Section 7

Hare's Implicit Use of Evaluative Principles

I will first summarize briefly Hare's 'derivation' of a utilitarian conclusion from the facts of any given case via the logic of evaluative words, in his Moral Thinking (1981) . I put the word 'derivation' within quotes because it is not a straightforward logical inference from premisses to a conclusion; rather, his 'derivation' means that if one (1) tries to decide rationally what one ought to do in the given case, (2) knowing all the facts and (3) following out the logic of moral judgments, then one will accept a certain 'ought'-judgment, and (4) this 'ought' prescribes an act which maximaizes the satisfaction of preferences of all persons involved in the case. As is well known, Hare distinguishes the critical thinking from the intuitive thinking in moral discourse, and I am here talking only about the critical thinking (which assumes full rationality and sufficient information; I will touch upon the problem of rationality later). Despite the criticism I am going to put forward in the following, I still appreciate Hare's method for justifying an evaluative conclusion in this manner, leaving the gap between Is (description) and Ought (universalizable prescription) as it is.

Let us begin with his simplest model case, the car-bicycle example (Hare 1981, 6.2). Adam wants to park his car but Eve has put her bicycle in the only vacant parking lot; Adam prefers to park, whereas Eve prefers her bicycle to stay where it is, but it is assumed that Adam's preference is stronger than Eve's (this presupposes the interpersonal comparison of preferences). Assuming, further, that Adam has perfect knowledge about all this, and he is ready to decide what he ought to do by critical thinking, what ought he to do? Hare answers as follows:

Since Adam wants to decide what he ought to do, by universalizability he must seek an 'ought'-judgment which is acceptable even if Adam's position and Eve's position are reversed; thus if he is to accept

(a) Eve ought to move her bicycle in order to enable me to park there,

he must be ready to accept

(b) If I were in her position, I ought to move my bicycle.

But for this, he must know what it is to be in Eve's position with her preferences, because the proposed act will frustrate some of her preferences; and because a rational decision must be made in the light of knowledge and logic. Further, this knowledge in the context of critical thinking satisfies what Gibbard (1988, 58) named the Conditional Reflection Principle: 'I cannot know the extent and quality of others's sufferings and, in general, motivations and preferences without having equal motivations with regard to what should happen to me, were I in their places, with their motivations and preferences' (Hare 1981, 99). Hare regards this as a conceptual truth (by virtue of 'know' in the moral context). In short, Adam has to represent Eve's preference for unmoved bicycle within himself by his own acquired preference equal in strength with hers; this preference is a consequence of his knowledge, by the Conditional Reflection Principle.

Then, the problem for Adam is now reduced to a rational decision as regards his own conflicting preferences. And since if his own two preferences conflict the stronger wins, a rational or prudential choice is to satisfy the stronger (here, 'rational' roughly means that the choice survives the criticism in the light of logic and sufficient knowledge). And the same solution applies in our case too: since Adam's original preference is assumed to be stronger than Eve's preference now represented by Adam's acquired preference, the rational solution is to choose an 'ought'-judgment which satisfies (maximaizes the satisfaction of) Adam's overall preference (everything considered), in this case (a). Thus the 'ought'-judgment is justified in terms of a rational acceptance based on facts and logic.

This solution seems very attractive. Unlike Sidgwick, Hare seems to have dispensed with the notion of 'one's good on the whole' or of 'people's good on the whole', thereby avoiding Sidgwick's 'Quantitative Whole'. However, on a closer examination, similar problems reappear (I was still unaware of this in my 1994). First, even in the case of intrapersonal comparison of one's own preferences, the comparison is not always among contemporaneous preferences, often involving conflicting preferences ranging over different times; the question of prudence will lose much significance if we may restrict our attention only to contemporaneous preferences! What is mostly at issue is the problem of diachronic rationality, requiring the comparison of preferences at different times. Hare is of course aware of this problem; that's why he spent many pages (Hare 1981, 101-106, 124) for discussing 'now-for-now' and 'now-for-then' and 'then-for-then' preferences, emphasizing that 'there are obvious analogies between other people's preferences and our own preferences in the future' (124). This clearly shows that, although Hare did not use the notion of 'one's good at a moment', Sidgwick's problem for prudence (rational self-love) reappears in a different form in Hare's theory too. And in view of Sidgwick's discussion, the essential problem is: how to weigh the importance of preferences at each moment?

