Pigou s Ethics and Welfare

Transcription

1 Pigou s Ethics and Welfare Satoshi Yamazaki (Kochi University: yamazaki@kochi-u.ac.jp) ⅠIntroduction Although Arthur Cecil Pigou ( ) is generally considered the founder of welfare economics and a venerable authority in the discipline, economics was not his main interest at the beginning of his academic life. It can be easily realized from his first book, Robert Browning as a Religious Teacher (Pigou 1901a), that his original focus of concern involved the philosophical and ethical field. He came to economics later via philosophy and ethics (Robinson 1971, 814), and thereafter his academic life was essentially dominated by it. This might explain why his early thoughts, especially his ethics, have rarely been fully examined by philosophers or economists 1. In his inaugural lecture at Cambridge in 1908, however, Pigou clearly stated: Economics and Ethics are mutually dependent. The practical art of social service requires them both. The first is hand-maid to the second. It is essential for the full fruitfulness in practice of either of them that the other should progress. May I add that it is an urgent need for the economist that he be also a student of Ethics? (Pigou 1908b, 13-4). We may naturally expect that Pigou himself had acquired some ethical theory 2. But what is his ethical theory? What is his theoretical position in ethics? As I noted above, there have been few comprehensive studies on it. Is it utilitarianism? If so, what kind of utilitarianism? Is it the traditional hedonistic type or some other type? This is the first problem to be tackled in this article. Unlike conventional understanding 3, I will show that Pigou s utilitarianism differs from that of Bentham or J.S. Mill or Sidgwick in some crucial aspects 4, and that Pigou is to be regarded as an ideal (non-hedonistic) utilitarian, a * I am grateful to Professor Tiziano Raffaelli (Pisa University) and Professor Richard Arena (Nice University) for their helpful comments on the previous version. 1 His bibliography may also have something to do with this. Pigou did not publish any systematic study of ethics. Yet, this does not mean that he was not well acquainted with it. Indeed, his commitment to the subject is evident from the emphasis he placed on the significance of ethics and the interdependence of ethics and economics. 2 For instance, in 1900, Pigou achieved the top grade in the Moral Science Tripos at Cambridge University. 3 Edgeworth (1913), Schumpeter (1954), Blaug (1978), Hutchison (1953), O Donnell (1979) and Collard (1981; 1996) regard Pigou as a scholar who is located in the stream of Benthamite or Sidgwickian traditional utilitarianism. However, they do not consider Pigou s thought from the strictly ethical or philosophical point of view. The arguments seem to be merely supplementary observations incidental to economics. 4 Despite some different elements among them, Bentham, J.S.Mill and Sidgwick are classified as hedonistic utilitarians. For this, refer to Frankena (1973), Uchii (1988) and Quinton (1989). 1

2 position represented by G.E. Moore and H. Rashdall. Furthermore, we know that Pigou s economic welfare is a part of welfare as a whole, but what then are the substantial contents of his non-economic welfare? Again, we also know that non-economic welfare is indirectly related to economics through economic welfare. Nevertheless, is there any direct relation between them? This is the second problem I will address. In general, economic welfare is obtained through the desire satisfaction principle. I will show the existence of the need satisfaction principle in Pigou s thinking, and that he depended on this principle at important contexts in his welfare economics. Furthermore, those parts of welfare that are brought about through the satisfaction of needs do not necessarily coincide with economic welfare. Hence, contrary to conventional understanding, we can see that some parts of non-economic welfare are intended to be promoted directly in his economic perspective. Before the main argument, it is advisable to state the general definition of utilitarianism as a moral principle 5. Basically a moral principle requires (1) a theory that clarifies the good thing in itself and (2) a theory that clarifies what conduct is right or how we ought to behave. Generally speaking, on the one hand, all kinds of utilitarian theories include teleology or consequentialism in (2), while on the other hand, in (1), various versions of the theory of value can be assumed: hedonism, desire satisfaction, preference, and so forth. Therefore, various types of utilitarianism are derived according to their different approach to the theory of value. However, in (1), utilitarian doctrine does not include all sorts of value as its ultimate goal. To explain briefly, public values such as justice or fairness are not regarded as values in themselves in utilitarian theory. They are defined as good insofar as they promote individual good (welfare, happiness and the like). This can be called the notion of methodological individualism in the theory of value. The first task of this article is to identify Pigou s ethical position by examining his documents on the basis of (1) and (2). We will discuss Pigou with regard to (1) in section II and with regard to (2) in section III. In section IV, the relation between his ethics and welfare economics will be re-examined. So far the study of Pigou s ethical thought has been carried out referring mainly to 5 For this point, refer to Uchii (1988, 164-5). Indeed, the definition of utilitarianism ranges from the wider to the narrower sense. As for the definition in the narrower sense, Sen and Williams (1982), for example, define utilitarianism as the combination of consequentialism, welfarism and sum-ranking. On the other hand, with regard to the definition in the wider sense, Riley (1988), Kelly (1991) and Singer (1993) among others look upon utilitarianism as a compound moral principle. Even if we assume the definition in the wider sense, the theoretical distinction between utilitarianism and deontology, such as justice or fairness, will not be unjustified. The distinction is certainly made valid by methodological individualism in the utilitarian theory of value. That is, the utilitarian doctrine never recognizes intrinsic goodness of justice or fairness. Anyway, this point might relate to the interpretation of the utilitarian principle itself. 2

