Canadian Journal of Philosophy - 24HourAnswers...Canadian Journal of Philosophy Two Kinds of Moral Reasoning: Ethical Egoism as a Moral Theory Author(s): Jesse Kalin Source: Canadian

Canadian Journal of Philosophy

Two Kinds of Moral Reasoning: Ethical Egoism as a Moral TheoryAuthor(s): Jesse KalinSource: Canadian Journal of Philosophy, Vol. 5, No. 3 (Nov., 1975), pp. 323-356Published by: Canadian Journal of PhilosophyStable URL: http://www.jstor.org/stable/40230576 .

Accessed: 13/06/2014 10:33

Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at .http://www.jstor.org/page/info/about/policies/terms.jsp

.JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range ofcontent in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new formsof scholarship. For more information about JSTOR, please contact [email protected].

.

Canadian Journal of Philosophy is collaborating with JSTOR to digitize, preserve and extend access toCanadian Journal of Philosophy.

http://www.jstor.org

This content downloaded from 141.161.91.14 on Fri, 13 Jun 2014 10:33:08 AMAll use subject to JSTOR Terms and Conditions

http://www.jstor.org/action/showPublisher?publisherCode=cjp

http://www.jstor.org/stable/40230576?origin=JSTOR-pdf

http://www.jstor.org/page/info/about/policies/terms.jsp


CANADIAN JOURNAL OF PHILOSOPHY Volume V, Number 3, November 1975

Two Kinds of Moral Reasoning: Ethical Egoism as A Moral Theory

JESSE KALIN, Vassar College

Ethical egoism, when summarized into a single ethical principle, is the position that a person ought, all things considered, to do an action if and only if that action is in his overall self-interest. The criticisms standardly advanced against this view try to show either that it is subject to some fatal logical flaw or else that, even if logically coherent, it can give no account of the basic parts of morality. Both these objections are mistaken, however, and it is the point of this paper to make this clear. Central to my argument is the distinction drawn in Section 1 between two kinds of moral reasoning and hence two kinds of moral reasons. I call these 'traditional' and 'nontraditional' (the latter could be termed 'conventional' or 'institutional' without much change of

meaning). Both are present in the writings of contemporary moral

philosophers but have not been emphasized or seen as crucial. On the basis of this distinction, I shall in Sections 2-5 show how ethical egoism accounts for morality by developing the view that moral systems as we

normally conceive them are to be understood as systems of nontraditional reasons and as such derive their rationality and their status as reasons with force from traditional reasons which are egoistic in nature. I shall then in Sections 6-8 show how, when understood in this

way, ethical egoism is immune to the logical points that have been rais- ed against it.

1

There are in fact two distinct kinds of moral reasoning, and cor-

responding to them different analyses of morality, moral principles, and moral behaving. On the traditional conception, moral reasoning is the activity of discovering moral principles and moral rules. Moral

principles are more basic and general than moral rules and specify the

323



Jesse Kalin

criteria that determine them. Presumably, there is one group of all the

possible or likely principles and rules which is the group of true ones. The activity of discovery itself is in no way essential to the nature, existence, or correctness of these principles and rules; it is simply the

process by which they become known. Furthermore, it is something about the principles and rules discovered, namely, their truth, and not

something about the activity of moral reasoning whereby they are discovered, that accounts for their force.

On the nontraditional conception, moral reasoning is not the ac-

tivity of discovery, but the activity of creation. It is the activity of es-

tablishing and adopting moral principles and moral rules. This activity is essential to their nature, existence, and correctness and hence is not

simply the process by which they become known, for prior to or apart from this activity, they are not there to be known. It is the process of moral reasoning itself which brings them into being as actual principles and rules with force.

Let us consider the nontraditional view first. The following is a minor illustration of moral reasoning as this sort of activity. Suppose five boys wish to play ball. They have a baseball and a bat, but obviously they cannot follow standard rules of baseball. Instead, they must adopt special rules to govern themselves in this situation. These rules will be actual rules - i.e., ones to which players can appeal and which will provide reasons for certain decisions - only if they are in fact adopted by the five boys. First, however, the boys must also mutually adopt certain principles governing how they will determine the rules: for instance, that they will all abide by the consensus, and that they will

give equal weight to each person's interest. If they do not adopt such

principles, they will be unable to decide on rules and thus unable to

play their game. But note that while the principles mentioned are perhaps the most

plausible ones, they are not the only possibilities. Instead of consensus, they could adopt principles which allowed their leader, or

perhaps the person owning the bat and ball, to choose the rules and to do so without giving their interests equal consideration. Sometimes

people want to play more than they want to avoid being taken advan-

tage of.

Having adopted, let us suppose, the principles of consensus and

equal consideration, the boys can proceed to the special rules. These rules will govern such things as foul balls, outs, walks, bases, etc., and settle such questions as: How many bases should there be? How far from home base should they be? Should batters be able to walk? Should there be a limit on foul balls? and so on. They nrrght decide to have one base, 20 yards long, no walks, no teams, and unlimited fouls. They could just as well have decided to have one base 30 yards long, no walks, teams, and a limit of three fouls. There is nothing in the rules to

324



Ethical Egoism as a Moral Theory

prevent them or guide them, for they are making the rules up, not dis-

covering them. That the rules are being made up does not mean that reasons can-

not be given for them, but the reasons given will be such as to appeal (either directly or indirectly) to the interests they have in playing the

game. For example, the claim that they should have only one base, since otherwise it is likely that too many people will be running at one time, leaving no one to field, is a direct appeal to their interest in hav-

ing a continuing and manageable game, whereas the claim that

dividing into teams of three and two would be unfair and violate the

principle of equal consideration, is an indirect appeal to their interest in not being taken advantage of.

Yet whatever these reasons might be, they are not, because they cannot be, reasons which, in virtue of their own truth or soundness, justify a particular rule in the sense of making that rule itself true or in force. These considerations give the players reasons to adopt a certain rule, but do nothing more. They cannot validate that rule. No matter what else is true, it must be adopted to be an actual rule, and in the im-

agined circumstances, such adoption is sufficient to put it in force. If the players ignore or are unaware of relevant considerations, foolish or even bad rules can become actual rules, which nonetheless will

provide genuine reasons for acting as they dictate. Mutual adoption, of course, makes no sense unless there are both

several persons involved and the matter is one of concern to each. Nontraditional reasons and reasoning, therefore, is essentially in-

terpersonal in character. This means that it will be needed only when the goals of different individuals come into conflict with each other.

Otherwise, there will be no use for a set of reasons designed to override individual considerations and to mediate between conflic-

ting needs and desires. Such reasoning would be both pointless and

impossible for a person alone who has no normal interaction with others.

This presupposition underlying nontradtional reasons gives us an

easy way to locate traditional reasons and traditional moral theories. If the reasons and theories are meant to apply to a person alone, apart from others, as well as to persons together, then it must be traditional in character. The force of such reasons can rest only in their truth and not in their having been adopted.

To illustrate, let us take the case of Robinson Crusoe, and let us

grant that he is entirely alone on his island and not in interaction with

any persons beyond it. Is it still possible for him to engage in moral

reasoning? On the traditional view, the answer is yes. For such

philosophers as Kant, Mill, and Moore, who are typical of this conception, Crusoe will be able to raise and answer questions which are

genuinely moral and not just egoistic.

325



Jesse Kalin

Thus, Crusoe might be tempted to suicide while despairing over his loneliness, but Kant thinks reason can discover suicide to be intrin- sically immoral. It is immoral because (supposedly) it could not become a universal law of nature. Similar considerations hold for the

temptation to go to pot and not develop his talents. Nothing in Kant's

arguments depends upon what others will or will not in fact do, or even that there are others involved in the situation. His arguments de-

pend only on what could and could not be willed to be a universal law of nature. Utilitarianism is also applicable to Crusoe. Crusoe can still discover that he has a duty to maximize happiness

- to seek the

greatest happiness of the greatest number, and he can apply that principle to his own case. As for Moore, if maximization of good in the world is one's moral duty, it is one's duty wherever one might be. One

always has the duty to do what one can. Crusoe, for instance, might be faced with alternatives realizing different amounts of beauty; if so, he would be morally obliged to choose that action which maximized this

good. On each of these philosopher's views, Crusoe will be faced with

moral issues requiring moral decisions. Clearly, questions will arise about his relations to animals and to himself, questions concerning such matters as cruelty and sadism, sodomy, and masturbation. In con- trast, on the nontraditionalist view, none of these questions can occur as moral questions since they will in no way involve other people or a conflict of interests between persons and there can be no question of

mutually adopting rules to adjudicate future conflicts. One can briefly summarize the difference between these two con-

cepts of moral reasoning as follows: For the traditionalist, the point of moral reasoning is to discover the truth of a particular proposition. Moral reasoning is just "reasoning about a case." Crusoe can do this as well as anyone else. The task of moral philosophy is to discover the

principles of such reasoning, the criteria of moral truth and falsity, and

perhaps the basic moral truths. For the nontradionalist, the point of moral reasoning is the attain-

ment of mutually satisfactory rules of interaction: it is to bring about agreement in the strong sense and thus settle (frequently in advance) conflicts of interest. Moral reasoning is "reasoning with another about a case." Such reasoning presupposes on-going mutual acceptance of certain principles (such as that of equal consideration) just as playing baseball does, and involves settling upon or "designing" rules by which to "adjudicate" issues. These rules are not moral rules because they are independently true propositions discoverable by rational reflection, but because they are in fact adopted by the parties involved when operating from these presuppositions in a determinate situation. The task of moral philosophy is to determine the presuppositions of such successful interaction and adjudication, the basic principles

326




that make it possible, and perhaps what it would be reasonable to adopt on their basis. But such principles and rules will still need to be actually adopted (usually tacitly) in order to be in force.

