chapter. But we must notice at once that this was the way in which the Austrians rediscovered marginal productivity. Theirs was, however, a marginal productivity with a difference. In order to clarify this point, let us recall the usual distinction between marginal physical productivity and marginal value productivity. Marginal physical productivity of a ‘factor’ is the increment of product that results from an infinitesimal increment in the quantity of that factor. Marginal value productivity of a ‘factor’ to a firm 15 is this physical increment multiplied by the corresponding increment in this firm’s total revenue or gross receipts. Both these concepts do enter into the Austrian theory. But they do not enter on its ground floor, and they had been developed independently of it. 16 Fundamentally, the Austrian marginal productivity was indeed a value productivity but one that did not presuppose the price of the product: it was not physical marginal productivity multiplied by any price but physical marginal productivity multiplied by some consumers’ marginal utility. It was on this basis that they worked out their theory that was at the same time a theory of production and of distribution: the tools of their barter theory, forged beforehand, then came in to implement it and to show how this works out in a private-property economy. Now, this conception of marginal value or utility productivity makes obvious common sense only in the case of a Crusoe economy. Crusoe may indeed be reasonably supposed to value his various scarce means of production according to the satisfactions he knows to be, on the margin, dependent upon their possession. To use Wieser’s term, he may indeed be supposed to impute these satisfactions to those means (his own ability to work being just one of them) and so, for his own practical purposes, to perform a subconscious process of imputation. But if we are to maintain that a similar process of imputation constitutes the innermost meaning of the mechanisms of ‘acquisitive society’ (Tawney’s term), it is necessary to construe this imputation as vicariously performed by firms that do not psychologically experience those consumers’ satisfactions and in any case want to maximize pecuniary gains instead. To prove this is the real problem. So far as it can be proved at all, it can be proved only by showing that the barter or price mechanisms of free markets will operate in such a way as to insure the results that would follow if fac- 15 Both physical and value productivity are in the first instance defined with refer ence to individual firms. But the Austrians, especially Wieser, as well as J.B.Clark, being interested in the social process as a whole, attempted to proceed directly to social productivities, social values, social marginal utilities. This created another class of difficulties that are typical of an early stage of analysis, difficulties that Marshall and Wicksell were much better equipped to avoid. Not much came from the discussion on Social Value, however, and we shall not go into it beyond noting that this desire to reason in social terms accounts for the curiously semi-socialist constructions of Wieser and Clark in which society itself plays the role of directing agent. 16 This is why in this case we find a different set of forerunners: while in matters of marginal utility we have Dupuit, Gossen, and so on, we have here mainly Longfield and Thünen. This is also the reason why Menger, if indeed he was familiar with Thünen’s work, failed to discover in it anything bearing directly upon his own. History of economic analysis 882 tors were first evaluated as Crusoe evaluates them and if these utility values were then turned into exchange values or prices in the same manner in which the utility values of consumers’ goods are turned into exchange values in a simple consumers’ goods market. 17 Even to posit this problem, which is obviously neither trivial nor without interest, would have spelled considerable achievement. But Menger and Wieser, barring technical defects, went almost the whole way toward actually solving it, and thereby also solved the fundamental problems of allocation of resources (production) and of the pricing of these resources (distribution). However, the construction involved in applying the method of imputation was not only far removed from any actual mental processes that may be credited to any deciding agents—this does not greatly matter considering the ‘as if’ that enters this as it does so many other scientific constructions—but also unnecessary. In order to determine the prices of factors and their distributive shares we do not need to know their utility values first. All we need to know is consumers’ tastes, the technological conditions of production, and the initial distribution of ownership of ‘factors’; then the principle of maximum net revenue, implying a principle of minimum cost, will do the rest. But the Austrians insisted on asserting their fundamental idea at every step and, in order to accomplish this, thought it necessary to divide up between the factors the use value of the product just as the receipts from the sale of the product are actually divided up between factors, the idea being that the former process (a methodological fiction) would yield the explanation of the latter process (a reality). Therefore, their problem of Imputation (Zurechnung) took the following form: to find the utility functions of producers’ goods