3 his Wealth and Welfare and The Economics of Welfare. But, as Pigou himself admits, those ethical elements revealed in them are only part of his ethical idea. Here, it is important to note that Pigou s basic as well as relatively systematic ethical thought is rather shown in Some Points of Ethical Controversy (Pigou 1907b) and The Problem of Theism, and Other Essays (Pigou 1908a), which have rarely received attention. For the reconstruction of his ethics, it is indispensable for us, first of all, to analyze and to identify the inner logic of those documents. And it may be worth comparing Pigou with Moore in passing. As will be ascertained, both held similar views in their criticism of hedonism, in the denial of metaphysical ethics and in the adoption of intuitionism in metaethics, pluralistic value theory and teleological principles. Despite any point of difference between Pigou and Moore, they share interesting points of ethics to be examined previously some comparisons between the moral philosophy of Moore and another well-known Cambridge economist, J.M. Keynes, were made 6. However, very little consideration has thus far been given to the relation between Moore and Pigou. Unlike Keynes 7, Pigou did not seem to attend the circle run by Moore or to have personal contact with him. Hence, any reference to the relationship between Moore and Pigou is necessarily concerned with their theoretical aspects in moral philosophy. ⅡPigou s method of ethics and his theory of value In the introduction of Some Points of Ethical Controversy (Pigou 1907b), Pigou says as follows: The word Good is applied in ordinary life to two entirely distinct classes of things. It is sometimes used for good absolutely and in itself, and sometimes for useful as a means to promote something thus absolutely good. My paper has nothing to do with good in this second sense. It is concerned exclusively with certain points of controversy touching things good in themselves (Pigou 1907b, 99). After stating the above, Pigou presents three points at issue. The first is the method by which ethical inquiry ought to be pursued, namely, metaethics. The second is the contents and the qualities of intrinsic good. The third is the way in which the good things of one being should be related to those of others. For the purpose of examining Pigou s normative ethics (not metaethics), the latter two points will form the core. Nonetheless, the first point seems attractive when we compare Pigou and Moore. As described later, it is through the consideration on this first point that the meaning of Pigou s exposition of the concept of 6 For example, refer to Shionoya (1991) and Dostaler (2007). 7 On the relation between Pigou and Keynes, see Ambrosi (2004) and Aslanbeigui (1992), and on the biographical aspect, see Johnson (1960) and Saltmarsh and Wilkinson (1960). 3

4 welfare and his citation of Moore at the beginning of Wealth and Welfare can be elucidated. 1 Method of ethics In relation to the first point, Pigou argues the method by which the question, what is good? can be answered. However, we must understand that good relates not to a property that we assert belongs to a thing, but to a thing to which good as a property belongs 8. To answer the question, he refers chiefly to two major principles: one is the a priori method of deduction from the nature of things and the other, the method of direct perception [intuition] (Pigou 1908a, 81). Now, let us look at what views Pigou expresses on each method. Pigou mentions T. H. Green as a representative of the a priori method and examines his metaphysical ethics. According to Pigou, in Green s theory, it is argued that the human spirit is timeless in terms of metaphysics and based on this, what is not timeless, such as happiness (a temporal thing), can never be good in itself. Moreover, Pigou indicates that certain scholars assert, without any demonstration, that the good must be one (Pigou 1908a, 81). Pigou counters those scholars, including T. H. Green, who espouse the a priori method, declaring [a]s applied by Green, it seems to contain a definite formal fallacy (Pigou 1908a, 81). [F]or, even if it be true that the spirit of man is timeless, is there not a logical chasm between the premise that something is and the supposed inference that something else is good? (Pigou 1908a, 81: italics are original). As a result, to him, the whole method of inquiry, which seeks to determine a priori what things must be good seems mischievous and idle (Pigou 1908a, 81-2). Thus, the a priori method as metaethics is completely disapproved of by Pigou. Moreover, it is understandable that he also refuses naturalism in metaethics for the same reason. From his youth, Pigou was cautious about the logical gap between positive and normative judgment (Pigou 1907c, 359). The essence of Pigou s criticism is shared by Moore. Moore explains as follows: by metaphysical Ethics I mean those systems which maintain or imply that the answer to the question What is good? logically depends upon the answer to the question What is the 8 In this connection, let me quote a passage from Aslanbeigui (2002): For Pigou, welfare and good are cognates. Like G. E. Moore s concept of the good, Pigou s concept of welfare is indefinable: welfare refers to unanalyzable states of conscious life. Unlike Moore s concept of the good, Pigou s welfare can be brought under the category of greater and less (Aslanbeigui 2002, xxxv: italics mine). However, here, it does not seem that discrimination between good and the good, that which is good is made. States of conscious life correspond to the good in the latter sense; therefore, the good is analyzable and definable. Moreover, Moore s concept of the good can also be brought under the category of greater and less, e.g. more good (Moore 1903, 148). Therefore, the difference between Pigou and Moore pointed out in the above would not be recognized. 4

