by G.E.M. Anscombel Basil Blackwell, Oxford, 1953). They take immense trouble stating convincing arguments for what they take to be the truth that it is not truth, but continuing the conversation, that matters (See Richard Rorty in Contingency, Irony and Solidarity, Cambridge University Press, Cambridge, 1989). They argue with deep sincerity that it is impor- tant to play with ideas, rather than believe them, and then, though one would have thought they wanted to convince us, defeat this end by telling us that when they said this, they were only playing (Richard Rorty, loc. cit.) They assert that all intellectual remarks are ironic, and expect this remark to be taken as being literally true (Richard Rorty in Contingency, Irony and Solidarity and Hans-Georg Gadamer in Truth and Method, New York, 1975. Having themselves transcended the limits of thought, they feel themselves to be in the best possible position to tell us that thought has limits. Sometimes they provide logical proofs of the validity of logic (Descartes in the Meditations uses logic to show that God cannot deceive us about logic), thus presupposing the truth of what they are trying to prove. At other times they provide logical proofs of logic's invalidity, thus pre- supposing the truth of what they are trying to disprove (Graham Priest, in In Contradiction; A Study of the Paraconsistent, Nijhoff International Philosophy Series, Dordretch, 1987). They are worried in case it is not con- sistent to hold that some inconsistencies can be true (Graham Priest, loc. cit.). Without having established the whole truth about anything, they

598 JONATHAN HARRISON

announce that nothing short of the whole truth about everything is worthy of belief, and claim to know the partial truth that there is no such thing as partial truth (F.H.Bradley in Appearance and Reality, Oxford University Press, Oxford, 1893). Without themselves understanding the whole of any one language, they state that it is impossible to understand anything in isolation from the whole of language, and appear to think they understand what they themselves say, isolated though their remark is (See Holism; a Shopper's Guide, by Jerry Fodor and Ernest Lepore, Blackwell, Oxford, 1992). They are convinced of the useless truth that only what is useful can be true. 2° They try to prove that in reality nothing is real, and that it is no illusion that everything is illusory. They provide rational arguments for the desirability of faith, but nevertheless put their faith in reason (Many theologians come close to doing this). They prove that all persons are really one and the same, but nevertheless allow their friends to buy them drinks (Arnold Zuboff, in a paper read to the Cambridge University Moral Sciences Club). From the centre of the self they lament that the self has no centre. As a result of careful reflection upon the function of material sense organs they conclude that there is no such thing as matter (Hume in A Treatise of Human Nature, edited by L.A.Selby- Bigge, Oxford University Press, Oxford, 1888, Book I, Part IV, "Of Scepticism with Regard to the Senses"). They argue most rationally in favour of irrationalism (Richard Rorty, in Philosophy and the Mirror of Nature). They provide their readers with the most moral reasons for believing that it is wicked to be good. 21 They altruistically proclaim that all men should seek their own interest, and not bother about the interest of the person making this proclamation. 22 They think that there are no opinions that may not be legitimately expressed, but try to prevent people from expressing the opinion that there are some such opinions. They immodestly proclaim their modesty (everybody), openly advocate the advantages of deception, and hold that not forgiving one's neighbours deserves severe punishment in this world and the next (almost everybody, from time to time). They listen to the story of the Pharisee who thanked the Lord that he was not as other men are, and thank the Lord that they are not as that Pharisee. They insist that anarchism should be imposed by force upon the protagonists of government. 23 Their concern for the welfare of man is such as to deplore universally doing good because of the harm it would do (Bernard Williams in Morality: an Introduction to Ethics, Harper and Row, New York, Evanston, San Francisco and London, and Derek Parfit in Reasons and Persons, Oxford University Press, 1984). They write as if moderation were a virtue that one could not have to excess. 24 Though, like honest men, they say nothing of which they are not sure, they are, on the basis of past experience, rightly convinced that most of what they say must be wrong. 2s I myself am one hundred per cent serious in saying that I am only ninety five per cent serious about everything I say. 26

ETHICAL EGOISM 599

(IIt) THE FALLACY OF PRAGMATIC INCONSISTENCY AND ETHICAL EGOISM

(A) The alleged "overt" pragmatic inconsistency of egoism There is a well known objection to Ethical Egoism to the effect that Egoistical Egoism commits the fallacy of pragmatic inconsistency, because it entails that one ought not to say out loud that one is an Egoistical Egoist or advocate Egoistical Egoism to others. It is in everyone's interest to be egoistical himself, but for everybody else to be altruistic. In telling other people that it is their only duty to pursue their own interest, one is behaving in a way that is contrary to one's interest.