2

Nontraditional moral reasons have no force if they are not based in mutual adoption. Lacking such a basis, they are not really reasons at all.

Consequently, a person can, by refusing to adopt them or by withdrawing from a previous agreement, avoid their force. The

significance of this fact lies in the extensiveness of such reasons. The

previous section has, by way of illustration, established some of the moral reasons we in fact use as nontraditional, but gives no reason to

suppose that these are more than a small subclass of moral reasons as a whole, most of which would seem to be traditional in character. If, however, the class of nontraditional moral reasons were very exten- sive or even identical with the class of moral reasons itself, this would make morality contingent in a radical way and would force us to think about it in terms other than the usual utilitarian and deontological ones. What I shall do in the next three sections of this paper, sometimes in a highly programmatic way, is develop a theory of

morality as a set of solely nontraditional reasons. At bottom, what is at issue between the traditionalist and the non-

traditionalist is how a moral principle or rule comes to be in force. For

any given moral situation, one can imagine more than one principle which is applicable. Which is in force and how do we tell?

The tradionalist will say that the answer is at least theoretically sim-

ple. A moral principle or rule is in force if it is both applicable and true. He could also add by way of explication that it is in force if it is reasonable or more reasonable than any other applicable principle since it is just such reasons that establish the truth of moral principles and rules. It may of course remain a very unsimple thing to actually determine which principle is in fact true or supported by the best reasons.

The nontraditionalist will reply that this answer is only partially correct. He will agree that if there are any moral principles or rules which are both true and applicable in the above situation, they will also be in

force, but will argue first, that there are only a few traditional principles or reasons, second, that these are egoistic in character and con-

sequently, according to a common usage, not moral principles or

reasons, and third, that unless nontraditional moral principles and reasons can be established using these egoistic ones as their basis, the

only reasons in force in such cases, because the only traditional reasons available, will be egoistic ones. This would mean that disputes

327



Jesse Kalin

could not be morally resolved and thus could not be resolved without recourse to some manner of force.

In the next sections, I shall discuss these three claims of the non- traditonalist and shall argue that morality as a set of nontraditional principles and rules can be based on egoistic reasons. Thus, if successful, these sections will present an egoistic theory of morality even though egoism itself is not considered a set of moral reasons.1

3

Let us begin by looking at matters at their most general level. What are the possible basic principles that might apply to practical situations and be true?

I. Principle of Egoism: A person ought, all things considered, to do an action if and only if that action is in his overall self-interest. II. Principle of Equal Consideration: A person ought, all things considered, to do an action if and only if that action is in the general interest (where each person's welfare and interests are coordinate with every other person's welfare and interests, and each person is regarded as an "end in himself"). I shall discuss only these two, since they are the most plausible can-

didates, though there are other possibilities such as an Aesthetic Prin- ciple ("A person ought, all things considered, to do that action max- imizing the world's aesthetic qualities -

beauty, sublimity, etc.") and a Theological Principle ("A person ought, all things considered, to obey the will of God in all things.") (For convenience, I shall drop the 'all things considered', but unless otherwise specified, 'ought' is to be understood in this sense.)

To label these as principles is not enough. More important to know is principles of what? Principles are not simply principles, but are always principles of some enterprise or activity, such as plane geometry or politics. Principles can be judged to be reasonable or unreasonable, correct or incorrect only by reference to such an activity (though this in itself may not be enough). What, then, are the activities of which the above might be principles? The activity of practical reasoning or practical assessment would appear to be the obvious answer, but this will not do, for this "activity" can be divided into two distinct and separable modes of practical assessment which reflect the

1 For the purposes of this paper, I shall accept this distinction as valid and consider 'moral' to have a technical philosophic sense excluding egoism. See note 3 and Section 8.

328




distinction between reasoning about oneself considered alone and

reasoning about oneself and others made in Section 1. These are: A. Personal (or private) assessment of actions with a view to

deciding only what / ought to do. Here one asks the question: "Which principle ought I to use, given my sole interest is to assess and determine just my own conduct?" B. Interpersonal (or public) assessment of actions with a view to

deciding what we ought to do. Here one asks the question : "Which

principle ought we to use, given our interest is to assess and determine each other's actions, resolve disagreements, etc.?" Let us consider the activity of personal assessment first. Are there

any grounds for saying that one principle rather than the other is the

proper principle of this activity? There are two sorts of consideration relevant here. First, the activity may be of a certain sort such that one

principle is better suited to obtain the goals that are a constitutive part of it. Second, there may be traditional reasons which are applicable to this activity which favor one principle above the other.

It seems clear that either Principle I or Principle II could be used by a person to assess his own actions and to decide by himself what it really was he ought to do in a situation. There is nothing about Activity A as such that makes I a better principle of it than II. But there are traditional reasons favoring I as its principle, and I shall sketch them out in this section, though I will not develop them.

Parallel to, and the basis of, Principle I is a claim about what are reasons for acting, namely:

I'. Only a person's own wants and desires give him reasons for acting.

Similarly for Principle II: II'. Both a person's own wants and desires as well as the wants and desires of others give him reasons for acting. If I' can be established, then so can I since a person's self-interest

consists in those wants and desires most important to him as deter- mined by his own informed preferential valuation. There is no restriction on what a person can want or have an interest in. His wants may be selfish, confined to his own pleasure and advancement, or they may be nonselfish, directed toward the pleasure and well-being of another, or they may even be impersonal, scientific or artistic, for example. This means that the egoism embodied in I and I' is not as such egotism. The ethical egoist may have an interest in the welfare of others, but if he does, it is only because that other has some special connection with his wants and desires, such as being loved. Only in virtue of this connection can another's wants and desires provide reasons for acting according to I and T. Not so with Principles II and IT. According to them, other's wants and desires, and hence their interests, provide reasons irrespective of any connection with the agent's own wants and desires,

329



Jesse Kalin

and hence provide reasons to any and everyone in a position to act upon them.

What reasons can be given in favor of I' over IT? The egoist's position is made plausible by arguing first, that we can be more sure of I' than of M', and second, that all arguments to show that IT must be added to I' fail. The support for the first of these steps consists in noting such things as that it is easier to be sceptical about IT than about I', suggesting that the latter has a certain primacy the former lacks. One can, for instance, intelligibly suppose that others' wants and desires do not provide one with reasons for acting without undercutting practical reasoning; practical reasoning is still possible since there are one's own wants. One might think that one could also just as easily suppose that one's own wants did not, while others' wants did, provide one with reasons, but this is not so. In that case practical reason becomes very odd if not unintelligible, for how can M's desire for food for M

give N a reason to get food for M and yet not also give M a reason to get food for M? This makes no sense whereas the earlier assumption, that M's desire gave M a reason but did not give N a reason, does, for in this case M's wants have a connection to M they do not have to N, namely, the fact that they are his.

This asymmetry between the agent's wants and other's wants, for- tified by appeal to the sort of experience one has in crisis situations where it does not seem unreasonable to prefer one's own well-being to that of others when the two are incompatible, supports the egoist's claim that he can dismiss 1 1' as a set of traditional reasons having force, at least in the absence of any counter-argument to show that the wants of others must be included in the base set of traditional reasons. He, of course, has the task of refuting such arguments when offered, but has gained a foothold from which to operate.

A complete survey of such counter-arguments is beyond the scope of this paper. Typically, they assert that a consideration cannot be a reason just for one person when it applies to others as well. If M's wants and desires are reasons at all, this is a general fact about them and one which must be accepted as such by all rational agents. This means that they will be reasons for them as well. But this is not convin- cing. If M is playing chess, his desire to win gives him a reason for an in- capacitating move of his Queen, but it gives his opponent, N, no reason to assist this strategy or to in any way help secure M's goal of winning. Still, N can fully acknowledge that M has good reasons to move as he does. Here reasons are clearly relative to agents, and it seems false that IT must be added to I'.

Since II' can therefore be dismissed, for traditional reasons, as having less plausibility than I', II can also be rejected as the principle of the activity of personal reasoning, leaving I. Is I then in force with respect to A? Yes, with one qualification. The principles of an activity are in

330




force only on those engaged in that activity. For those not engaged in it, they have merely hypothetical force and provide only hypothetical, rather than actual, reasons. Can one not engage in A? Perhaps, but I find this difficult to imagine. Can one really avoid the activity of personally deciding what to do? It would seem that the mere presence of

practical alternatives would make it unavoidable, and that this is a necessity which comes with being human. If this is so, then one needs no reasons for engaging in the activity of personal assessment, and

Principle I, as a set of traditional reasons, will be fully in force.