from given utility functions of consumers’ goods. It was greatly complicated for them by their technical deficiencies and gave rise to a considerable literature—positive and critical—which, starting from the original contributions of Menger, Wieser, and Böhm- Bawerk, explored various blind alleys and produced more heat than light. It is, however, not necessary for us to go into this. 18 But whatever we 17 Some critics bluntly asserted that the Austrian theory of value, being a theory of the evaluation of consumers’ goods, was inapplicable to the behavior of people who produce not for the satisfaction of their own wants but for markets. This, of course, indicates complete inability to understand Menger’s reasoning. Others found circular reasoning in the proposition above because producers who produce for the market value their products for the sake of the money they hope to get, which presupposes the idea of expected prices, the very thing that is to be explained. The error involved in this argument should be clear. I use this opportunity to mention at once the two other vicious circles that have been charged to the Austrian theorists (often by members of their own group), though both will have to be mentioned again. The second is: we know only from a man’s behavior (actual choices) which one of several quantities of goods he prefers; it is therefore circular to explain his choice by his preference. The third is: we value money for the sake of the goods that it will buy; hence, in the case of a purchase by means of money it is circular to hold that price is determined by the utility values of the two things to be exchanged. 18 Professor Stigler is emphatic in expressing the same opinion. He gives, however, a small sample of this literature (op. cit. p. 5n.). His judgment to the effect that some of General economics 883 may think of the technical merits of the theory of imputation, it expresses a profound truth which the simple statement that production and distribution are matters of evaluating productive services does not convey in itself. And it yields a satisfactory theory of costs. Discussion of the marginal utilities of means of production in the spirit of the theory of imputation easily leads to the recognition of the relevance to these marginal utilities of the elements of complementarity and substitutability 19 of factors and of their alternative uses. By this route the Austrian arrived at what has been called the alternative-use or opportunity theory of cost 20 —the philosophy of the cost phenomenon that may be expressed by the adage: What a thing really costs us is the sacrifice of the utility of those other things which we could have had from the resources that went into the one we did produce. Sporadically, this theory of cost had turned up in the past, especially in J.S.Mill’s Principles, but only to explain special cases which failed to fit into the older schemata. As a general theory and as an explanation of the fundamental social meaning of cost— both in capitalist and in socialist society—it was new. There should be no doubt also that it makes a much better theory of distribution. But I want to advert particularly to the fact that it emphasizes a phenomenon that got all but lost in Marshall’s analysis. Consider the allocation of a requisite of production, say, labor, of a given type in a two-commodity economy. As we go on allocating more and more labor to the production of commodity A and less and less of it to the production of commodity B, marginal utilities of the commodity A will fall and marginal utilities of commodity B will rise. We may express this by saying that the utility returns from the A-production are decreasing and that costs in the A-production are increasing: we have derived, from the marginal-utility principle, a new ‘law of decreasing returns,’ which is independent of any physical law of decrease, and which will assert itself even in face of a physical law of increasing returns. the monographs on the subject have been grossly incompetent is, I am afraid, only too justified. 19 The idea of substitutability was, of course, familiar to Thünen. But Menger was the first to formulate it explicitly: ‘it is…certain that not only can fixed quantities of goods of higher order be combined in production in the way in which we observe this in chemical products…[But] general experience teaches us that a given quantity of any good of lower order can result from very different combinations of goods of higher order.’ (Grundsätze, p. 139 [trans. by J.A.S.]). To say the least, this foreshadowed the ‘law of variable proportions’ and even the concept of ‘equal product curves.’ Moreover this formulation is superior to Marshall’s later ‘principle of substitution.’ 