I call this the fallacy of "overt pragmatic inconsistency" because the inconsistency arises only if you assert Egoistical Egoism out loud; there is not necessarily any inconsistency if you simply think it quietly to yourself. When there is an inconsistency in believing a statement or theory, as opposed to asserting it, I shall describe the theory as committing the fallacy of "covert pragmatic inconsistency". It would be pragmatically inconsistent both to think and to say that one did not exist. But it would be pragmatically inconsistent to say that it is wrong to talk, but not incon- sistent to think it. And it might be pragmatically inconsistent to be conceited enough to believe that one was humble, if one thought that humility demanded that one underestimate oneself (assuming, of course, that one took humility to be a virtue), but not pragmatically inconsistent, but only mildly comic, to say that one was humble without believing it. Most remarks, of course, are neither overtly nor covertly pragmatically inconsistent.

There are various other ways of committing the fallacy of pragmatic inconsistency, most of which I have not space to discuss here. Here I shall distinguish only between theoretical and practical pragmatical pragmatic inconsistency. (i) Asserting something is theoretically pragmatically incon- sistent if asserting it shows that what is asserted is false, as the remark that Juliet is complaining about in the quotation that heads this article, or if your asserting it shows that you do not yourself believe it, as does trying to get converts to solipsism. (ii) Asserting something is practically prag- matically inconsistent if (1) in asserting 2v it one is disobeying the injunc- tion one is propounding (if it is in the imperative mood) or (2) acting in a way that would be wrong if the moral judgement asserted were true. Putting up a bill saying "Post no bills", would be an example of the first, and writing in green ink that it was wrong to write in green ink an example of the second. 28

However, though asserting Ethical Egoism commits the fallacy of prag- matic inconsistency, showing that it was to commit the fallacy of pragmatic inconsistency to assert it would not show that Egoistic Egoism was false, as would showing that ethical Egoism was straightforwardly inconsistent. (The fact that it is pragmatically inconsistent to shout in libraries "It is

600 JONATHAN HARRISON

wrong to shout in libraries" does not mean that the statement that it is wrong to shout in libraries is false. Indeed, it is true.)

It is not irrational to do things that are pragmatically inconsistent. For example, it would be rational for me to shout in a library that it was wrong (or impossible) to shout in a library, if a lunatic threatened me with death if I did not, and it could be commendable as well as rational so to shout if another more benevolent lunatic offered to give £ 1 , 0 0 0 to children in Yugoslavia if I did. It would not be irrational for a philosopher to make pragmatically inconsistent assertions if he could get enough other philoso- phers to believe him. It could be rational to acquire covertly pragmatically inconsistent beliefs if I had any command over my beliefs.

The senses in which the proposition asserted and the action of asserting it are inconsistent are different. The proposition asserted is inconsistent in a sense that entails that it cannot be true. The action of asserting it is incon- sistent in that (at least) either one could not do it if the proposition asserted were true, that one's asserting it would entail or provide evidence for the falsity of what was asserted, or that it would be wrong or irrational or inconsistent or misleading to assert (or enjoin) i t if what was asserted were true.

(B) The alleged "covert" pragmatic inconsistency of ethical egoism In Reasons and Persons Derek Parfit discusses a more sophisticated version of the difficulty with Egoistic Egoism mentioned above, viz, the difficulty that it commits what I have called the fallacy of covert pragmatic incon- sistency. Parfit argues (if I may simplify a little) that Egoism has the property of implying that it is wrong even to believe Egoism, not just that it is wrong to assert it. (It may even have the slightly different property of implying that it is right to believe some theory incompatible with itself.)

The reason for this, according to Parfit, is that Egoism - or strictly speaking what Parfit calls "The Self-Interest Theory, S" - implies that it is rational 29 to break promises when one can get away with it. But, Parfit thinks, an Egoist might be such a bad actor that everyone would know when he did not intend to keep a promise when he made one. In this case no one would trust him or believe him when he made a promise. Hence he could get the reputation for being trustworthy only if he actually kept his promises, which promises he could not break if he acted upon Egoism; Egoism dictates that he should break some of them. In that case, the rational thing for a man to do would be to get himself to believe that Egoism was false, for then he would not believe that it was rational for him to break his promises. He would then, or so Parfit supposes, keep them, and so get the reputation for trustworthiness he needed.