4

Things are different when it comes to the second activity of in-

terpersonal or public assessment. First, this activity is of such a sort that

Principle 1 1 is better able to achieve its constitutive goals than Principle I. Interpersonal assessing is an activity of practical reasoning in which rational agents share a practical problem and must come up with a reasoned answer as to what they ought to do in the given situation. It will be natural for them to approach this problem by asking what

ought to be done. Principle I is in general not adequate for this sort of

inquiry for it will emphasize differences and conflicts in interests

among the reasoners, and will frequently dictate not a course of action

they ought to pursue, but opposed actions for each individual.

According to it, there is no one, common thing that ought to bedone or state of affairs that ought to exist. Therefore, I, if made the principle of B, would tend to encourage the forceful rather than rational settle- ment of practical problems and would indeed make interpersonal reasoning generally pointless. Matters are quite the opposite with

respect to II. Since it allows equal consideration to each and embodies a concept of the common good, agents can rationally arrive at common decisions and shareable courses of action. Thus II makes a better

principle of B than I. The second difference is the fact that the activity of interpersonal

assessment does not have the necessity the activity of personal assessment does. It can be avoided in a variety of ways. One can, for in-

stance, simply refuse to discuss practical matters with others or to

attempt any sort of co-operation. Thus, engagement in B is itself open to assessment as reasonable or unreasonable. The relevant question will be, "Ought I to enter into this activity (as conducted here) or not?" This is a personal question, hence such assessment must be made from the point of view of the first activity. If the line of argument in Section 3 is sound, then the only traditional reasons one can have for engaging in the activity of mutually assessing actions are self-interested, egoistic ones. Usually, these considerations make such engagement

331



Jesse Kalin

reasonable. Devices enabling common courses of action, including peaceful resolution of differences, are normally to our long-term advantage even if they place restrictions on the pursuit of our short-term interests. They offer more security than their alternatives, as well as make possible most of the institutions and goods of civilization.

Thus, the force of Principle II is contingent on engagement in Ac- tivity B which is contingent on engagement in Activity A and traditional reasons following from Principle I. However, it has not yet been established that II is a set of nontraditional reasons or that B is an activity of nontraditional reasonng. There is nothing about the nature of the activity of interpersonal reasoning as such which requires that its basic reasons be nontraditional. Whether they are to be nontraditional or not depends not on the nature of B but on the relation of Principle I to Principle II.

Indeed, some writers have adopted the position that II can be understood as a principle derivative from I without any intermediary agreement. All that is required is that a person be in a position such that the mutual adoption of II would bean egoistically reasonable thing to do. Then, it is claimed, reason requires him to follow the rules that would have resulted had agreement actually taken place. This move makes moral reasons and the moral point of view, as constituted by Principle II and Activity B, traditional reasons, although not the most fundamental ones.2 What I shall argue is that this view is mistaken and that Principle I requires that II and B be nontraditional in nature.

What is unsatisfactory about the view that II has derivative yet traditional force in virtue of I? It is this: In order for it to be rational according to I for a person to put I aside in favor of II when reasoning with others, that person must be assured that he and the others are employing the same principles and rules, otherwise such reasoning will be confused and even personally dangerous. These must be fully determinate, but no determinate principles or rules, actually usable in interpersonal reasoning, can be derived from I. Therefore, one can have this assurance only if the principles and rules governing B are standardized by the device of mutual choice and agreement.

Some writers have taken exception to this by arguing that there are certain universal moral rules central to any morality, such as "Do not

2 Modern writers presenting a version of this conception include Russell Grice in The Grounds of Moral Judgement (Cambridge; Cambridge University Press, 1967), who tries to use it to establish a form of utilitarianism; Kurt Baier in The Moral Point of View (New York; Random House, 1965); Richard Brandt in "Rationality, Egoism, and Morality," Journal of Philosophy, v. 69, no. 20, Nov. 9, 1972, pp. 681-697; and Michael Scriven in his chapter "Morality" from Primary Philosophy (New York; McGraw-Hill, 1966).

332




harm others/' "Do not murder/' "Do not lie/' and "Help those in need/' which are constitutive of morality and hence derivable from I as the basis of entering B. It is agreed that such rules are central to in-

terpersonal reasoning, that they normally follow from II, and that one could not engage successfully in B without employing them in some form or other. But as they stand, they are morally unusable because

they have a core of indeterminacy which can be removed only by a mutual decision to understand them in a specific way. For instance, while all moral communities have principles and rules forbidding causing harm or unnecessary suffering to others, they will and can vary in what they admit as a 'harm' and as 'unnecessary' (foreclosing a

mortgage is not, usually, unnecessarily harming another among us). There is no way to make these notions have a determinate content ex-

cept, in the final analysis, mutual agreement.

Similarly, exactly what counts as an act of murder and what is self- defense or justifiable homicide is not given by the rule against murder insofar as it can be based on I or other available traditional reasons. Is

killing one's wife's lover murder and morally culpable? It has not

always been. And whether it is or not cannot be derived by analysing the rule or the purposes it serves. Just where the line is drawn will de-

pend on the particular values and interests of the moral community in

question, may vary from community to community, and thus must be

agreed upon and mutually adopted in order for it to be appealed to as a moral rule in force (as contrasted with a moral rule it would be desirable to adopt). Indeed, not only is there a question about what counts as murder, there is a question about what counts as a man which must also, in the final analysis, be decided in the same way. This is what has happened in the controversy over abortion.

These "central" rules, therefore, provide a schema which can be filled in in more than one way without violating reason. Their content

gets specified in terms of exceptions made, excuses accepted, and cases regarded as paradigmatic. But what actually count as legitimate excuses, exceptions, and paradigmatic cases, and how in fact these considerations are to be ranked when they conflict is settled only by the mutual adoption (usually tacit) of a particular practice and are thus nontraditional in character.

Indeed, in the light of these observations and the discussion of five

person baseball in Section 1, it will be seen that there is no one determinate form of Principle 1 1 and that people could construct a system of

interpersonal reasoning using less- or even non-egalitarian principles. II is a nontraditional principle and the version of it given in this paper conveys its essential nature, but any actual version of it, governing an

on-going activity of interpersonal reasoning, will have a determinate

333



Jesse Kalin

form which is the result of a common understanding agreed to by its users.3

So far, I have argued that there are both traditional and nontraditional practical principles, and that egoism is a traditional principle which obtains its force from its reasonableness, while equal consideration, or whatever is taken to be the basic principle of morality, is a nontraditional one which obtains its force from mutual adoption. This means that morality is nontraditional in character and is to be understood as instituted in a complex way in accord with egoistic considerations.

More specifically, I have argued first, that the principles and rules

characterizing interpersonal or moral reasoning receive their

rationality from the egoistic purposes of those using such an institu-

tion; second, that since these principles and rules require a determinate form and content which cannot be derived from their rational

basis, their status as actual principles and rules of a particular moral

system depends not upon their general reasonableness but upon their

(at least tacit) adoption by the members of that system; third, that in-

terpersonal or moral reasoning is not an activity one must engage in; and fourth, that because of this, the actual moral principles and rules of a particular moral system will have force on an individual only if that

person has accepted membership in that system and has adopted those rules and principles as governing his, that is, their, activity of moral reasoning.

These last two considerations suggest that on this account one can

"opt out" of morality and that doing so will indeed be rational if a

system of interpersonal reasoning is no longer in one's overall self- interest. I shall discuss this possibility next.

5

Morality, when it is conceived as a system of nontraditional principles and rules, is to be understood as an institution. Institutions do two things: they regulate natural behavior by establishing public

3 It is because of this variability in determinate principles that content is not a

defining feature of morality. Rather, something is a morality in the technical sense used by this paper in virtue of the presence of most of a set of interpersonal activities which include praising, blaming, advising, punishing, excusing and justifying, and which are a necessary background for such moral attitudes and emotions as remorse, resentment, repentance, forgiveness, and vengeance. Thus, "adopting the moral point of view" is best understood as engaging in and

submitting to these activities thereby making oneself vulnerable to their accom-

panying attitudes and emotions.

334




systems of formal and informal, explicit and implicit rules, and they thereby bring into being kinds of behavior and goods otherwise non- existent. Thus, institutions constitute as well as regulate. Etiquette, for

example, may regulate behavior which exists independently of its rules, such as eating, but in doing so, brings into being other behavior and goods which have no existence outside these rules, such as rudeness and manners.

If the goods that the ethical egoist, operating from the basis of V, is interested in were all noninstitutional in nature, his need for a system of interpersonal rules, that is, morality, might not be very great. But in fact the egoist does need such systems and specifically the institutional

goods they create and maintain. This is because these goods are usual-

ly essential means to the satisfaction of the egoist's central and more

long term interests, including those of getting along peacefully with others and establishing enduring relationships of friendship and love, and are unobtainable apart from utilization of, and hence

membership in, such an institution. They include praising and blaming others, punishing and rewarding them, asking and giving advice, and

especially justifying and excusing one's behavior to others. These become possible with an institution because II, as a nontraditional moral principle mutually adopted in a determinate form, embodies a common but institutional objective good against which all actions are to be publicly measured. The value pluralism of I is thereby avoided and people are able to profitably ask "What ought we to do?/' usually in its elliptical form of "What ought to be done?"