5 nature of the supersensible reality? All such systems obviously involve the same fallacy the naturalistic fallacy (Moore 1903, xviii: italics are original). To hold that from any proposition asserting Reality is of this nature we can infer or obtain confirmation for, any proposition asserting This is good in itself is to commit the naturalistic fallacy (Moore 1903, 114). Then, what is the method espoused by Pigou to answer the question regarding good in itself? He replies: [t]he only way to know whether anything is good is by looking at it. We are confronted with the world of reality and of imagination and we perceive some things to be good and some bad, just as we perceive that some are yellow and others red (Pigou 1908a, 82). The famous philosopher who gave the example of yellow in an argument about the definition of good and metaethics is Moore himself (Moore 1903, Ch.1) 9. Furthermore, at the beginning of Wealth and Welfare, we find a depiction, the significance of which has been seldom explored. There, Pigou first quotes the famous passage from Moore s Principia Ethica: If I am asked What is good? my answer is that good is good, and that is the end of the matter. Or, if I am asked How is good to be defined? my answer is that it cannot be defined, and that is all I have to say about it (Moore 1903, 6). Welfare, Pigou immediately continues, means the same thing as good. It, too, cannot be defined, in the sense of being analysed. indeed, it is the chief task of ethics to say, whether and in what way, particular things belong to welfare (Pigou 1912, 3: italics mine). Pigou identifies the concept of welfare with that of good. And he, as well as Moore, claims that welfare (good) cannot be defined in the sense of being analyzed. Therefore, in Pigou s thinking, it is implied that there is no way to define the good, that which is good 10, without direct perception, namely, intuition 11. As described above, for Pigou, welfare and good are mutually interchangeable concepts. It follows that the argument of good and that of welfare are two sides of the same coin, and that those propositions which hold true for good are also true for welfare. Indeed, as is shown below, Pigou defines the good as well as welfare as states of conscious life in his Wealth and Welfare and Economics of Welfare (Pigou 1912, 3; 1952a, 10). Thus, Pigou espouses intuitionism in metaethics. Particularly, as we can see in the 9 It is indicated that if we perceive that something has the attribute yellow, we do so directly without any deduction or inference. That is, some proposition which must be simply accepted or rejected, which cannot be logically deduced from any other proposition (Moore 1903, 143). 10 It is essential to distinguish between good and the good (which has intrinsic value). Moore (Moore 1903, 8-9) as well as Pigou says that only the former cannot be defined. 11 This indefinability of good or welfare logically leads to the statement: propositions about the good (welfare) are all of them synthetic and never analytic (Moore 1903, 7). 5

6 reference to Moore, his doctrine of the indefinability of good and the method of intuitionism based upon that indefinability play a fundamental part in Pigou s ethical reasoning. Intuitionism itself has a long tradition in British moral philosophy, and Sidgwick, Moore s teacher, pointed out the indefinability of good before Moore, yet, the student was more influential than his teacher 12. Harrod notes that Sidgwick s arguments, however, are not so pointed as Moore s, nor collected into so formidable a battle array; they are to be found scattered about in his large volume on The Methods of Ethics; and his style is altogether less readable (Harrod 1951, 76). Although, both Pigou and Moore learned the basics of the discipline from Sidgwick s writings (Pigou 1908a, viii), it is no less interesting to see the difference between them Pigou s theory of value Now, let us leave the first point (metaethics) and turn to the second point: substantial contents of the good. It will show Pigou s theory of value in normative ethics. On the description of contents of good things in themselves, Pigou says first of all: Many ethical thinkers hold that the only good things are states of conscious life. If this view is put forward as a necessary universal proposition, those who believe that our knowledge of what is good is obtained only through particular perceptions cannot accept it. They have no ground for holding that all the good that there is has fallen within their perception, and, apart from such ground, it is impossible for them to make any exclusive statement about good. They are entitled, however, to say and in this I am among their disciples that states of conscious life are the only good things of which we have present knowledge. This view is disputed by Mr. G. E. Moore and other writers who accept his important doctrine organic good. This doctrine does, indeed, in some instances appear exceedingly plausible (Pigou 1908a, 82-3) In this regard Shionoya made a remarkable statement. The most crucial philosophical problem in Sidgwick s intuitionism was to explain the rightness of ultimate ends in terms of rationality. Since rightness is concerned with the optimum means to given ends, it is reductively explained by the goodness of ends. But where rightness applies to ends as such, it is ultimate and unanalysable. It is here that Moore became famous for the doctrine of the indefinability of good. Unlike Moore, Sidgwick tried to explain the rightness of ends in light of rationality, and put forward, as self-evident axioms, what might be called a structure of rationality in the phase of philosophical intuitionism (Shionoya 1991, 23). 13 Although there has been some mention of the similarity between Sidgwick and Pigou in terms of ethics, concrete and detailed studies in regard to this are rare. 14 Moore himself uses organic relation or organic whole rather than organic good. But, this is a negligible terminological difference. Incidentally, other writers mentioned in this quotation seem to include some members of the Bloomsbury circle at that time. On this, see Levy (1979). 6