And this directly leads to the second point. We also have the corresponding problem related to benevolence: how to weigh the importance of one individual's preferences against another individual's preferences? These problems are analogous but independent, as Sidgwick clearly saw. But Hare seems to think these problems can be somehow solved in terms of universalizability, which I now realize is wrong. Although I was quite impressed by Hare's preceding argument, I always felt uneasy about his step from particular preferences to the final preference ('prefertence all told') for choosing an 'ought'-judgment. There is the problem of 'correct representation' of another person's preferences (or of one's own at different times), in the first place. This problem seems quite analogous to the measurement of length, which has to postulate a unit of length and a definition of congruence at different locations; it does not make sense to postulate the existence of the absolute length, independent from these postulates. In our preference case, we have to have, likewise, a criterion for telling whether this preference is the same as , or greater than, another preference, in strength, if two are not contemporaneous within oneself, or if two belongs to different individuals. Intrapersonal comparison, as well as interpersonal comparison, depends on such a criterion. Hare is aware of this, when he says: 'We imaginatively suppose that we could have the choice of having one of these experiences or the other, and form a preference now' (Hare 1981, 125); thus introducing the present (informed) preference as the criterion of the comparison (for a clearest statement of Hare's problem, see Griffin 1988, 76).

Hare has provided a more systematic statement on this point:

The problem of ordinality versus cardinality is more difficult. On the face of it it looks as if our method does not require us to be able to measure utilities in constant units or with a constant zero point. For it is enough if I, who am making a moral decision by critical thinking, can say 'Jones prefers outcome J1 to outcome J2 more than Smith prefers outcome S2 to outcome S1'. We do not have to be able to say how much more. This is because in our method of critical thinking we are not summing utilities, but, from an impartial point of view, which treats Jones' and Smith's equal preferences as of equal weight, forming our own preferences between the outcomes. (Hare 1981, 123)

But notice that, in this quoation, Hare is already assuming two things: First, we already know whether or not a preference of Jones' is equal (in strength) to a preference of Smith's; Second, we form our own preference from an impartial point of view. If we wish to provide a criterion of interpersonal comparison of preferences, the First begs the question. And the Second seems to be nothing but a restatement of Sidgwick's Principle of Rational Benevolence, although it is formulated in terms of preferences. Recall that, according to Sidgwick, if we treat one's own good impartially through time, it is the Principle of Prudence (Rational Self-Love), and if we treat different persons's good impartially, that is the Principle of Benevolence (and let me add that Sidgwick's notions of good and pleasure already incorporate preferences; see Sidgwick 1907, 110-1, 127). As Hare himself mentions after the preceding quotation, cardinal (quantitative) utility is obtained if we accumulate enough (infinite, though) preferences. And, further, let us notice that Hare spells out the content of this impartiality, in his reply to Griffin in Seanor and Fotion (1988), as follows:

They [informed preferences] are formed 'from scratch . . . from an understanding of the objects before us'. With this I agree; but (...) 'understanding' means understanding of their nature, not of their objective or generally accepted value; and though the judger judges from his own particular point of view, the exclusion of appeal to his own antecedent preferences or values means that any judge who truly represented to himself the situations-cum-preferences would form the same order of preference. This order of preference is thus objective in the sense that all rational informed judges, judging from a universal point of view which excludes their own other preferences, will share it. (Hare 1988, 238)

Notice how close Hare comes to Sidgwick in the last part of this quoation. In a word, if we have full information and form a preference, excluding our own personal preferences, that's a preference from an impartial point of view, and it is the same for everyone with the same conditions. I just wonder where Hare has established that this in fact holds! I cannot agree with him that 'the exclusion of appeal to his own antecedent preferences or values means' that everyone has the same ordering; this is not a proof, but merely an assumption.