There is a very delicate relation between the questions "What

ought I to do?" and "What ought we to do?" As long as one is

operating in terms of interpersonal reasons, reasons which can always be publicly shared, "What ought we to do?" is dominant and provides the grounds for answering "What ought I to do?" This is to say that in these cases II is dominant over I. If in some situation (one involving cheating, for example), II gives a different answer than I, II is then in force. This means that if one cheats for personal advantage one has no reasons one can give to others when caught to either justify or excuse such cheating. I, because it is egoistic, does not provide reasons that bear upon them, and hence cannot give others a reason to regard one's cheating favorably, while II does not normally provide reasons for cheating at all. This means that being caught in these circumstances

places one in a precarious position. One can stand by one's action on the grounds of the ultimate rationality of I, but this in effect will be to

deny one's membership in the moral community and will require others to resort to non-moral reasoning, in particular, to threats and force. Or one can acknowledge the force of II and admit having behaved wrongly by repenting, apologizing, making restitution, etc. This will reconfirm one's membership in the moral community,

335



Jesse Kalin

though it may not avoid punishment (acceptance of punishment is a

sign of sincere repentance), and will enable one to continue to share the benefits of a system of interpersonal reasons.

Still, II is dominant over I conditionally, and thus there is a sense in which "What ought I to do?" is always the governing question. As long as the answer is to engage in a system of interpersonal reasons because as a whole and in the long run this is more advantageous than not, II is allowed to be effective and to override I in particular cases. But when cases arise which show that interpersonal reasoning is no longer ad-

vantageous, all things considered, I becomes overly dominant and leads one to drop out of this activity. These are cases where one has no or few self-interested reasons to continue to impartially adjudicate conflicts or where the goods to be lost by doing so are much greater than those to be gained.

If a person in such a position chooses actually to opt out of the moral institution, its principles and rules will cease to have force on him and will no longer give him reasons for acting. Still, they will continue to include him within their scope since most moral rules are universal in character and therefore cover everyone, members and

strangers alike. The result of this is that members of the institution will have reasons to treat him in certain ways

- for instance, to fine him, or

put him in jail, or otherwise make him a public example - because the

rules say that no one should be allowed to do the various things he has done.

A system of moral rules, thus, has two "faces," an internal one and an external one. Internally, that is, as a member, the moral rules provide nontraditional reasons for all those mutually acknowledging them. Punishment for violations is then to be regarded as restitution or paying one's moral debt to others whom one has wronged. One's reason for accepting the punishment is that one has done a moral wrong which must be redressed. Externally, or as a nonmember, the moral rules as embodied in an on-going institution indirectly provide traditional egoistic reasons in that the institution is to be regarded as a system of prices set for various goods. One has to pay only if caught and whether the risk is worth it depends on the severity of the penalty, the likelihood of getting away with it, etc. "Punishment," then, is understood not as restitution, but as merely the cost of one's crime if one is unlucky and as a threat or danger otherwise. It has no significance except prevention and deterrence, and one has no reason to accept it.

That moral considerations have only a contingent force which can under the proper circumstances be escaped is crucial to an egoistic account of morality, for if the moral considerations could not be avoided when, for instance, one's life was at stake, the account would cease to be egoistic. The rest of this section will be devoted to an examination of the conditions that make it rational to enter into a system of in-

336




terpersonal reasons, revealing in the process when it is rational to

depart from such a system and why its reasons will then cease to have force.

Under what conditions is it rational to agree with others to subordinate I to II and thereby enter into a system of nontraditional in-

terpersonal reasons? Rawls, in another context, offers the following as the conditions of rationality for establishment of institutions and these are relevant to this question:4

(1) Members are self-interested, though not necessarily selfish or

egotistical; (2) "Their allegiance to the established practices is founded on the prospect of self-advantage;" (3) Members are rational, i.e., they know their own interests, are

capable of tracing out the likely consequences of different practices, are capable of deliberate and sustained action, are capable of

putting the long run ahead of immediate gain, and are not envious; (4) Members have "roughly similar needs and interests;" (5) Members are "sufficiently equal in power and ability to assure that in normal circumstances none is able to dominate the others." Conditions (1) and (2) are replaced by the adoption of ethical

egoism as the basic principle of practical reason. Condition (3) is standard, except for the restriction on enviousness which is not needed here. (It should be noted that condition (3) is closely related to (5) and

merges with it at several points; thus, inability to reason out conse-

quences or to control one's passions augments the power of others.) What is of special interest are conditions (4) and (5).

(4) requires a rough similarity in the needs and interests of the members of the moral community. In what way is such homogeneity a condition for adopting and abiding by principles of interpersonal reasoning? In particular, why is it a condition such that its failure may release one from the force of some set of moral principles?

Adopting the moral point of view involves recognizing the other's interests as co-ordinate with one's own - "recognize" not in the sense of "discover it to be a fact" but in the sense of "allow it in practice to be the case." Prior to such recognition, there is no moral equality between people's interests and welfares, for while Principle I gives a person a right to pursue his well-being whatever it might be, it also

permits others to ignore or even interfere with that pursuit should it be in their interest to do so. Engaging in the activity of interpersonal reasoning, since it involves granting a co-ordinate status to others, re-

quires restricting this latter sort of behavior. Such self-restriction is rational only on the prospect of long-range satisfaction of self-

4 John Rawls, "Justice as Fairness/' Philosophical Review, v. 67,1958, pp. 169-171.

337



Jesse Kalin

interest, that is, only if one stands to gain by it more than one stands to lose. This will be the case only if people's interests are similar enough to be reconcilable into one practical system so as to allow the con- struction of a common and objective good as part of the determinate content of II. If interests are so dissimilar as to be irreconcilable, it is likely that a person will lose more of value by restricting himself to II than he will if he deprives himself of the benefits of a moral system. In such event, moral relations will become impossible and irrational because harmful.

A similar irrationality results when condition (5) fails to obtain. In this case, recognition of the other's interests is unwise not because they are so different from one's own, but because one stands to gain nothing by it. There are two possibilities: one is weak and cannot readily resist the power of the other, or else one is in a position where one can get away with whatever one chooses, either because one is very powerful or because one is very clever and the system very inef- ficient. In the first case, moral recognition, because it is un- reciprocated, gets one nowhere and disarms one by making one think nonegoistically. I n the second case, it gives one less than can be gotten without it.

The theory advanced by this paper is committed to saying of a person who has opted out of morality because of the failure of (4) or (5) that the basic moral principles and rules have no force on him. He is not subject to them, just as persons in New York are not subject to Brazilian traffic laws. Thus, he cannot be said to violate these rules nor can they be appealed to by others to give him a reason for not doing as he is doing, even supposing him to be fully rational. This is because these rules will give such reasons only on the condition of their mutual adoption, and here this is both not the case and justifiably not the case.

What reasons can then be given among persons who do not accept a common moral system to lead them to form a moral community by adopting the principle of equal consideration and thereby placing themselves within the activity of interpersonal reasoning? There are two possibilities.

The first kind of appeal argues that moral relations are more valuable than the special interests that separate people. To think otherwise is to be seriously mistaken as to the nature of happiness and personal welfare, and thus as to where one's true interests lie. A life without moral relations is an impoverished life, one which no wise man would choose if he could safely avoid it. It is, for instance, a life without friends or family in the important sense of these terms.

This appeal is least effective in those situations where people's con- tinued well-being, or even their lives, are at stake. If basic interests are genuinely incompatible, the sacrifices involved in forming a moral community will not be reasonable. It is most appropriate when

338




directed at a person with superior power, but even here is apt to be ineffective. It is unlikely that such a person will think that the goods provided by a moral institution are better in quality and quantity than those otherwise available, for their value is most apparent to those who already experience them. And, indeed, it is possible that with respect to his own case his valuations are sound, in which event he will be rationally correct to continue to reject these goods and the activity in which they are embedded.

The second kind of appeal is threat. The implied argument has this form: "It is not at present in your interest to make concessions to our interests, but it will be, for if you do not make such concessions, we will act in a way harmful to your welfare." In order for such an appeal to have cogency, it must be credible. This means that people must have the power to do what they threaten, and must perhaps exhibit some of that power. But if the other party is very powerful and is secure in that

power, it is unlikely that he can really be threatened. In such an event, he is rationally entitled to continue in his "immoral" ways.

When both these appeals have failed, reason has been exhausted and all that remains is actual force if that is feasible. This both provides the grounds upon which violence can be justified and explains why the old concepts of "tyrannicide" and "regicide" and the newer one of "revolution" are important in an adequate political theory. They allow for a means sometimes necessary before a moral community can be formed.

This completes the first part of the argument. Ethical egoism is best understood not as a lone principle which is to be applied as it stands to each action, but rather as the foundation for establishing a system of

interpersonal, or moral, reasons which are nontraditional in character. The resultant principles and rules will embody some conception of common good, usually involving the recognition of persons' interests as morally equal, and will have force among parties only if mutually adopted. In normal situations, these moral reasons will be decisive, but, because they are contingent, cases can arise where it is rational to

opt out of the moral institution, with the result that they lose their force. This means that moral reasons will not always be superior to

egoistic ones. The conclusion to be drawn is that morality rationally rests on one or both of two conditions -

general good will and equality of power. In the case of good will, the moral relations are themselves of personal value, and hence members willingly accept some version of II. However, such acceptance is rational only if it is shared. Failing reciprocation, the institutional equality characterizing morality can be based only on equality of power, which in the final analysis will be the

power of hurting one another, that is, of adversely affecting each other's interests.