7 We can confirm several matters from this statement. Pigou puts it in question that the view the only good things are states of conscious life is put forward as a necessary universal proposition. This view seems to him a kind of a priori or metaphysical proposition. As a matter of fact, we do not perceive all the objects in the universe. Hence, it follows that proclaiming the only good things are such and such as a universal consideration is contrary to truth. It is, to Pigou, a dogmatic judgment and a bad example of metaphysics. According to Moore, in order to answer What is the good? we must know both what things there are in the Universe and also which of them are good. Upon this question then our Metaphysics would have a bearing, if it can tell us what is real. It would help us to complete the list of things which are both real and good. But to make such a list is not the business of Ethics (Moore 1903, 119). Apart from a metaphysical viewpoint, Pigou argues that within the region of present experience, states of conscious life are the only good things. Also, Moore does not admit the positive role of metaphysics in ethics. As an aside, Pigou argues that there is some obscurity in Moore s organic whole (Moore 1903, 27-36). According to Pigou, although, for instance, it looks apparent that the state of a man who enjoys a daub believed to be a beautiful picture is worse than that of one who enjoys a really beautiful picture, this plausibility is due to an ambiguity. He explains as follows. When a man enjoys a daub believing it to be a beautiful picture, he may either recognize what its quality really is and falsely regard it to be beautiful, or believe that its quality is different from what it is and precisely judge the imaginary quality which he has conceived instead of the real quality to be beautiful. In the former case, according to Pigou, things are worse for the man s error, but the error belongs entirely to relations within his consciousness, never to the relations between his consciousness and an object. On the other hand, in the latter case the error concerning the relation between his consciousness and an object Pigou does not admit that things are worse for the error. He continues, I reject the doctrine of organic goods so far as it conflicts with the view that goodness within the region of present experience belongs only to states of conscious life (Pigou 1908a, 83). It seems to him that Moore s doctrine judges things in the latter case to be worse for the error, but the essence of the doctrine of organic whole does not consist in the point as such. I will return to this doctrine later. Now, let us examine what kinds of conscious life Pigou thinks are good. Certainly, this is a formidable question. The only procedure open to us is to look out the people we know or can imagine and try to judge directly of their goodness as concrete wholes. When we have brought to bear on this task a sufficiently wide experience, it is possible that we may be able to disentangle the various elements in their consciousness upon which their goodness seems to depend, together with the nature of this dependence. If that could be done, 7

8 our notions would be to some extent simplified and clarified. The most daring attempts in this direction that have hitherto been made select, from among the various elements of which any total state of consciousness is composed, some single element, and declare that to be the only element affecting the goodness of the state. All the other elements may vary in any direction and to any extent without altering this, but the least variation in the chosen element does alter it (Pigou 1908a, 84: italics mine). The way Pigou argues above is to analyze elements of the total state of consciousness and isolate each element, then to reflect whether each one affects the goodness of the state. It is the process through which we can see something is good or bad in itself, or good as a means. This is, in principle, analogous to Moore s method of absolute isolation (Moore 1903, 188). According to Moore, in order to ascertain what has intrinsic good, it is necessary for us to isolate an element from others that compose the whole. The method which I employed in order to shew that pleasure itself was not the sole good, was that of considering what value we should attach to it, if it existed in absolute isolation, stripped of all its usual accompaniments. And this is, in fact, the only method that can be safely used, when we wish to discover what degree of value a thing has in itself. The necessity of employing this method will be best exhibited by a discussion of the arguments used by Prof. Sidgwick and by an exposure of the manner in which they are calculated to mislead (Moore 1903, 91). as well as to avoid confusion between means and end, it is absolutely essential to consider each distinguishable quality, in isolation, in order to decide what value it possesses. Prof. Sidgwick, on the other hand, applies this method of isolation only to one element in the whole he is considering. He does not ask the question: If consciousness of pleasure existed absolutely by itself, would a sober judgment be able to attribute much value to it? (Moore 1903, 93: italics are original) As Pigou himself pointed out, traditional utilitarianism regards pleasure as the only element that affects the goodness of the states of conscious life. Pigou, taking up Sidgwick, argues that he admitted the goodness of the good will only insofar as it made a contribution to pleasure (Pigou 1908a, 84-5). Briefly, even though various scholars have claimed various sorts of intrinsic good (e.g. love, the good will, pleasure, etc.), [t]he point is that each of these schools regards the goodness-in-itself of any total state of consciousness as dependent upon one element alone (Pigou 1908a, 85). In fact, Sidgwick 8

9 maintains that things such as volition or cognition can be good, but as a means to the sole intrinsic good, pleasure (Sidgwick 1907, 395-7). Moore refutes that: It seems to me, then, that if we place fairly before us the question: Is consciousness of pleasure the sole good? The answer must be: No. And with this, the last defence of Hedonism has been broken down. In order to put the question fairly, we must isolate consciousness of pleasure (Moore 1903, 95). In this respect, Pigou follows Moore. This is crucial to our understanding of his position in ethical thought. In a way, similar to Moore s method of isolation, Pigou judges, for instance, the intellectual power to be good as a means, while things such as pleasure, the good will and love to be good in themselves. Of course, these are not all. He admits that there are some other elements with intrinsic value (Pigou 1908a, 86). Therefore, the goodness of any conscious state is, to use a mathematical phrase, a function of several variables (Pigou 1908a, 87). In typical utilitarianism, pleasure is the only variable. However, Pigou claims the plurality of variables, which clearly means a pluralistic value theory in ethics. We may call this the functional theory of good. Yet, if it is called a function at all, those elements which are regarded as independent variables of the function (e.g. pleasure, satisfaction, the good will) must be measurable. However, as Pigou explains, strictly, for magnitudes of any kind to be measurable means that a unique and reciprocal correspondence, a one-one relation, can be established between the magnitudes in question and cardinal numbers. Pleasures, satisfactions, utilities are intensive magnitudes and are not measurable. Nevertheless, an intensive magnitude may sometimes be correlated with an extensive magnitude and so may be capable of being measured (Pigou 1951, 289). Briefly, the functional theory of good seems to present a figurative rather than a strict explanation useful for understanding. Pigou then considers the character of this function under the following headings. First, whether the dependent variable will always be enhanced when the value of any or all of the variables is increased. Second, whether the sign of any of the independent variables governs that of the dependent variable. And third, if the previous two questions were answered negatively, whether any further proposition concerning the character of the function can be established. As to the first question, it is conceivable that in all cases the dependent variable, that is to say, the goodness of a total state, is increased by an increase in the quantity of pleasure or happiness contained in it. However, Pigou objects to this. For instance, if one derives pleasure by carrying out unsound deeds, of which one is aware in one s own mind, this cannot lead to the betterment of the total state. Furthermore, he considers such cases as enthusiasm and intensity of love, yet they are thought to be more or less similar to the case of pleasure (Pigou 1908a, 87-8). That is, increments of any elements do not 9