I myself cannot see that Egoism does imply that one ought not to believe Egoism, whether this implication is unfortunate for Egoism or not. (I do not think that it is; nor I think, does Parfit.) Egoism does not imply that one must break all one's promises. Egoism dictates only that one should

ETHICAL EGOISM 601

break certain promises if one can get away with breaking them. Hence there is no need for an egoist to go to the lengths of persuading himself that Egoism is false, in order to prevent himself from breaking promises which he will be discovered breaking. Egoism implies that one should not break such promises. Egoism does not imply that one ought to break promises even if one will be found out. If it did, there must be simpler ways of stopping oneself from making promises which will be found out than getting oneself to disbelieve Egoism, if one does believe it. And I myself have very little control over my beliefs, and would not at will be able to produce a belief that Egoism was false, even if I did not believe that it was false already.

There is another reason for thinking that Egoism does not dictate that it is necessary to eliminate one's unfortunate tendency to be detected when making lying promises. People who could not dissemble well enough to get away with making lying promises would in fact be very valuable. If they knew they could not get away with making lying promises, they would not make them. They would therefore be much sought after for positions of trust, which fact would be very useful to them. A man who blushed violently whenever he made a promise he did not intend to keep would be universally trusted when he made a promise and did not blush. He could promise that other people were sincere when they made promises. Hence such a man could advertise his services in the local newspapers, and make a lot of money. People who had already lost their reputations for honesty, but needed to be believed, could hire the blusher's services. Hence Egoism would dictate that one become a man who could not make lying promises without blushing, rather than dictate that a man who already had this pecu- liarity should stop himself from being an egoist, as Parfit supposes. At the least, it would dictate this so long as not everyone allowed Egoism to make him become a man who blushed when he made a lying promise (supposing a potion which produced this propensity were for sale). In that case, the blushers would become a drug on the market, and their services without any value.

(IV) THE FALLACY OF PRAGMATIC INCONSISTENCY AND ACT UTILITARIANISM

(A) The alleged "overt" pragmatic inconsistency of act utilitarianism To assert Utilitarianism is not always to commit the fallacy of overt prag- matic inconsistency, as it is always overtly pragmatically inconsistent to say that one does not exist) ° It would only sometimes be inconducive to the greatest good to say that one ought to seek the greatest good. I would hesitate to make the remark to children, and young men whom Aristotle thought should not engage in moral philosophy. But I feel that I can say this to my present readers without there being a great danger of their

602 JONATHAN HARRISON

abusing it, if only because most of them will know about Utilitarianism already, and so have overcome the dire temptations to which those who believe in Utilitarianism are subject. If they have yielded to these tempta- tions, they have been consistent enough Utilitarians not to admit that they have.

The Utilitarian Sidgwick has himself been accused of committing the fallacy of pragmatic inconsistency in propounding the fact that Utili- tarianism itself demanded that it ought not to be propounded to everybody. 31 But though to propound to all people that Utilitarianism should not be pro- pounded to all people, but only be propounded to some, is pragmatically inconsistent, to propound to some that it ought only to be propounded to some is not - or not so long as the ones to whom it is propounded are just those who will not abuse it should they come to believe it.

(B) The alleged "covert" pragmatical inconsistency of act utilitarianism I suspect that that believing a proposition, as opposed to acting on it, 32 has bad effects is not a reason for thinking that it is false. For example, I suspect that believing one of Parfit 's contentions, that the boundaries between I and thou break down, so that it is unclear whether the person who benefits from my taking thought to the morrow is myself or someone else, would have bad effects if it were generally believed. Men would take less care of them- selves, not knowing whether it would be they or some other person who would benefit from their doing so, and this would be a bad thing; it was because men did not take care of land that was held in common that it was sometimes "privatised". But this does not show that Parfit 's view is false, though it may be some small reason for putting Reason and Persons on the Index of Forbidden Books (which reason is greatly outweighed by the book 's merits). And if Pascal 's God saved only those who acquired the irrational belief that there was a God, it would be rational to produce this irrational belief, and it would also (ex hypothesi) be true.