339



Jesse Kalin

6

Still, ethical egoism cannot serve as the foundation of morality if it is logically incoherent, and this is the prevailing opinion. It is therefore necessary to turn to the most important recent attempts to substan- tiate this claim and show that they do not succeed.5

These objections form a family and their core is succinctly expressed by Richmond Campbell in his article, "A Short Refutation of Ethical Egoism." There he argues:6

(i) If an agent ought to do something in a given situation and another agent ought to do something in the same situation, then it is not logically impossible for them to do these things in that situation. (ii) There is a situation S where an agent M would benefit most from doing X and another agent N would benefit most from doing what would prevent M from doing X. (iii) If IE E is true, then M ought to do X in S and N ought to do what would prevent M from doing X in S. (iv) If M ought to do X in S and N ought to do what would prevent M from doing X in S, then it is not logically impossible for M to do X in S and for N to do what would prevent M from doing X in S. (v) It is logically impossible for M to do X in S and for N to do what would prevent M from doing X in S.

Therefore, from (iii), (iv), and (v), we must conclude that IE E is not true. (i) is self-evident according to Campbell; (ii), the description of a

possible situation; (iii), the application of ethical egoism to this situation; (iv), the instantiation of (i) with respect to the situation given in (iii); and (v), a consequence of (ii).

Campbell's argument, and the others related to it, depends for its success on premise (i). In order to provide a refutation of ethical egoism, (i) must be true in virtue of the ethical concepts the egoist

5 Previous defenses of egoism in the literature which are relevant here include my articles: "On Ethical Egoism/' American Philosophical Quarterly Monograph No. 1, 1969, pp. 26-41; "In Defense of Egoism/' in David Gauthier, ed., Morality and Rational Self-interest (Englewood Cliffs, N.J.; Prentice-Hall, 1970), pp. 64- 87; and "Baier's Refutation of Ethical Egoism/' Philosophical Studies, v. 22, 1971, pp. 74-78.

6 Richmond Campbell, "A Short Refutation of Ethical Egoism/' Canadian Journal of Philosophy, v. 2, no. 2, Dec. 1972, pp. 249-54. 'IEE' stands for Impersonal Ethical Egoism, "the view that everyone ought (morally) to do what will benefit him the most in any given situation" (p. 249).

340




must use, for otherwise its force will depend on substantive ethical judgments which beg the question. It should be recalled here that the proper and careful formulation of ethical egoism uses the concept 'ought, all things considered' rather than 'ought' or 'morally ought' and is:

I. A person ought, all things considered, to do an action if and only if that action is in his overall self-interest.

Consequently, if there is to be a refutation of I, (i) must in some way be implied by it and must hold in any system in which the concept 'ought, all things considered' has a part. I shall argue that (i) fails to meet these conditions and that this failure severely undercuts all attempts to refute ethical egoism on other than substantive grounds.

To see why (i) fails to have the force required of it, it is helpful to speculate on why (i) might appear to be self-evident, (i) says that, if M ought to do X in S and N ought to do Y in S, and if M's doing X in S is logically incompatible with N's doing Y in S, then either M ought not to do X in S or N ought not to do Y in S. Why does this conclusion follow from these premises? The reason lies, I think, in the following argument:

1. If M ought to do X in S, then it ought to be the case that M do X inS. 2. If N ought to do Y in S, then it ought to be the case that N do Y in S. 3. Therefore, if M ought to do X i n S and N ought to do Y in S, then it ought to be the case both that M do X in S and that N do Y in S. 4. If when M does X in S, N cannot do Y in S, then it cannot be the case that it ought to be the case both that M do X in S and that N do Y in S. 5. Therefore, if it ought to be the case both that M do X in S and that N do Y in S, then it cannot be the case that when M does X in S, N cannot do Y in S.7

This might appear to be merely the same argument as before, but if so, it at least makes explicit what was there implicit, namely, the logical connection employed in 1 and 2.

I shall use ' a' to refer to direct ought-statements of the form 'M ought to do X in S' and ' /}

' to refer to ought-statements of the form 'It ought to be the case that M do X in S', 'X ought to be done in S by M',

7 This argument is even more persuasive if one substitutes for 'it ought to be the case that M do X in S\ 'X ought to be done in S'. Then, if (i) were not correct one could conclude that "X ought to be done in S and -X (or Y) ought to be done in S." And, indeed, Campbell understands the argument in this way: "This use of 'ought' is, I take it, unconditional or categorical in that it entails that both X and Y ought to be done, together, and hence that it is possible for both to be done either simultaneously or in sequence" (p. 252).

341



Jesse Kalin

and 'X ought to be done in S\ Any system of reasons in which an a- statement implies a 0 -statement is a strong system of reasons, and any in which a-statements do not imply /3-statements is a weak system.

What I shall ultimately argue is that ethical egoism is a weak system in which its ought-statements have no /3-consequences; that (i) holds

only in strong systems for only then do premises like (1) and (2) obtain; and that therefore, ethical egoism has not been refuted. But first, more about strong and weak systems.

Strong systems, being those in which one can infer from a- statements to ^-statements, imply that there is something that ought to be the case. They thus embody a set of objective and shareable

goods, that is, goods whose value is not (as far as the system is concerned, anyway) relative to persons' wants or interests or to other personal attributes and which consequently give everyone (in the system) a reason to effect them. Such goods can be happiness as such, or justice, or many other things. Whatever the common good is, it is only because it is a common good that anyone ought to do anything. For instance, given utilitarianism, it is solely because happiness ought to be maximized that Jones ought to help Smith pull his cow out of the ditch. Such goods provide the ultimate reasons. This means that in a strong system the foundation of a-statements is a set of objective values which can be expressed in the form of /3-statements. 'M ought to do X in S', therefore, implies 'It ought to be the case that M do X in S' and 'X

ought to be done in S' only because the latter are independently true and the original source of the former.

Weak systems, where one cannot infer from a-statements to any /3-statements at all, do not imply that there is anything that ought to

be the case and, indeed, are incompatible with any such objective goods. In such systems, the fact that M may have conclusive reasons for doing X in S ('ought, all things considered') does not mean that X or

anything else ought to be done. Such a system does not promulgate any common ends or states of affairs, not even the following of itself. It will say that M has conclusive reasons for himself doing X in S and hence that he, M, ought to do X in S, and that N has conclusive reasons for doing Y (or not-X) in S and hence that he, N, ought to do Y in S, and this is all.

This means that weak systems must have a different logical structure, and therefore a different foundation, than strong systems. In particular, they cannot include any /8-statements which are independently grounded. Rather, they must presuppose relativistic values and reasons which are dependent on person-specific characteristics such as the agent's wants and interests. Only then can it be the case that M can have a reason for doing X in S without others also having a reason for seeing that X is done, that is, without it being

342




the case that X ought to be done in S by M, as is required if a- statements are not to imply /3-statements.

Because a-statements do imply /3-statements in strong systems, however, (i) will have force in them. This is made clear by the argument on p. 341. So also will several corollaries of (i) which have found their way into various other "refutations" of ethical egoism, including:

One ought never to prevent someone from doing what they ought, all things considered, to do.8

That a person ought not to do something is a consideration against causing him to do it.9

That a person A ought to perform a given action is a consideration in favor of any other person B who is able, helping to bring it about that A does so.10

It is easy to see how these principles are true given ^-statements and hence how, thinking that a-statements must imply ^-statements, they appear to follow given any a-statement. But this is a mistake, for while strong systems will include Campbell's principle (i) and its corollaries, they will not do so because these principles follow from the

meaning of 'ought' or 'ought, all things considered' as these occur in a-statements. Rather, they will be included because /^-statements

are independently present in the system and because these are founded in a set of objective goods the objectivity of which insures and re-

quires them. This means that (i) and its corollaries will not hold in weak

systems, even though these use some of the same ethical vocabulary that strong systems employ, because there are no objective goods in such systems upon which they can be founded. It should now be clear

8 See Baier's The Moral Point of View, pp. 95-96, and my "Baier's Refutation of Ethical Egoism." In a recent reply to this article, Baier has substituted the principle "One ought never to aim at preventing another from doing what he ought" (see "Ethical Egoism and Interpersonal Compatibility/' Philosophical Studies, v. 24, 1973, pp. 357-368).

9 Warren Quinn, "Egoism as an Ethical System," Journal of Philosophy, v. 71, no. 14, Aug. 15, 1974, p. 460. 1 have altered Quinn's formulation here, and in the next

principle, by dropping the 'morally' from before the 'ought' to keep it consistent with premise (i). This use of 'moral' is discussed in Section 8.

10 Quinn, p. 460. Kant uses a similar principle in the fourth illustration of the Categorical Imperative.

343



Jesse Kalin

that (i) can be self-evident as Campbell claims only if it is also self- evident that value-relativism and I' are false, and this seems hardly the case.