20 The latter—very felicitous—term is due to D.I.Green, ‘Pain Cost and Opportunity Cost,’ Quarterly Journal of Economics, January 1894, and has gained wide currency in the United States owing to the vigorous sponsorship of Professor Knight. The most exhaustive treatment of this whole set of problems is to be found in H.J.Davenport’s Value and Distribution (1908), who preferred the equivalent term Displacement Cost. History of economic analysis 884 (c) Interdependence and Equilibrium. If we behold the bustling crowds that work and trade in order to make a living, we shall indeed have little difficulty in linking up their behavior with appetites for gain and appetites for goods. But it will be by no means obvious to us that the process that generates real income can be satisfactorily described, so far as its formal logic is concerned, by any simple principle or that there is any inherent logic to it at all. The history of analytic effort in this field is the history of a growing awareness, partial at first, ever more general later on, of the presence of a logically coherent economic process, an awareness that first attained conscious formulation in the works of such men as Cantillon, Quesnay, A.Smith, Say, and Ricardo. But it was only in the period under discussion that the conception of an economic cosmos that consists of a system of interdependent quantities was fully worked out with all its problems, if not quite satisfactorily solved, at least clearly arrayed and with the idea of a general equilibrium between these quantities clearly established in the center of pure theory. This was the achievement of Walras. So soon as we realize that it is the general- equilibrium system which is the really important thing, we discover that, in itself, the principle of marginal utility is not so important after all as Jevons, the Austrians, and Walras himself believed. But analysis of Walras’ schema at the same time discloses the fact that marginal utility was the ladder by which Walras climbed to the level of his general-equilibrium system. If the marginal utility principle ceased to be all-important after this level had been reached, it was nevertheless all-important heuristically. This observation sheds new light on the achievement of Jevons and the Austrians. They, too, found the ladder. Defective technique only prevented them from climbing to the top of it. But they did climb as high as their technique permitted. In other words: we must see in the Jevons-Menger utility theory an embryonic theory of general equilibrium 21 or, at all events, a particular form of the unifying principle that is at the bottom of any general- equilibrium system. Though they did not make it fully articulate, mainly because they did not understand the meaning of a set of simultaneous equations, and though they saw in marginal utility the essence of their innovation instead of seeing in it a heuristically useful methodological device, they are nonetheless, just like Walras, among the founding fathers of modern theory. This also holds for J.B.Clark. Later critics were so delighted with their own technical improvements and so anxious to renounce communion with Jevons and the Austrians that they entirely failed to perceive this. In what sense was a revolution effected? And did this revolution produce a new theory of the economic process? The answer to the first question will depend upon what we mean by that much misused word. If we mean change that is both thoroughgoing and discontinuous, then the claim of these pioneers of modern theory—the claim to having revolutionized the ‘pure’ part of economics—should be admitted. For though J.S.Mill’s shaky structure invited reconstruction on the lines actually adopted by Jevons, Menger, and Walras, and though Marshall may be said 21 This aspect is particularly in evidence in Wieser’s Natural Value. General economics 885 to have done much the same thing by reform rather than by revolution, the controversies of that age testify strongly to the break that had occurred. We are apt to smile at Gossen’s boast of having accomplished a Copernican feat. But this boast was less unreasonable than it may seem at first sight. The replacement of the geocentric by the heliocentric system and the replacement of the ‘classic’ by the marginal utility system were performances of the same kind: they were both essentially simplifying and unifying reconstructions. The comparison strikes us as ridiculous only because of the different intellectual standings of astronomy and economics. Similarly, we smile when we learn that the Negro soldier-statesman Toussaint L’Ouverture (1743–1803) described himself as the Buonaparte of Santo Domingo. But this is because France’s importance in the world is so much greater than is Santo Domingo’s rather than because there is any obviously ludicrous disproportion between the two men, each taken with reference to his environment. 22 This bears upon the habit, which has developed especially in the United States, of describing the ‘marginalist’ theory as neo-classic. Considering how much of the old framework and the old attitudes was taken over by the ‘marginalists,’ we might feel inclined to approve of it. Marshall’s efforts to preserve continuity—and still more a semblance of it—lent additional support to the implied, and somewhat derogatory, appraisal of the revolutionaries’ performance. But so far as pure theory is concerned, there is no more sense in calling the Jevons-Menger-Walras theory neo-classic than there would be in calling the Einstein theory neo-Newtonian: as we know already the term Eclectics, as applied to Marshall and his followers, is still more misleading. For this, however, he had only himself to blame. The second question, as it stands, must of course be answered in the negative. No theory in the sense of pure theory can ever be a theory in the sense of complete analysis of the phenomena to which it refers. Factual assumptions are as important as is the analytic apparatus that distills results from them. 23 22 I believe, in fact, that it is impossible, after analysis of the two careers which display certain striking similarities from their splendid beginnings to their melancholy ends, to say with confidence that less personal energy and genius went into Toussaint L’Ouverture’s achievements than into Napoleon’s. But economics is to astronomy, as Santo Domingo is to France. 