10 necessarily bring about goodness of the total state 15. Then, the answer to the first question proves to be negative. The second question is whether it is true or not that the sign of any of the variables governs in any general way that of the function. With regard to this, Pigou remarks: It may be held that pleasure is essential to a good state in the sense that any predominance, however slight, of pain over pleasure must always render the state as a whole bad (Pigou 1908a, 88). Put another way, whenever the sign of a variable of pleasure is negative, so is that of the function. Obviously, this is the typical view of traditional hedonistic utilitarianism and Pigou himself contradicts it. Besides, as for the good will, the case is the same as for pleasure (Pigou 1908a, 88-9). Accordingly, the answer given to the second question is also negative. Thus far, it follows from these two points that Pigou s views are separate from the hedonistic utilitarian doctrine, in the mainstream of which were Bentham, Mill and Sidgwick. In this respect, both Moore and Pigou hold entirely the same opinion. it is extremely important to decide rightly what kinds of consequences are intrinsically better or worse than others. And it is on this important point that the theory in question [hedonistic utilitarianism] seems to me to take an utterly wrong view. It maintains that any whole which contains more pleasure is always intrinsically better than one which contains less, and that none can be intrinsically better, unless it contains more pleasure (Moore 1912, 144: italics are original). Lastly, to the third question: if the two preceding questions should be negatively answered, [c]an any general proposition be laid down as to the nature of the function on which the quantity of goodness present in any state of consciousness depends? (Pigou 1908a, 89). According to Pigou, we can recognize the possibility of the cases given below. Let G, x and y denote the goodness of the total state, one element and another element, respectively, then, under certain circumstances, we have Δ(ΔG/Δx)/Δy > 0. That is, the effect of x on G is enhanced by the increment y; for example, in the case that x is happiness and y is virtue (Pigou 1908a, 89). This proposition is interesting since its implication is similar to some aspects of Moore s idea of organic wholes. Moore uses the term organic wholes in a restricted sense, signifying the value of a whole does not necessarily coincide with the sum of the values of its elements. It is certain that a good thing may exist in such a relation to another good thing that the value of the whole thus formed is greater than the sum of the values of the two good things (Moore 1903, 27-8). Suppose a whole is composed of A and 15 As we can see below, this point clearly leads to the argument of the controversial relationship between economic welfare and total welfare. 10

11 B; let the intrinsic value of A, B and the whole be V A (A), V B (B) and V(A, B), respectively 16. Then, the theory of organic wholes indicates V(A, B) V A (A) + V B (B), i.e. V(A, B) > V A (A) + V B (B) or V(A, B) < V A (A) + V B (B) (of course,= is also conceivable). If so, on what principle does it occur? The exposition by Moore himself does not seem clear. However, Pigou s functional theory of good described above is expected to yield it. An addition to one element might, via direct channels, improve the state of the whole. Further, the state of the whole might be also improved since the addition makes the effect of another element on the whole more powerful. This is the essence of Pigou s functional theory. Based on his theory, we can hypothesize the following: as a simple example, let the separative evaluation of A, B be V A (A) = A, V B (B) = 2B and the value of the total state V, V(A, B) = (2 + A)B + A, respectively. The value of the whole containing both A and B is not simply V A (A) + V B (B). Nonetheless, in the absence of any other elements except A (B) (A or B = 0), the whole V(A, B) must be identical to the single form V A (A) (V B (B)). Now, with the addition of A (>0), ΔV/Δ B = 2 + A at V(A, B) becomes larger than ΔV B /ΔB = 2. This is what Pigou implies. If A = 0 and B = 2, then V = 4,and if A rises to 1, then V = 7. In this case, the total difference, ΔV, is 3. Besides it is greater than 1, the contribution made through the direct channels of A. It is the addition to A that makes the coefficient of B more efficient. Let me summarize the main points made thus far. Moore claimed that his method of absolute isolation is the only way by which we can identify the things having intrinsic value, and in what degree, and can distinguish between the intrinsic good and instrumental good. Hedonists hold that pleasure is the only good thing and every other thing can be (instrumentally) good insofar as it is useful to the former. Is it true that the goodness of everything but pleasure is no more than instrumental? The boundary between hedonists and non-hedonists is fixed according to the answer to this. With that method of isolation, perhaps those things which at first seem to be intrinsically valuable may prove to have only instrumental value, and the reverse is also conceivable (Moore 1903, 91; 187). According to Moore, because of failure in the application of that method, Sidgwick committed an error of judging the intrinsically good things to be instrumentally good. That the value of a pleasurable whole does not belong solely to the pleasure which it contains, may, I think, be made still plainer by consideration of another point in which Prof. Sidgwick s argument is defective (Moore 1903, 94). As for Pigou, a thing like intellectual power is neutral with respect to the goodness of the total state: In itself, however, when its effects are abstracted, it does not seem to me to affect in any way the goodness of the states. It is said to be a means to good. However, 16 The reason why each function is different from the others is that each element must be evaluated separately in accordance with the method of absolute isolation mentioned before. 11