Ethical egoism as formulated in I and as described in the earlier sections of this paper is precisely such a weak system of reasons. It, when considered at its fundamental level, apart from subsidiary systems of nontraditional reasons, contains only a-statements. Campbell, and all those who attempt a similar attack, have therefore imported into their

arguments principles which the egoist both rejects and is not logically committed to. Rather than being self evident, (i) and its corollaries are

dependent on substantive value judgments which beg the question and, consequently, the "refutations" of ethical egoism based on them fail.

7

There are objections to this treatment of Campbell's argument that must be considered, but first let me connect strong systems of reasons with nontraditional systems. A strong system is characterized by a set of objective goods which that system embodies. These are the foundation of its ^-statements and can be of two sorts - natural or constructed.

The first sort requires that there be traditional reasons which support the claim that certain goods are objective and nonrelative in value. Because such values will therefore be part of the nature of things, the /3-statements expressing them will apply to and have force on everyone, irrespective of their desires and interests. This will yield a set of strong, traditional reasons. The egoist, of course, must contend that there are no reasons or values, that is, no ^-statements, of this kind, and it is at this point that the arguments outlined in Section 3 become crucial, especially those whose intent is to show that I' does not commit one to II'.

The second sort, however, requires only that there are traditional reasons supporting the mutual selection of some particular set of goods. These reasons can be egoistic and this common good will then be "constructed" out of the personal relativistic goods of the individuals involved and will include as essential elements institutional goods as defined in Section 4. Since a constructed good is not a natural part of things, the system of reasons founded on it will be a set of strong but nontraditional reasons. The /3-statements of such a system, while they will still apply to everyone in virtue of their universal and objective form, will, because they express nontraditional reasons, have force only on those who are members of the system and who have mutually accepted its underlying goods. Furthermore, the system

344




as a whole will be subordinate to the weak system of ethical egoism. Even though it is a strong system, its strength is conditional and may in extreme situations dissolve.

This means that one cannot tell from its logical form whether a 0- statement is backed up by egoistic or nonegoistic reasons. And it means that one cannot conclude from the mere presence of strong systems that egoism is false. What matters is not whether a system is

strong or not, but whether it consists of traditional or nontraditional reasons. And the only way one can determine this is philosophically.

The result is that the available strong systems, because they are nontraditional given the argument of Sections 3 and 4, have an es-

pecially complicated and deceptive logical structure which may foster confusion in analysis. This is compounded by the fact that traditional and nontraditional systems share the same ethical vocabulary and that, while in fact strong systems are subordinate to a weak system, they normally do not recognize this fact and present themselves as noncon-

tingent. The resultant danger is that the logic of an ethical sub-system will be taken as the logic of ethical systems as such.

Just exactly this sort of mistake has been made, I believe, by many if not most critics of egoism and it will be helpful here to look at a few

specific cases. Their objections as I shall present them center around the contention that weak systems of reasons are really a fiction and im-

possible. What misleads these critics and underlies their difficulty in

recognizing as genuine the observation that a-statements do not as such imply ^-statements is, as stated above, one or both of two things: the fact that the same words, 'ought' and its derivatives, have several different meanings, and the fact that ethical language in its ordinary, as contrasted with its philosophical, occurrence, is almost always used as

part of some ongoing institution of interpersonal reasoning, that is, as embodied in a strong system which is not fully aware of its nature.

Exactly what does 'ought, all things considered' mean when it occurs in I or in a-statements like 'M ought, all things considered, to do X in S'? It means that the agent in question has conclusive reasons, tak-

ing all relevant reasons into account, to do a certain thing or to act in a certain way. It does not mean or imply that it is desirable that the agent do this even though we also frequently use 'ought' to mean "most desirable," "good," and "best." These two senses of 'ought' must be

separated; the former is "deontological" and the latter is

"teleological." Its being the most reasonable thing for an agent to do is

quite a different thing from its being a desirable thing for him to do. It

may not be desirable at all, yet still what he has conclusive reasons for. A purely egoistic situation (i.e., one in which sub-institutions of nontraditional reasons have not been introduced) would abound with such cases, as do competitive games. Thus, if it is in A's interest to do X, which is contrary to my interest, then while it is the case that A ought,

345



Jesse Kalin

all things considered, to do X since he has conclusive reasons for doing X (as provided by I), it is not the case that it is desirable (or good, or the best choice, etc.) that he do X because first, there is no common good in this situation which means that value terms must always be implicitly relative, and second, A's doing X is not desirable as far as I am concerned.

Can it be shown that this is incoherent and that these two senses of 'ought' are not independent? Two recent attacks on ethical egoism have attempted to do this. George Carlson, for instance, argues that the egoist's analysis of competitive games is faulty and that they do not support his contention that there can be a system of reasons where one can believe that someone ought to do something and yet neither want him to do it nor have a reason to help him. The core of Carlson's argument is:11

...in Kalin's example what I really believe is not that this is how my chess opponent ought to move, period (i.e., for the sake of simplicity, in such a way as to put my king in inextricable check), but that this is how he ought to move if the game is to be won by him (according to the rules); that he ought therefore to try to move that way. Since, however, (ex hypothesi) I do not want him to win the game, it is plausible to assert that I do not really believe that he ought to succeed in moving in the prescribed way, and despite the fact that (with Kalin) I can meaningfully assert (I) ["This is how he ought to move."] (so long as its suppressed hypothetical force is understood). I conclude that this particular putative counter-example does not in fact support Kalin's key (more general) claim, namely, that one's belief in what anyone else ought to do or ought to make the case does not commit one to wanting any one else to acting so. As with the previous competitive game example, the state of affairs one believes ought to be the case (e.g., the attempt of one's opponent to make a field goal or to inextricably check one's king), and that state of affairs which one wants not to see realized (e.g., the actual scoring of the field goal or the actual checkmating of the king) are not the same.

This argument turns, not on the analysis of game-directives as hypotheticals, but upon the meaning of 'ought' or 'ought, period'. Why can't I believe my chess opponent really ought, all things considered, to put my king in checkmate? Because I do not want that state of aff irs to come about. That is, because that state of affairs is, as far as I am concerned, undesirable and therefore ought not to exist. But while this latter belief may be true, it also employs a different sense of 'ought'

- its teleological rather than deontological sense - and in no way contradicts the first belief. Carlson's mistake is to fail to distinguish these senses and thus to think that a-statements automatically imply /3-statements.

11 George R. Carlson, "Ethical Egoism Reconsidered/' American Philosophical Quarterly, v. 10, no. 1, Jan. 1973, pp. 25-33. Carlson is referring to my article "In Defense of Egoism."

346




It is this that leads him to speak of states of affairs "one believes

ought to be the case" (p. 28). As should be clear by now, the ethical

egoist (as well as the astute gamesman) does not, and does not have to, believe that there is any state of affairs that ought (all things con-

sidered) to be the case. I n Carlson's example, he believes only that one

party has conclusive reasons for checkmating his king, and in addition finds this a for him undesirable state of affairs. Clearly, what he believes and desires concerns one and the same state, and is, given the

separation of these two senses of 'ought', quite compatible. Carlson's

analysis appears correct only if one persists in thinking of competitive games as strong systems when they are not.

Warren Quinn is also unsuccessful in a related attempt to show that

/ought' always has a sense expressible by 'good'. Quinn argues:12

When a system issues a directive that Jones ought to <|> , it is, in a perfectly obvious sense, positively oriented toward Jones's <|>-ing. Our bedrock intuition concerning the coherency of such a system therefore demands that it extend this positive orientation in a suitable approbative. This is why it seems unacceptable that an impersonal system S should regard the fulfilment of its directives as bad. But, by the same

reasoning, (3) is also unacceptable; for to regard the satisfaction of an overall directive as neutral is to regard it as not good. Hence, by elimination:

P4: It is good from the point of view of an impersonal system S that one do what, all

things considered, one S-ought to do.

and a little bit later:

If a system issues a directive that Jones <j>, then the system must be positively oriented toward Jones's satisfying the directive, and therefore must regard it as in some sense good that he do so. The important thing to note about this inference is that prudence constitutes no counter-example; for clearly it must be prudentially good for Jones to do what he prudentially ought to do. Thus, some general principle of inference from 'ought' to 'good' (but perhaps no principle transferring 'ought' interpersonally) seems valid in every system.

Seen in the light of the preceding discussion, Quinn's arguments are unpersuasive. Indeed, it is hard to find in them more than the in-

sistence that a system must take a positive orientation toward the

satisfaction of its directives, which is countered by the egoist's in-

sistence that in weak systems this is not so. The egoist's "bedrock intuitions" seem as good here as any one else's.