However the argument against Utilitarianism is not simply that it is a theory bel ief in which has bad effects, but that Utilitarianism is false because it would have bad effects to believe it, when it itself enjoins the doing of what has good effects. But the fact, that Utilitarianism sometimes, but not always, enjoins that one should disbelieve Utilitarianism, would have no tendency to show that it was false. An interesting form of the argument from bad effects is that Utilitarianism should be rejected because it would be biologically harmful for mankind to be Utilitarians. But then, it might be biologically harmful for men to believe that their moral beliefs are the useful products of evolution - because in that case they might pay less attention to them - but this would not show that their moral beliefs were not the useful products of evolution. (It could be biologically advan- tageous to have beliefs that Utilitarianism would in fact justify, without knowing that it did, although biologically disadvantageous to have them for this reason.) 33

ETHICAL EGOISM 603

(V) THE FALLACY OF "INDIRECT" PRAGMATIC INCONSISTENCY 34

Anyone seeking his own happiness will often, though by no means always, have to perform actions which will be in his interest only so long as others do not do the same. For example, he may promote his own interest by breaking promises, but if everyone were to break similar promises, the result would be contrary to his interest. This is because he benefits from the institution of promising, which would not flourish if everyone did what he did. (This is additional to the fact that it is not in a person's interest to state out loud that offlers ought to act in their interests. It would not be in the interest of an egoist to persuade others to be egoists, whether or not he was successful in getting all egoists to act on egoism.)

The same is true of seeking the greatest good. I might do more good than harm if I gave the money that the Inland Revenue think I owe them to the poor, but, on account of the cumulative ill effect of such actions, if everyone were to do this, government, which some consider a necessary institution, could not exist, and harm would result. 35 Hence if everyone were to do something which did more good than harm, the result would be more harm than good.

It follows from this that recommending Act Utilitarianism commits what I shall call the fallacy of "indirect" pragmatic inconsistency. An action of recommending something is directly pragmatically inconsistent if in per- forming this action the end which it asks one to seek will not be brought about by one 's recommending that it be brought about. The action of recommending a principle is indirectly pragmatically inconsistent if the end the principle one is recommending enjoins one to seek would not be brought about if everybody were to do what one is recommending. Whether recommending a principle is (directly) pragmatically inconsistent depends (1) upon the principle and (2) upon the consequences of recommending i t ] 6 Whether or not it is indirectly pragmatically inconsistent depends (1) upon the principle and (2) upon what would be the consequences if everyone were to act on one's recommendation, whether everyone actually does act on it or not. 37

There will thus be a "contradict ion in the will ''38 of anyone recom- mending Act Utilitarianism, because, if his object is to secure the maximum amount of good, he must hope that his recommendation will not be uni- versally followed, for though its being followed by some but not by others will maximise good, its being followed by all will do harm. Hence in asking others not to pay their taxes in order to help the poor, my action is indi- rectly pragmatically inconsistent in that I am making a recommendation that would defeat my end, of producing good, should it be universally adopted.

Though it is indirectly pragmatically inconsistent to recommend Act Utilitarianism, it cannot be indirectly pragmatically inconsistent to recom- mend Cumulative Effect Utilitarianism. Cumulative Effect Utilitarianism

604 JONATHAN HARRISON

entails that the actions one is recommending are wrong if harm would result in their being performed on all occasions by everybody. Hence, whereas Act Utilitarianism demands that we perform actions that it would be to commit the fallacy of indirect pragmatic inconsistency to recommend, Cumulative Effect Utilitarianism forbids such actions.

(VI) THE FALLACY OF PRAGMATIC INCONSISTENCY AND THE LOGICAL CONSISTENCY OF UTILITARIANISM

I suspect that some philosophers confuse the fact that Utilitarianism implies that one should (sometimes) not recommend Utilitarianism with the fact that it both implies that one should and that one should not recommend Utilitarianism (((p implies q) and (p implies not-q)) implies not-q). This would be a fatal difficulty for Utilitarianism. But Utilitarianism does not imply this.

There would also be a difficulty if Utilitarianism implied its own falsity, for in that case one could use the principle that a proposition that implies its own falsity must be false ((p implies not-p) implies not-p) to reject Utilitarianism. But Utilitarianism's implying that Utilitarianism should not be believed or asserted out loud - which would mean that one was com- mitting the fallacy of pragmatical inconsistency if one asserted it - is not the same thing as its implying its own falsity. (The same is true of Ethical Egoism.)