23 The marginal utility theorists were (like most theorists to this day) only imperfectly aware of the formal character of their analysis. Preoccupied as they were with what they believed to be psychic facts, they thought that they were teaching much more about economic reality than was actually the case. I take this opportunity to comment once more on a meaningless controversy that nevertheless seemed important to many an able man, viz. the controversy about subjectivism and objectivism in pure theory in general and in the theory of price in particular. Actually, the ‘subjective’ theory must always appeal to ‘objective’ facts (data) if it is to produce concrete results; and any ‘objective’ theory must always state or imply postulates or propositions about ‘subjective’ factors of behavior. In other words, any complete subjective theory must be also objective and vice versa, and differences on this score can only be due to differences of emphasis on different parts of the analyst’s task. Yet the ‘issue’ was accepted as real and gravely discussed by all scientific parties alike. History of economic analysis 886 Moreover, economic life is a unique historical process and our authors had no explanatory schema of economic change other than the one they had inherited from A.Smith; even if they had had one of their own, their marginal utility theory would have been completely neutral to it. Finally, they had no more explicit dynamic schemata than had their ‘classic’ predecessors and they struggled through the inadequacies thereby created very much in the same manner. It might seem that, if our question were reformulated in the light of these three qualifications, the answer should be Yes: the marginal utility theorists certainly seem to have succeeded in creating a use-value schema of economic statics that was complete in itself. Unfortunately, we must qualify still further. Not all of the problems of pure theory were capable of unique solution within the marginal utility theory. We have had examples already, namely the theories of enterprise and of capital: as regards these, marginal utility theory completely failed—and very naturally—to restrict the range of possible difference of opinion. Another case of failure of the unifying power of the marginal utility principle is the theory of interest. This is the main reason why we shall have to enter into a separate discussion of the distributive shares after all, although the occasion will be used to touch also upon a few other matters (below sec. 5). Before we do this it will be useful to discuss Marshall’s attitude to the Jevons-Menger analysis. 4. MARSHALL’S ATTITUDE AND REAL COST The reader is requested to refresh his memory of what has been said in the preceding chapter on the subject of Marshall’s attitude to contemporaneous and earlier work closely similar, in fundamentals, to his own. In view of the thoroughly un-Marshallian interpretation of the work of the Austrians and, to a lesser extent, of Jevons, which has been presented in the preceding section, 1 it is necessary to consider Marshall’s different appraisal of this work and some of the arguments by which he supported it. 2 I submit that the impression the reader of Marshall’s Principles is bound to receive, in spite of qualifications that Marshall inserted occasionally, is this. Marshall maintained (1) that English ‘classic’ analysis stood in need of corrective reinterpretation here and there but that there was nothing fundamen-tally wrong with it; (2) that Jevonian and Austrian criticisms had been largely due to failure to understand and to interpret it properly; 3 (3) that the posi- 1 A point of particular importance should be emphasized once more. We have seen in ch. 5, sec. 2, that the contents of ‘note XXI’ in the Appendix to the Principles constitutes the core of Marshall’s theoretical analysis. This note blocks out a system of general equilibrium. In the preceding section it has been maintained that the theoretical analysis of the Austrians also amounts to an embryonic—and highly defective—equilibrium system. 2 For the sake of brevity, Marshallian criticisms of Jevons will be treated as if they also had been directed against the Austrians and vice versa. The reader will have no difficulty in satisfying himself that this is permissible in the instances discussed. 3 Nothing illustrates this better than does Edgeworth’s famous phrase, the echo of General economics 887 tive contribution of Jevons and the Austrians consisted in elucidating the demand side of market phenomena, though Ricardo had, of course, not been ignorant of these rather obvious things; and (4) that in overstressing demand aspects Jevons and the Austrians had erred at least as much as Ricardo and Mill had erred in the opposite direction. These positions must be considered from three different angles. First, as regards Jevonian and Austrian criticisms of Ricardo, Marshall’s irritation, though unjustified in part, 4 was not unnatural. We may readily grant that Nicolaus Cusanus and Copernicus did not prove the geocentric theory to be ‘wrong’ but that they simply applied some corrections to it. Second, as regards the performance of Jevons and the Austrians, Marshall, by nature and acquirement, could