12 regarding other things such as pleasure, the good will and love (needless to say, he admitted that there are more elements), different evaluations (as intrinsic good) are given (Pigou 1908a, 86). Pigou, as well as Moore, clearly distinguished between good in itself and means to good, besides espousing the plurality of the former. From what his functional theory of good shows, several elements holding intrinsic value exist and compose a whole (the state of conscious life), which also has intrinsic value. Moreover, the sum of the value of each element does not necessarily equal the value of the whole itself, which can virtually be considered as Moorean organic wholes. Incidentally, there are some differences in Moore s opinion as stated in Principa Ethica and Ethics. In the former, he recognizes intrinsic value of beauty itself whether or not human beings contemplate or enjoy it (in this point, Moore and Pigou were different from each other). In the latter, however, it does seem as if nothing can be an intrinsic good unless it contains both some feeling and also some other form of consciousness (Moore 1912, 153: italics are original) 17. Both contradict each other (Scarre 1996, 117). It is said that the final position of Moore s theory in ethics is refined hedonism (Baldwin 1990, 132). It confines good in itself to conscious states and refuses goodness of the other things (e.g. beauty itself and objective conditions without human existence). If Moore s final opinion is indeed so, we can see barely any meaningful differences in the theory of the good between Moore and Pigou. According to their opinions, various kinds of consciousness besides pleasure may have intrinsic value and there is a difference in kind or quality concerning even pleasure. Pigou does actually take account of the qualitative distinction of satisfaction (Pigou 1952a, 17) 18. In this respect, their theories are very close to that of Mills. Indeed, Pigou refers to various kinds of intrinsic good in various places. Some of these are: ethical personalities (Pigou 1907a, 981); pleasure, the good will, love, virtue (Pigou 1908a, 86-9); character (Pigou 1901b, 240; 1937, 138); satisfactions, cognitions, emotions, desires (Pigou 1952a, 14); open-heartedness, sincerity, unselfishness, ardour in living, ardour in pursuit, happiness (Pigou 1965, 80-1) 19. Because, as we have seen, Pigou himself thinks good to be identical to welfare, it must be noted that all those elements are parts of welfare as a whole (i.e. a complex of states of conscious life). Particularly, it is 17 It implies that beauty is ontologically dependent on the mode of seeing as well as on the thing seen (Scarre 1996, 117). 18 Then, how should we compare or evaluate various sorts of utility different in quality? It is clear in Moorean and Pigovian theory. Any qualitative difference is reduced to the quantity of goodness. Still, there is nothing for it but to grasp the quantity of goodness by intuition. 19 The following remark seems to represent his overall notion: Mere quantity of life does not present itself to our consciousness as the only good thing; it may not even present itself as necessarily good at all. What we want to discover is the nature of the good life (Pigou 1908a, 124: italics are original). 12

13 worth paying attention to the moral property of the several elements in welfare. We can interpret them as excellence of human character. Shionoya states that the Moorean (non-hedonistic) theory of value can be understood to be that of excellence (Shionoya 1991, 10-3). The same understanding applies to Pigou 20. Schneewind argues that while Sidgwick took account of excellence in the deontological framework, it is possible to regard excellence as a non-hedonistic goal of teleology (Schneewind 1977, 202). His suggestion unexpectedly shows what Pigou as well as Moore was looking for. We should briefly discuss the problem of commensurability 21 and inter-personal comparisons. We can assume Pigou s approval of commensurability, judging from his functional theory of good and his statement that equal amounts of utility should be evaluated equally, no matter to whom it belongs (Pigou 1935, 120). And, the latter point is based upon the thesis that each counts for one and none for more than one, which has been taken as axiomatic by almost every utilitarian 22. Anyway, in such a case, we need some denominator for commensuration. However, in such a theory of value as the non-hedonistic, multi-stranded account of human good backed by Moore and Pigou, the denominator would ultimately be the quantity of good 23, which cannot be grasped except by our intuition, otherwise comparison itself becomes impossible in principle (i.e. incommensurable) 24. Such a criticism, as pointing out that an intuitionist in metaethics can never objectively show the validity of his intuition and thus that ambiguity is unavoidable, is too trite. Pigou himself recognizes it well (Pigou 1908a, 87). Although he resorts to the 20 Since Pigou interprets quality of man in Marshall as states of consciousness that mean welfare (Pigou 1925, 82), we can certainly see that he himself thinks welfare to include so-called human excellence. Pigou knew that Marshall studied Kant s works, and probably grasped what Marshall had in mind with respect to quality of man. 21 We cannot evade the subject of commensurability, when it comes to a pluralistic value theory. This is one of the most difficult problems in ethics or social philosophy. Regarding this, refer to Griffin (1982, 1986), Raz (1986), Anderson (1993). Of these, Griffin shows the affirmative attitude toward commensurability of the good things (Griffin 1982, ). 22 To it, Singer adds his explanation: Ethics requires us to go beyond I and you to the universal law, the universalisable judgment, the standpoint of the impartial spectator or ideal observer, or whatever we choose to call it (Singer 1993, 12). Also, it should be noted that accepting the view of equal respect on the ideal observer does not necessarily mean some of the sacrifices of one to others (Griffin 1986, 168). 23 In this respect, Moore is described as embracing monism (Anderson 1993, Ch.6). The same would be true of Pigou. 24 Concerning Rashdall who also espoused the pluralistic value theory (Rashdall 1907), it is indicated that the feature which forms the greatest attraction of Rashdall s Ideal Utilitarianism is also the source of its major difficulty. Rashdall largely overlooked the fact that if there is no common denominator of human good, it is unclear how we are supposed to evaluate the relative merits of actions productive of different species of benefit. The injunction to maximaise the amount of ideal human existence is virtually a dead letter if there is no principled way of comparing the relative worth of the different ingredients of the good (Scarre 1996, 120). 13