As we have seen, it has been argued that it would have harmful consequences to recommend Utilitarianism as if this were a reason for rejecting Utilitarianism (e.g., by Bernard Williams, loc. cit.). But to reject Utilitarianism, because it would have harmful consequences to recommend it, is to use an Utilitarian argument against Utilitarianism. It is therefore gratifying to be able to say that, if this were so, using this argument against Utilitarianism would itself be pragmatically inconsistent, for the fact that it would have harmful consequences to recommend Utilitarianism could not be a conclusive argument against Utilitarianism unless Utilitarianism were true; hence the person using it is presupposing the falsity of what he himself is arguing for.

In any case, to use the argument that Utilitarianism must be rejected because of the harm that would be done if everybody were to act on it, is not, as can be seen if it is spelt out more fully, to use an Utilitarian argu- ment against Utilitarianism so much as to use a Cumulative Effect Utilitarian argument against Act Utilitarianism. Hence this argument does not show that its user is assuming the truth of Utilitarianism in order to reject Utilitarianism. Utilitarianism, indeed, does not prohibit felicific actions which are such that harm would result if everyone were to do them; the trouble with Act Utilitarianism is that it enjoins them. Williams's argument is in fact a Cumulative Effect Utilitarian argument against Act

ETHICAL EGOISM 605

Utilitarianism, for it rejects Act Utilitarianism because it would have bad consequences if everybody were to act on it. Hence this objection to Act Utilitarianism is sound only if Cumulative Effect Utilitarianism is true. Since it is true, Act Utilitarianism must be mistaken.

(VII) CONCLUSION

Utilitarianism commits the fallacy of pragmatic inconsistency only to a limited extent, in that it itself would demand that s o m e t i m e s it should not be propounded to s o m e people. Act Utilitarianism, it is true, is °'indirectly pragmatically inconsistent", in that in recommending it we would be recommending, on the grounds that it would do good, something that would do harm if our recommendation were universally adopted. But Cumulative Effect Utilitarianism is not even indirectly pragmatically inconsistent.

Whether Utilitarianism commits the "fallacy" of pragmatic inconsistency or not, it is only comparatively rare that the fact that asserting something is pragmatically inconsistent means that it is false if it is asserted. When I say that I do not exist, and in similar cases such as shouting through the letter box that one is not at home, which one could do only if one were at home, asserting something shows that it is false. But there is a surpris- ingly wide spectrum of cases, to which I hope one day to give more atten- tion. And even in the case when my asserting a proposition demonstrates the falsity of what I assert, this fact does not mean that what I assert is con- tradictory. To put the matter in what the English call an Irish way, had ! not asserted that I did not exist, that I did not exist might have been true.

It is more usual for the fact, that asserting something commits the fallacy of pragmatic inconsistency, to indicate that the person asserting it does not genuinely believe what he himself is saying. This takes away any authority that would be given to this proposit ion by the fact that he is asserting it, for his asserting it would give authority to what he asserts only if he himself believed what he asserted.

The fact that asserting something is pragmatically inconsistent also tends to show that what the person making the assertion is saying is something that it would be foolish to believe. For what he is asserting is likely to be something that one would put forward only in the context of philosophical discussion, where it is quite common to maintain extraordinary proposi- tions on which it is quite impossible sensibly to act, even to the limited extent of performing the act of asserting them. The pragmatic inconsistency of asserting a philosophical view tends to show that the philosopher propounding the view has forgotten G.E.Moore 's contention - I put in it my own way - that when, such is the poor record of philosophy on such matters, a philosophical theory is contrary to common sense, it is much more likely that the arguments for it are invalid, or the premises on which they are based false, than that common sense is mistaken. The theorist's

606 JONATHAN HARRISON

common sense reasserts itself in the very act of asserting the view to which it is contrary.

In any case, it might be considered Philistine to let a concern for morality override one's dedication to truth, which is what one would be doing if one refused to enlighten other people concerning utilitarianism on the grounds that such enlightenment would be bad for them. And who knows but that a dedication to truth at the expense of goodness, consistently pursued, might do more good in the long run than a dedication to goodness at the expense of truth? If that is the case, to recommend preferring goodness to truth, as do those who object to recommending utilitarianism on account of the harm done by preferring truth to goodness, would itself be to commit the fallacy of pragmatic inconsistency.

ACKNOWLEDGMENTS

I am grateful to my friend David Rees for numerous helpful suggestions, almost all of which I have incorporated in the text. The reader is referred to Chapter 4 of J.A.Passmore's Philosophical Reasoning, Duckworth, London, 1961.