But the disagreement is not just a matter of intuitions. A system "wants" its directives followed only in the sense that the system embodies a set of common goods which it expresses in the form of |8- statements. Weak systems do not contain /3-statements because they do not embody a common good and therefore cannot "want" their

12 Warren Quinn, "Egoism as an Ethical System," pp. 465-466.

347



Jesse Kalin

directives followed. The satisfaction of a directive is, as far as they are concerned, neither good nor bad; they have no interest in the matter precisely because they contain no /3 -statements. (Is Quinn playing on the connotation of 'bad' when he says that this means the system regards the satisfaction of its directives as 'not good'? But then why say this, for if it is neutral, it is 'not bad' as well.) Consequently, it is a mistake to say, as Quinn does, that a weak system cannot regard the satisfaction of its own directives as neutral. He makes this mistake, I think, because he fails to note that the same word has different senses, particularly as used in philosophical contexts. Because of this, rather than show that the egoist cannot distinguish the two senses he does, Quinn in the end merely assumes that they are at bottom the same and thus begs the question.13

The existence of such propositions as "It must be prudentially good for Jones to do what he prudentially ought to do" does not help Quinn's case. Rather, it only shows that we are often not very careful in the language we use. What do we mean when we say this? Going by his other examples (see p. 467), this is the same as "From the prudential point of view, it is good that Jones did what was in his self-interest." But

'good' here simply means "correct" or "justified," that is, "supported by prudential reasons," and none of these have the value implications Quinn needs for his argument. At least, this is all 'good' need mean in this context and thus certainly all that the egoist will allow it to mean.

The ancestor of these two arguments against ethical egoism is the version given by Moore in Principia Ethica. Moore says:14

What, then, is meant by 'my own good'? In what sense can a thing be good for me? It is obvious, if we reflect, that the only thing which can belong to me, which can be mine, is something which is good, and not the fact that it is good. When, therefore, I

13 If my analysis shows that P4 is not sound, then Quinn's argument fails, for on the basis of it he wishes to show that "the egoist must admit the moral desirability of other people doing as they, all things considered, morally ought" and must therefore "further recognize a moral consideration in favor of his helping to make it the case that they do so" (p. 469). But it might be that my analysis is irrele- vant to P4. Even though Quinn presents himself as demonstrating that directives in an ethical system such as egoism must imply some kind of goodness, in which case counter-examples are relevant, it may be that P4 is true merely in virtue of Quinn's special understanding of 'impersonal system'. Unless I have missed something, he seems to say that a system 5 must be an evaluative system and that it is an impersonal one if it contains "approbatives" of the form "It is good (bad, etc.) that P" (see pp. 462-63). If this is what makes P4 true, then Quinn must also show that ethical egoism is necessarily an impersonal system in his sense. This he tries unsuccessfully to do in section VI of his paper (see pp. 351-2 below for a discussion of this argument). In either case his argument fails.

14 G. E. Moore, Principia Ethica (Cambridge; Cambridge University Press, 1960), pp. 98-99.

348




talk of anything I get as 'my own good', I must mean either that the thing I get is good, or that my possessing it is good. In both cases it is only the thing or the possession of it which is mine, and not (he goodness of that thing or that possession. . . If, therefore, it is true of any single man's 'interest' or 'happiness' that it ought to be his sole ultimate end, this can only mean that that man's 'interest' or 'happiness' is the sole good, the Universal Good, and the only thing that anybody ought to aim at. What Egoism holds, therefore, is that each man's happiness is the sole good - that a number of different things are each of them the only good thing there is - an absolute contradiction!

As the next to last sentence makes clear, Moore's argument relies on the familiar assumption that any a-statement ('that man's own

happiness ought to be his sole ultimate end') implies some /3- statement ('that man's happiness is the sole good and the only thing anybody ought to aim at'). This assumption came into play at the very beginning of Moore's account when he defined egoism "as a form of Hedonism . . . which holds that we ought each of us to pursue our own

greatest happiness as our ultimate end" (p. 96), which is acceptable, but then went on to explicate this as the view "that each man ought rationally to hold: My own greatest happiness is the only good thing there is . . ." (p. 97), which is not. Here, Moore has failed to recognize the two distinct senses of 'ought' and has therefore forced upon the

egoist a concept - 'my own good'

- which he need not use. For the

egoist, there are, strictly speaking, no "good things" at all, that is, no natural objective goods. Furthermore, while there are things which are "good relative to his interests" (and in that sense his own good), these constitute only one aspect of his theory and are not the fundamental reasons upon which that theory is based. Those are quite abstract, theoretical, not particularly teleological in character (see Sec- tion 3), and do not commit one to the kinds of logical connections a refutation of ethical egoism requires.15

As already mentioned, one reason these different senses of 'ought' have been difficult to see and keep distinct is the fact that any appeal to

ordinary language and usage will almost always confine the meaning of basic ethical terms to that sense they have in strong systems. This is both because we are usually always within such a system and it re-

quires a special effort to stand apart from it, and because nearly all social contexts, those in which we use the terms, are part of an institution of interpersonal reasoning. Given this, it is easy to think that

analyses derived from such references fully capture the meaning of the terms involved. Consequently, language presents a kind of blinders and can be misleading, for weak systems can both be con-

15 An attempt to make Moore's argument work is made by Thomas Nagel in The

Possibility of Altruism (Oxford; Oxford University Press, 1970), Part Three, pp. 79-146. I believe it fails for essentially the same reasons given in this section.

349



Jesse Kalin

structed and in part found, and these support a more formal, less content-laden account of ethical terms.

Baier, for instance, has recently advanced the claim that ethical

egoism is incoherent becuse it says that each person has a right to do what is in his own interest and yet allows that others may violate that

right and prevent its exercise. Rights, he argues, are connected essen-

tially with corresponding duties, especially the duty not to prevent their legitimate exercise.16

This is a familiar argument, but its great appeal is due to looking ex-

clusively at 'right' and 'rights' as they function in strong systems. There

they serve as a check on the behavior of others, bringing it into line with the common good they have been designed to promote. So it is not surprising that when we consult normal experience which is confined to such contexts, we find few if any cases where rights are not ac-

companied by duties and thus seem to confirm Baier's argument and his conclusion that the egoist cannot talk about having a right to do what he ought to do.

But here the appeal to ordinary contexts and usage is misleading, for there is a sense of 'right' which does not have these consequences, though it may have been invented by philosophers and its use confined to their writings. That is 'right' as "liberty," "rational justification," or "rational permission."17 This is the sense used by Hobbes when he speaks of each person having, according to the Law of Nature, the right to do whatever is necessary to protect his own life. He has this right because that is always the most reasonable, and therefore

always the right thing to do. But note that Hobbes and the egoist have not arbitrarily appropriated the term; it still retains most of the sense it has in strong contexts, for even here it emphasizes that the conduct is

justified, that the agent is doing nothing wrong in behaving as he does, and that no one else can legitimately complain. (This is true because in this situation there is no common good which can invalidate the egoistic justification or form the basis of a complaint.) Therefore, if the

egoist has made a mistake in using 'right', it is not a logical but a strategic one since 'right', in virtue of its normal connotations, usually confuses instead of clarifies.

16 Kurt Baier, "Ethical Egoism and Interpersonal Compatibility/' pp. 358-62.

17 Baier notes that the egoist might reply in this way and says that this account of 'right' will fail, though he does not in this article say why (p. 362). I do not see what his argument could be unless it in some way involves the view that reasons cannot be person-relative and I' commits one to IT.

350




Quinn also makes several appeals to what we would say in normal conditions to support important parts of his argument. At a crucial

stage he says:

Abstracting from specifically moral contexts, it seems obvious that we call an action

right ... only if we think it is a good choice . . . i.e., among the best choices available. And we call an action wrong. . . only if we think of it as a poor or badchoice. . .That

is, Tightness or wrongness from a given point of view is understood to be nothing other than a certain special sort of goodness or badness from that point of view . . .

'right' and 'wrong' express a type of goodness .... (p. 471)

This passage does not differ in substance from Quinn's earlier claim that 'ought' always implies 'good' and is subject to the same criticism. The problem here is why it might appear obvious to someone that

'right' and 'wrong' "express a type of goodness/' when there is at least a strong philosophical tradition in which the concepts are dis- connected. It is again, I think, because most practical contexts in which

'right' and 'wrong' are used are strong contexts, and in them the right choice, in addition to being conclusively supported by reason (so far as that system is concerned), is both a good choice (again, so far as that

system is concerned) and one backed by reasons ultimately grounded in a set of objective goods. This makes it seem that "we call an action

right . . . only if we think it is a good choice," where by 'good' we

appear to mean something other than 'correct' or 'justified'.18 But it will follow that 'right' and 'wrong' always express a type of goodness only if one can also show that all contexts in which they occur are

strong, and this Quinn does not do.

Perhaps this failure to step outside of the strong system one is nor-

mally in and attend to weak systems also accounts for the extreme "bonism" that has mislead many other critics of ethical egoism. Moore, for instance, could not really conceive of any ethical foundation other than the good and went so far as to say that judgments about

obligation and Tightness "can only mean that the course of action in

question is the best thing to do" and "obviously" "assert that more

good and less evil will exist in the world" if followed than if not followed.19 Such a position begins to have plausibility if one confines oneself to the structure of strong systems which always, even if they

18 That Quinn does not mean 'correct' or 'justified' is shown by his subsequent remark: "Moreover, the goodness involved in Tightness is obviously a result of

the good aspects of the right act .... Thus, moral Tightness must derive from a

type of moral goodness which . . . would seem to be impersonal" (p. 472).

19 Principia Ethica, pp. 24-25 (the emphasis on 'mean' is mine). The subsequent

quotes from Moore are also from this work. A similar view is advanced by Jan Narveson in Morality and Utility (Baltimore; Johns Hopkins Press, 1967), pp. 268-

71.