14 intuitive notion concerning goodness in itself at times, Pigou seems to base its validity not on arbitrary intuition of individuals but on a consensus of people s considered (intuitive) judgment (Pigou 1935, 128; 1947, 43; 1965, 1). Nevertheless, we can identify some devices contrived by Pigou. We can suppose a complex whole and one element of it, and two alternatives A and B, which would somehow affect it. Then, we assume that if A is more efficacious to the element than is B, the same thing will be true of the whole. Clearly, comparison of simple factors is less ambiguous. As for the simple component, the more objectively it is measured, the better. Now, according to Pigou, what things fulfill such a condition? According to him, they correspond to economic welfare and welfare as a whole. As we have seen, since welfare is identical to good, it also means the states of conscious life as a complex whole. And economic welfare is a kind of satisfaction as consciousness, and is one element of welfare (Pigou 1912, 3-4). On the relation between these two, he claims: When we have ascertained the effect of any cause on economic welfare, we may, unless, of course, there is specific evidence to the contrary, regard this effect as probably equivalent in direction, though not in magnitude, to the effect on total welfare; and, when we have ascertained that the effect of one cause is more favourable than that of another cause to economic welfare, we may, on the same terms, conclude that the effect of this cause on total welfare is probably more favourable (Pigou 1952a, 20: italics mine). This, Pigou explains, is the premise referred to as unverified probability by Edgeworth. Based upon the premise, the comparison between any two effects on a total complex state (welfare) can be reducible to that between those on its simple component (economic welfare). Without doubt, his ultimate interest is to improve welfare as a whole (Pigou 1952a, 11-2); however, welfare is a thing of very wide range. There is no need here to enter upon a general discussion of its content. A general investigation of all the groups of causes by which welfare might be affected, would constitute a task so enormous and complicated as to be quite impracticable. It is, therefore, necessary to limit our subject-matter (Pigou 1952a, 10-1). This passage shows that Pigou s concept of welfare was very extensive from the beginning and that he did not account for all the components of it, although they were placed in the category of conscious life. Then, even if overall welfare as the ultimate goal is to be enhanced, it is less practicable, for all its components have not been designated. Pigou s welfare concept is open-ended, since it has the possibility of adopting new elements seeming desirable for our life. Hence, in practice, to limit our subject-matter is rational as long as other aspects are considered. To use the terms of Anderson (1993, Ch.3), welfare as a whole concerns global maximizing and 14

15 economic welfare concerns local maximizing. In Pigou s thinking, the former is abstract, while the latter is practical. The reason why we may, in practice, narrow our consideration down to local level is that the local betterment has the prospect of global betterment (Pigou 1912, 11; 1952a, 20). As we have seen, since commensurability among inter-personal as well as intra-personal goods is assumed, the problem of inter-personal comparisons in Pigou comes to be touched upon. On this controversial problem, disputed by scholars like Robbins, Pigou expresses his overall view in Study in Public Finance (Pigou 1947) and Some Aspects of Welfare Economics (Pigou 1951). According to his concluding contention, whether comparison is possible or not is a difficult question, fundamentally, it is not nonsense to say that A is happier than B. Moreover, if the utilities of different individuals cannot be compared, a large part of that subject [welfare economics] is undermined. It finally leads to the wreck of the whole apparatus of practical thought (Pigou 1951, 292). His message is that, even though inter-personal comparisons are formidable or impossible to carry out strictly, the real world is not so conveniently created without them. What is imposed on us is to try endlessly to find better scales which almost all the people could accept. 3 The relation between the good of one being and that of others Now, let us consider the problem of how one s good is related to others good. This has a great deal to do with a core dilemma of normative ethics, which can be stated as follows: There are two views on the moral constituency, some philosophers affirming that the proper goal of action is the good of the agent alone, others that it is the good of mankind in general (Scarre 1996, 118-9). Pigou begins the argument with the question whether the good of A can compete with that of B or C? (Pigou 1908a, 90). Referring to Green again, Pigou points out that there cannot occur a conflict like that in Green s system on the ground that one s good includes that of others there. Briefly, Green considered the self to be public; therefore, his notion of individual good necessarily contained others good. In this respect, Pigou disagrees with him: This view, however, is not really defensible. Whatever may be the case in a world of ultimate angels law, it is evident that, in the actual world, good, even in Green s conception of it, may be competitive. The realisation of A s self does sometimes involve diminished opportunities for the realisation of B s. There are cases in which, so far as experience shows, A s good cannot increase without affecting adversely that of other people (Pigou 1908a, 90: italics are original). Moreover, according to him, this state of affairs leads to very important matters. Only in 15