NOTES

Ethical hedonism is the view that the only thing anyone ought to do is to maximise his own happiness. 2 Act Utilitarianism is the view that the only thing anyone ever ought to do is to maximise good. (It is sometimes taken for granted by Utilitarians that all good is human good.) If one also thinks that happiness is the only good, one is a Hedonistic Utilitarian. Act Utilitarianism is contrasted with Rule Utilitarianism, the view that one should act on useful rules, and Cumulative Effect Utilitarianism, the view that one should act in a manner that would have good consequences if everybody were to do the same. The fact that an act performed by almost everyone will have different consequences from the same act performed by only a few people is due to the cumulative good or ill effect of repetitions of the action. The reader is referred to "Utilitarianism, Universalisation and our Duty to be Just" in Volume I of my Ethical Essays, Avebury Academic Publications, Aldershot, 1993. 3 I cannot make up my mind whether Rule Utilitarianism commits this fallacy or not. 4 Solipsism is the view that one is oneself the only person in the universe, or that the universe contains nothing but oneself and one's mental states. 5 The mistake is not confined to Bertrand Russell 's lady. In Paul Feyerabend's Three Dialogues on Knowledge, Basil Blackwell, 1991, I find the following dialogue, between no less than three people, concerning the view that the world contains only one: Arnold: You mean solipsism, the idea that you alone exist and everything else is just a

colourful part of your personality? Maureen: Yes, but that cannot possibly the whole truth. Leslie: Are you sure? 6 I have not space to discuss the relationship between conversational implication and prag- matic inconsistency. There must be some relation sometimes, because saying "P, but I don't believe that p" is pragmatically inconsistent. If asserting p did not conversationally imply

ETHICAL EGOISM 607

that one believed it, there would be no inconsistency. However, the pragmatic inconsistency of saying that one exists does not involve conversational implication. 7 Pragmatic inconsistency seems to be somehow involved in Groucho Marx's telegram "Please accept my resignation. I do not want to belong to any club that will accept me as a member". Perhaps applying to join a club that you will not want to join if it accepts you is pragmatically inconsistent. Perhaps it is like trying to lock the only keys of one's safe in one's safe for safe keeping. A British minister has recently caused a furore by saying, sin- cerely, that there are occasions on which a minister may lie to the House of Commons. If this were one of those occasions, he would presumably have said that there were no such occasions. Is it ironic that it was the minister for open government who was defending lying, or not ironic because he was openly defending lying? 8 Epimenides the Cretan. "One of themselves, a prophet of their own, said, Cretans are always liars, evil beasts, idle gluttons. This testimony is true." Titus 1. 12 and 13. Even though this remark is pragmatically contradictory, and not contradictory, it constitutes a difficulty for those who believe that everything in the bible is true. For Paul not only reports the remark, but endorses it, i.e. he says that a Cretan has truly said that all Cretans are liars. (Philosophers take this remark more literally than it was intended to be by Paul.) I do not know of a philosopher who has maintained that all propositions are false, but there have been philosophers (for example, J.L. Mackie, in Ethics, inventing Right and Wrong, Penguin Nooks, Harmondsworth, 1977, who have held that all moral judgements are false. It would be pragmatically inconsistent to maintain this if it commits you to a moral judgement, say the moral judgement that everything is permissible. 9 Many sceptics, for example, Peter Unger in Ignorance, a Case for Scepticism, Clarendon Library of Logic and Philosophy, Oxford, 1975. Unger sometimes less drastically claims to know the interesting truth that there are no interesting truths about which one should not be sceptical. to Among many, Karl Popper in The Logic of Scientific Discovery, Hutchinson, London, 1959. Those more charitable than myself would say that Popper thought that it was only scientific generalisations that were uncertain. But what about the generalisation that scien- tific generalisations are uncertain? ~t Physicalists who hold the eliminative theory of mental events. The reader is referred to Richard Rorty in Philosophy and the Mirror of Nature, Blackwell, Oxford, 1980. ~2 In fairness I should point out that they are only pragmatically contradicting themselves if they believe what Quine has said about translation, which some of them may be wise enough not to do. See Quine's Word and Object, The Technology Press of the Massachusetts Institute of Technology, Cambridge Massachusetts, and John Wiley and Sons, Inc, New York and London, 1960. For further locations I refer to Robert Kirk's Translation Determined, Oxford University Press, Oxford, 1986. I have much sympathy with the claim that Philosophy says what everybody knows in language which nobody can understand. But if no one can understand it, how can anyone know that it says what everyone knows? ~3 Quine, Word and Object, loc. cit. In Peter Unger's Ignorance: The Case for Scepticism there occurs the meaningful sentence "I even have some reservations towards thinking that there actually are any sentences . . . . " 14 It is common to level this accusation at mystics, who commonly say that the object of mystical experience cannot be described, and then go on to describe it. It would be chari- table, but probably not right, to suppose that they mean that mystical experience cannot satisfactorily be described. 15 R.B.Braithwaite (in Scientific Explanation, Cambridge University Press, 1953) and D.H.Mellor (in The Warrant of Induction; an Inaugural Lecture, C.U.P., 1988) have both (intentionally, unlike most of the other philosophers I have mentioned) produced an induc- tive proof of the validity of induction). ~6 Protagoras said "Man is the measure of all things, of those that are that they are, and of those that are not that they are not." This remark is pragmatically contradictory, in that it describes what it is that is relative (e.g, "those that are") in a way that could only be correct