351



Jesse Kalin

are highly deontological in character, contain /3-statements, which in turn seem to use a sense of 'ought' that is capturable by 'desirable', 'good', and 'best'. In such a system, if one is to justify one's action to another, one can do so only by reference to a set of goods that he can share. Otherwise, the "justification" cannot succeed in producing the

agreement in actions that is its point in the strong system. This makes

plausible Moore's claim that "the only possible reason that can justify any action is that by it the greatest possible amount of what is good ab-

solutely should be realized" (p. 101); in a strong system this appears to be true. But unfortunately for Moore and his followers in this, some ethical systems of reasons are weak and actions in them are justified or

rationally supported by appeal to reasons other than an absolute, common good. Once it is admitted that there can be such systems, all of the Moore-type arguments become ineffective because about

something else.

8

There remains one final difficulty. While critics may grant, in view of the preceding discussion, that it is implausible to hold that weak systems of reasons are impossible or incoherent, they can still insist that such are not moral systems and that, if ethical egoism is coherent only as a weak system, then ethical egoism is not a theory of morality.

This is presented as a clinching counter-argument. "Alright," the critic seems to say, "we will grant that I is consistent but that, you see, doesn't matter, for it is not a moral theory, only a 'prudential' one." But this will not do, for it is in effect an attempt to resolve substantive matters by verbal decisions.

To begin with, it is not clear what 'moral' in this context means. Frankena20, in one of his attempts to determine the concept, suggests that the word denotes "a wider formal concept of morality and a narrower material and social one." A system of reasons is a morality in the wider sense if its adherent "takes it as prescriptive," "universalizes it," and "regards it as definitive, final, over-riding, or supremely authoritative" (p. 687), while it is a morality in the narrower sense if, in addition to these, there is a restriction on its content so that the effect of actions on other persons or sentient beings is always relevant to their correctness (p. 689). Ethical egoism is a morality in the first, but not in the second sense.

20 William Frankena, "The Concept of Morality/' Journal of Philosophy, v. 63, no. 21, Nov. 10, 1966, pp. 688-96.

352




Quinn denies even this much to the egoist and argues that

prescriptivity, universality, and overridingness do not constitute a

legitimate sense of morally ought' because "one wants to ask whether the overridingness that distinguishes the moral 'ought' is supposed to be cfe facto or normative. If the former, the account seems false; for it is certain that some people attach a greater weight to 'ought's they regard as non-moral than to those they regard as moral" (pp. 470-71). I am not sure what this observation shows. On the one hand, people often do what they believe they ought not to do, but here I take it they still regard this 'ought' as overriding though they do not heed it. On the other, why isn't it that the people Quinn speaks of are either unac-

quainted with the wider sense of 'ought' or are mislead by the

familiarity of the narrower one into thinking it the only sense? What one really wants to ask is whether overridingness is a part of the proper analysis of morality, and what people say and do in this regard is of limited help.

Overridingness as a part of morality has its basis in the fact that

morality originates in the need to decide what to do in determinate situations where ideally all alternatives except one are to be rationally excluded. On my account, the distinction between the wider and

narrower senses of 'moral' parallels the distinction between personal and interpersonal activities of reasoning and the fact that the former of

these has a 'wider' extent than the latter.

However, as noted earlier (see notes 1 and 3), I think there are good reasons for restricting the use of 'moral' as a .technical philosophical term to the second, narrower sense, and thus am inclined to concede

to those who insist that a morality is always a strong system of reasons.21

But this terminological stipulation neither eliminates the first sense

nor makes it unimportant. To denote it, I prefer the terms 'ethic' and

'ethical', now also to be used technically. Thus ethical egoism, while

not a moral theory, is an ethical theory. What is therefore at issue, and

surely what Campbell, Quinn, Carlson, Baier, et a/., are concerned

about, is whether moral reasons are rationally superior to ethical

reasons, that is, have overriding force, and this question cannot be

answered by noting that moral reasons claim such superiority, for

ethical reasons claim it too. The question is, whose claim is correct, and

no amount of analysis of 'moral' and 'ethical' will tell us that.

This paper, in developing a more complete account of egoism as an

ethical theory, provides an answer which denies the superiority of

moral reasons while at the same time accounting for their appearance as superior. The essential parts of this theory are first, the distinction

21 The egoist will accordingly modify Frankena's definition of the narrower sense

by adding to the third condition the phrase "within the activity of interpersonal reasoning."

353



Jesse Kalin

between traditional and nontraditional reasons; and second, the con- strual of systems satisfying the criteria for a morality as institutions of

strong, but nontraditional reasons. Given the first of these, practical directives have force on an individual only if they are based directly on traditional reasons or are based indirectly on them via mutual agreement on a set of nontraditional reasons. Given the second, even

though, as Baier says, moral reasons will have been designed to restrict and override egoistic reasons in interpersonal contexts, they will be able to do so only if they have force. Baier thinks they can be said to have non-contingent force, but he is mistaken (see Section 4). Because moral reasons are nontraditional in nature, their force must be con-

tingent in some way on the force of traditional reasons, and, since I' is more plausible than its alternatives, on egoistic ones. This means moral reasons are not as such superior to egoistic or ethical ones. They are

superior only in the context of an ongoing activity of interpersonal reasoning, which, while usually present, is itself a sub-context of the more basic ethical activity of personal reasoning. And while this activity is in fact more fundamental, it rarely presents itself as such but rather normally surfaces as a subpart of interpersonal reasoning (under, for instance, the guise of prudence)22 allowing that activity to give the false impression that it is unconditional and that its reasons have absolute force.

9

The argument of the last three sections has been that ethical egoism as expressed by I employs a use of 'ought, all things considered' that is consistent in all its implications. The damaging in- ferences cited by critics follow only when these terms are embedded in a nonegoistic system and, consequently, they cannot be used in a refutation of I without begging the question. Thus I cannot be dismiss-

22 It is now misleading to say 'prudential' theory and 'prudential sense of "ought'" because all our ordinary talk about prudence occurs within a strong, interpersonal context, where it is implicit that such reasons are limited in force. Morality, that is, recognizes prudence as a virtue among other virtues, and treats it accordingly. Prudence is founded, for instance, not on I', but on the moral obligation to look after our own well-being as well as the well-being of others. Hence it does not deny the soundness of IT. Ethical egoism, while related, is not prudence in this sense. Baier, for instance, frequently refers to egoism as that part of practical discourse which is "prudential" (see "Ethical Egoism and In- terpersonal Compatibility," pp. 362-67) and it may be this misclassification which makes him so convinced that moral reasons are always superior to egoistic (or ethical) ones, since moral reasons are always superior to (merely) prudential ones.

354




ed as logically incoherent. This means that egoism can in fact serve as the foundation of morality when the latter is understood as a system of nontraditional reasons.

There is something to be gained and something to be lost in doing this. What will be gained is a correct understanding of morality. This is what one desires when one begins to ask "Why?" and is really the ma-

jor motivation of the egoist as an ethical theorist. It is natural to think of ethical egoism as a teleological theory, and of course insofar as it says that M ought to do X in S because X partially achieves a certain good, namely M's overall interest, this is so. But the theory does not, philosophically, begin with such goods. It begins, rather, with a desire to avoid what it senses as an irreparable loss, the sacrifice of one's life or interests to the interests of others, and with the Kantian (and not the

Platonic) question about this desire, "What is the most reasonable

thing for me to do in this situation?" As a moral investigation, it is thus the search for moral truth in the form of conclusive reasons and therefore often keeps one on a sceptical path. Along the way, the

egoist's desire to avoid personal sacrifice becomes confirmed by the

discovery that there is no nonconstructed objective, common good and this is given expression in Principle I. It remains, however, that his motivation and goal in this philosophical investigation is not simply to validate his own interests but to be rational.

What will be lost is the unity of this rationality. Practical reason now becomes unified, or materially the same, only within a system of in-

terpersonal reasons; given such a situation, there is then some one

thing which is the reasonable thing to do. But lacking it, reason tells one person to do one thing and this same reason tells another person in the same situation to do something different and perhaps something incompatible. Practical reason itself, to use Kant's image, no longer reflects that harmony of nature which is the basis of a

Kingdom of Ends. For a condition of morality to obtain, therefore, it is not sufficient, as Kant and the Enlightenment tried to maintain, that

everyone simply be rational. People may be completely rational, but this in itself may not be adequate because nature can be intractable

through its prejudicial distribution of interests and resources. Kant had some awareness of the effect on the unity of practical reason caus- ed by the whims of nature which disjoined happiness from virtue and made the rational pursuit of one's own happiness immoral and the moral pursuit of it irrational (because ineffective), for he tried to save this unity by means of the moral postulates of an immortal soul and an omniscient and onmipotent God who would make sure that moral sacrifice received its proper reward. But such postulates are desperate, not rational. When nature plays its games, we must choose between the ethical and the moral, and reason is now on the side of the ethical.

355



Jesse Kalin

Reason then reflects this non-Kantian nature and separates men instead of unifying them, making the Kingdom of Ends a lost hope.

December 1974

356



Canadian Journal of Philosophy - 24HourAnswers...Canadian Journal of Philosophy Two Kinds of Moral Reasoning: Ethical Egoism as a Moral Theory Author(s): Jesse Kalin Source: Canadian

Documents