It is of course permissible to call this impartiality (with respect to preferences) another kind of universalizability. But as I have already pointed out by analogy to the measurement of length, this new kind of universalizability is concerned with new concepts, i.e. the strength of preference, which enables us to compare different people's preferences (assuming, for the sake of argument, that the strength is somehow ascertained), clearly distinct from the Principle of Justice or from 'the same facts, the same Ought' sort of principle. In forming our preferences (i.e. preferences all told) in critical thinking, it is logically possible to assign a different weight to my own or to other's preferences (that's why Hare has to mention 'an impartial point of view'); even if I represent other's preferences by my own (which is the job of the Conditional Reflection Principle), I can still distinguish them from my original preferences (notice that Hare is assuming Archangelic knowledge in critical thinking) so that it is up to me with what weight I should treat them. Harsanyi is quite clear about this when he discusses the 'conversion ratios' (between different utility functions) when we form a social-welfare function from individual utility functions (Harsanyi 1977, 57). Even if each individual has a cardinal utility function, one's unit of utility may be different from another's, and that's why we need conversion ratios for a social-welfare function. Notice that such ratios amount to weights of preferences. Thus, whether we call the impartiality in question 'universalizability' or 'benevolence' or whatever, the fact remains that it is non-tautological and needs substantive justification; you may attribute it to the concept of morality, but then the question 'why should we be moral?' becomes heavier and nothing is improved.

In short, even if we allow Hare to assume the 'correct representation' of others' preferences, including the strength of these preferences, he has to face the following dilemma: Either (1) the strength of each individual's preference does not have a common scale or unit, or (2) it has; but if (1), Hare cannot meaningfully talk about an impartial treatment of everyone's preferences, and if (2), he still has to assume a unique way to assign a weight to each individual, i.e.the impartial weight, which cannot be justified on logical grounds alone. Further, if in case (2) Hare tries to derive the impartial weight from rationality, Hare is no different from Sidgwick.


Griffin 1988, 76:

Suppose I know a lot about your experiences. I can correctly, fully, even vividly, represent them to myself. But my being able to represent to myself the feel of your experience is, in a way, too much of a good thing. It leaves me with one perception of the feel of my own experience and a second perception of the feel of yours. There is still a gap. How do I get the two experiences on to one scale?

Sidgwick 1907, 110-1:

It would seem then, that if we interpret the notion 'good' in relation to 'desire', we must identify it not with the actually desired, but rather with the desirable:----meaning by 'desirable' not necessarily 'what ought to be desired' but what would be desired, with strength proportioned to the degree of desirability, if it were judged attainable by voluntary action, supposing the desirer to possess a perfect forecast, emotional as well as intellectual, of the state of attainment or fruition.

Sidgwick 1907, 127:

but, for my own part, when I reflect on the notion of pleasure,----using the term in the comprehensive sense which I have adopted, to include the most refined and subtle intellectual and emotional gratifications, no less than the coarser and more definite sensual enjoyments,---the only common quality that I can find in the feelings so designated seems to be that relation to desire and volition expressed by the general term "desirable", in the sense previously explained. I propose therefore to define Pleasure----when we are considering its "strict value" for purposes of quantitative comparison----as a feeling which, when experienced by intelligent beings, is at least implicitly apprehended as desirable or----in cases of comparison----preferable.


To 6. Universalizability and Benevolence

To 8. Conclusion / Bibliography

[back to Sidgwick&Hare index][back to Uchii index]


July 20, 1998. Last modified April 17, 2006. (c) Soshichi Uchii

webmaster