608 JONATHAN HARRISON

if absolutism was correct. See Plato's Theaetetus, translated by John McDowell, in the Clarendon Plato Series, Oxford, 1977. ~7 Perhaps Aristotle, in Chapter 9 of De Interpretatione, translated by J.L. Ackrill in Volume I of The Complete Works of Aristotle, the revised Oxford translation, Princeton University Press, Princeton, 1982. t8 I do not know that anyone has ever actually done this. 19 Anyone who thinks that there are no truths except those established by empirical science is bound to make this mistake, i f - as I think it is - that all truth is scientific is not a scientific truth. That it would be a philosophical truth if it were true that there were no philosophical truths has been described as "the paradox of philosophy" by W.H.F. Barnes in The Philosophical Predicament, London, 1950. 20 It is a pragmatist view that the true is identical with what it is useful to believe, but it could be argued that it is not useful to believe that the true is nothing more than the useful. ~ See The Moral Society, its Structure and Effects (popularly known as "To Hell With Morality" by Ian Hinckfuss, Departments of Philosophy, Australian National University, 1987. In conversation Hinckfuss has conceded that "a morally inclined reader may see his reasons as moral reasons, even though he (Hinckfuss) sees these reasons as merely altruistic or practical". Friedrich Paulsen (quoted in Marcus G. Singer's Morals and Values, Charles Scribner's Sons, 1977) puts in the mouth of an imaginary moral nihilist the words " . . . he is contemptible [my italics) who has not the courage to do what he pleases, but lets all sorts of imaginary scruples defraud him of the pleasures of the moment." 22 All known egoistical hedonists. There may be many unknown egoistical hedonists who have been wise enough not to proclaim that they are, but since they have not proclaimed that they are, we do not know whether there are any or not. 23 The anarchist Bakunin looked forward to a time when those in superior positions would occupy inferior ones. See M.A. Bakunin, Statism and Society, translated and edited by Marshall S. Shatz, Cambridge University Press, Cambridge, England, 1990. 24 Arguably, Aristotle, in Book II of the Nicomachean Ethics, in The Works of Aristotle Translated into English, translated by W.D. Ross, Oxford University Press, Oxford, 1925. 25 I myself am in pessimistic moods inclined to combine the belief that everything that philosophers say is false with the incompatible belief that, since by now they must have exhausted all the available alternatives, some of the things they say must be true. Unfor- tunately I do not know which.

It is often alleged to be believing a contradiction (the "Preface Fallacy") to believe some such thing as that not everything one has asserted is true, but in fact there is nothing wrong with believing this. Though the propositions one believes are contradictory, there is no absur- dity involved in believing them, so long as, like a sensible person, one is only say 99 per cent certain of each of 100 things, of which one is 100 per cent certain, that they cannot all be true. People who worry about the Preface Fallacy are also inclined to confuse believing that all of a set of beliefs are true and at the same time believing of a given one of them that it is false, with believing that all of a set of beliefs are true, and that one or other of them (unspecified) must be false. The reader is referred to "Is "You can't fool all of the people all of the time" an analytic proposition?", by D.A. Rees, in Mind Volume LX, 1951, pp. 97-99. 26 The habit of pragmatically contradicting oneself is not confined to philosophers. Yeats (The Circus Animals' Desertion) and Coleridge (Dejection: an Ode) have both written poems lamenting their inability to write poems. Not all the above are clear cases of pragmatic incon- sistency. Some of them involve type fallacies. But that they involve type fallacies does not exclude them from being also pragmatic inconsistencies, as pragmatic inconsistency can occur together with a type fallacy. For example, if I say that all propositions are false, I am guilty of a pragmatic inconsistency in that I am doing something that I ought not to do if I think that all propositions are false, i.e., asserting one of them, and also (rightly or wrongly) saying something that applies to itself, since the proposition that all propositions are false is a proposition. I do not wish to suggest that every time I assert a proposition that applies