16 the afterlife would the conflict with each one s good be settled. The fact is that one s own good is liable to conflict with that of others by and large. Here, by referring to Sidgwick, Pigou argues that in his system of moral philosophy; admitting that fact ended up asserting two inconsistent propositions 25. Sidgwick, as he declared, intuited and recognised as equal authority two moral imperatives, both that he ought to pursue his own good regardless of others good, and also that he ought to pursue the good of the whole regardless of his own good. The case is the same with those who hold that A is an end in himself in the sense that his good ought never to be sacrificed to anything outside himself, and who also maintain a like proposition with regard to B. These propositions are only consistent with one another provided that the pursuit of a man s own and of others good dictate to him the same course of conduct (Pigou 1908a, 91). Pigou also adds Sidgwick s pessimistic hope of a future life (or afterlife) for the reconciliation of two inconsistent moral prescriptions 26. The intense criticism of Sidgwick s dualism (egoism and utilitarianism) by Moore is already well known. Pigou also refers to it, yet he, following Russell, does not see Moore s reasoning of this criticism as proper. For all that, he himself never admits Sidgwick s egoism: I am not, however, really concerned to defend Sidgwick here.the contradiction, for the solution of which he hints that a new world may need to be invoked, does not exist at all. I do not perceive that I ought to pursue my own good regardless of others good (Pigou 1908a, 92). Here, if we do not inquire closely whether or not Pigou s understanding hits the mark, his statements make it certain that ethical egoism as a major teleological moral principle was turned down by him 27. In this essential point, his view also differs from that of Sidgwick. 25 These two propositions were, to be exact, called the principle of prudence and the principle of benevolence, respectively (Sidgwick 1907, 497; 507). For a detailed examination of these, see Schneewind (1977), Shionoya (1984) and Schultz (1992). 26 On account of this pessimism, in his later life, Sidwick leaned more toward the esoteric speculations and concluded that immortality was necessary for the reconciliation of these contradictions. For example, Dostaler (2007, Ch.2) briefly describes Sidgwick s life and his company. 27 The situation of conflict between the good things of different people itself does not necessarily mean self-contradiction. Sidgwick intuited the validity of the principle of prudence as well as that of benevolence. Because his two principles possess conditional form, they do not contradict each other at the logical dimension; however, at the practical dimension, they become contradictory to each other, since their prescriptions are rarely consistent. This is what the contradiction in question signifies. The reason why Pigou says that there is no contradiction is that he refuses the principle of prudence and admits only that of benevolence in Sidgwickian terminology. 16

17 Further, Pigou comments on justice or rights-based argument. Taking up the phrase of McTaggart (a notable Hegelian-cum-British Idealist), Pigou questions if everyone has claims to possess the good which cannot be altogether cancelled by any amount of good possessed by other people (McTaggart 1906, 17). His remark on this right-oriented proposition is as follows: But this, if it be true which I do not myself believe, and which, given that states of consciousness are the only things ultimately good or bad, cannot as a matter of fact be the case merely asserts that fairness as between people is one element among good things (Pigou 1908a, 92. italics mine). If things like fairness should be included in the intrinsic good, theoretically we must abandon methodological individualism in the theory of value, which leads to a deviation from the utilitarian principle (Uchii 1988, 164-5). Anyway, Pigou himself does not take such a line. As for rights, for instance, property rights, he thinks them to be a mere product of legislation or social expediency which should be ultimately regulated by social welfare (Pigou 1947, 4-5; 43). ⅢThe relation between good and right In the previous section, we examined (1) Pigou s theory clarifying what is the good thing in itself. Now, let us inquire into (2): his theory that clarifies what conduct is right or how we ought to behave. The point here is to ascertain the logical relation between good and right in his notion, and it is an easy task. The morally right is justified by his theory of good as discussed in II. As the basic matter stated previously, Pigou has distinguished between intrinsic good and instrumental good (Pigou 1907b, 99). And, since the former is limited to the category of states of consciousness, it inevitably follows that other things, including morally right actions, can be good if and only if they are ultimately useful to intrinsic good. Our ultimate goal, of course, is to promote the greatest possible amount of goodness in itself (Pigou 1907d, 344) 28. Whatever form of ethical theory we may hold, we are all agreed that it is reasonable, in choosing a course of action, to take some account of its 28 It continues: Since, however, everything that there is produces effects, we cannot always work with success towards this end by manufacturing things that are good in that way. For these things may produce other things bad in themselves, and the badness of the effects may outweigh the goodness of the original cause (Pigou 1907d, 344). Similarly, In human affairs it is rarely possible to demonstrate absolutely even though our criteria of good be agreed that one course of action is better than another. The data are always imperfect (Pigou 1937, 137). He calls this practical difficulty. I am not certain whether a viewpoint like that was common at that time; still it was a main issue in the 5 th chapter of Moore s Principia. 17