ETHICAL EGOISM 609

to itself, this is a fallacy. For example, the proposition that every proposition implies itself, implies itself. 27 Sometimes it is only writing it that is overtly pragmatically inconsistent, as it is prag- matically inconsistent to write, but not to say, "It is wrong to write." 28 Clement Freud, in a book full of deliberate pragmatic inconsistencies, that once intrigued my children, but which I am now unable to locate, tells the story of a boy who removes a message tied to the leg of a carrier pigeon. On it is written the words "Do not tie messages to my leg. I am not a carrier pigeon." 29 I personally think that this is a misstatement of egoism. Egoism is the view that it is right to seek only one ' s own interest. And so far from this theory being the preponderant theory for many years, it has been held by hardly anybody, and has had very little influ- ence. I think Aristotle assumed that one would perform one 's ordinary duties before seeking eudaemonia, and gave something like an intuitionist account of these duties. He did not think is was morally right to seek one ' s eudaemonia. 30 One would have supposed that it would not be pragmatically inconsistent for other people to say that I do not exist, however. But anyone saying "Harrison does not exist", if he believed that Harrison did not exist, would be using the proper name "Harrison" with a view to refer- ring to Harrison, when this name could be used to refer to Harrison only by someone believing that Harrison does exist. 3~ Jonathan Glover (in Utilitarianism and its Critics, The Macmillan Publishing Company, New York, and Collier Macmillan, London, 1990, has suggested that Sidgwick is also guilty of a kind of pragmatic inconsistency in not mentioning the fact, that Utilitarianism some- times ought not to be propagated, until the end of The Methods of Ethics, by which time some of the damage to his readers by its propagation it will have been done. 32 An example of the consequence of believing something as opposed to acting on it, is my hair 's standing on end if I think that what I see is a ghost. Over a long enough period, I am more likely to secure my ends if I act on true beliefs rather than on false ones. 33 These matter are discussed more fully in "Some Reflections on the Ethics of Knowledge and Belief", in Volume II of my Ethical Essays, Avebury Academic Publications, Aldershot and Brookfield, 1993, and in "Sociobiology and Ethics", in Volume III (1995). 34 The adage "It is the last straw that breaks the camel ' s back" is simply not true. It is really rather obvious bad logic apart, that no act of straw dropping by itself ever does the camel any harm. The camel is broken by the collective dropping of many straws which dropped individually are quite harmless whether other straws are dropped or not. The mistake about which I am complaining is made by David Lyons, among others, in The Forms and Limits of Utilitarianism, Oxford University Press, Oxford, 1965. 35 There are a large number of such cases listed in Derek Parfi t 's Reason and Persons, loc. cit., though, so far as I know, he himself neither realises that the actions that puzzle him would be forbidden by Cumulative Effect Utilitarianism, nor how common such cases, of which he gives imaginary examples, are in everyday life. 36 I say that whether or not the action of recommending a principle is pragmatically incon- sistent depends both on the principle and the consequences of the action, because the fact that recommending a principle, that did not itself enjoin doing good, did harm, would not constitute a pragmatic inconsistency. 37 1 cannot make my mind whether or not recommending Ethical Egoism would be indi- rectly pragmatically inconsistent as well as pragmatically inconsistent. It is true that one person seeking my happiness might perform actions which would not be conducive to my happiness if they were performed by others besides himself. But ethical egoism does not rec- ommend that everyone seek my happiness but that every one seek his own happiness. In any case, it seems to be inconsistent to recommend that others seek their happiness, when their doing so is detrimental to one ' s own, whether the adverse effects on me of their seeking their happiness are cumulative or not. 38 The expression "contradiction in the will" is Kant 's . See his Fundamental Principles of the Metaphysics of Morals, of which there are many translations.