not help being severe toward their deplorable technique. We are no longer concerned with this. Third, however, as regards the fundamental meaning of that performance, Marshall’s interpretation amounts to serious though of course unintentional misstatement. It is on this point that our argument in the preceding sections needs countercritical supplement. Marshall illustrated his criticism of what he conceived to be overemphasis on demand aspects by a famous simile. ‘The “cost of production principle” and the “final utility” principle are undoubtedly component parts of the one all-ruling law of supply and demand; 5 each may be compared to one blade of a pair of scissors. When one blade is held still, and the cutting is effected by moving the other, we may say with careless brevity that the cutting is done by the second; but the statement is not to be made formally and defended deliberately.’ 6 Even when we narrow down the Austrian analysis to this particular Marshall’s voice, viz. that Marshall had dispersed the mists of ‘ephemeral criticism’ that for a time had clouded the ‘eternal heights.’ 4 This should be evident from our discussions in Part III. But we also know that Marshall’s irritation was less unjustified with respect to criticisms of Mill than it was with respect to criticisms of Ricardo. And though Marshall did not admit this in so many words, he may be called as a witness for our view: he never espoused the specifically Ricardian elements in the ‘classic’ structure, such as, e.g., the labor-quantity theory of value, which he quietly modified so that it is no longer what Ricardo had intended it to be. 5 The reader should observe that this statement, though it may be held to conform to Mill’s teaching, is thoroughly un-Ricardian. It is Malthusian. 6 Principles, p. 569. The passage is an almost verbatim repetition of another passage on p. 428 that leads up to the statement ‘that, as a general rule, the shorter the period which we are considering, the greater must be the share of our attention which is given to the influence of demand [utility] on value; and the longer the period, the more important will be the influence of cost of production on value’ (p. 429). Strictly interpreted, this statement is of course as true as it is trite. In its general import, however, it gives a wrong lead. This can best be explained by an analogy: it is wrong to say that foreign exchanges are determined by supply and demand in the case of paper currencies, and by the gold mechanism in the case of gold currencies; what should be said is that the factors behind supply and demand determine foreign-exchange rates in any case, but that in the case of gold currencies the gold mechanism will in general History of economic analysis 888 point—doing which implies neglect of all its wider aspects—we have still to recognize that its essential achievement was precisely the new theory of supply and cost that it yielded. It is in this sense only that Jevons’ saying should be understood: ‘Value depends entirely upon utility’ (Theory, p. 1). Hence it is meaningless to accuse either Jevons or the Austrians of wishing to minimize the importance of the very theorem which they were the first to deduce rationally and which Wieser called the ‘law of cost.’ They stood in no need of being told about the two blades of Marshall’s pair of scissors. What they aimed at showing was that both blades consist of the same material—that both demand and supply (no matter whether the case is one of exchanging existing commodities or one of producing them) may be explained in terms of ‘utility.’ In appearance at least, there is more to another form that Marshall gave to what was substantially the same indictment. Both Jevons and the Austrians were in the habit of expressing themselves in terms of causal chains, which ran from the value of consumable goods to the value of resources as if the utility of a quantity of a consumers’ good were first determined independently and then, in turn, determined causally the value of the producers’ goods that went into its production. It was child’s play for a superior technician to point out that this was inadmissible since the utility of the consumers’ good depends upon its quantity and the latter upon its cost. Jevons and the Austrians were held up to ridicule as people who, like school children, had to be taught that ‘when three balls, A, B, and C, rest against one another in a bowl …the position of the three mutually determines one another under the action of gravity,’ because they were contending instead ‘that A determines B, and B determines C’ (Principles, p. 567). But Marshall of all men should have realized that this criticism takes advantage of deficiencies in technique, in particular of a glaring inability to understand the logic of interdependence, and entirely fails to do justice to the substance of the position criticized. To keep again to Marshall’s simile, what Jevons and the Austrians really did was not the nonsense imputed to them in that passage but something very different; they discovered precisely that the position of the balls is to be accounted for by a single principle, gravitation in the case of mechanics, utility in the case of economics. Half the generosity lavished upon Ricardo might have revealed the great achievement behind the poor technique and reduced criticism to the one point that could have been justifiably made though Mar- prevent departure from gold parities beyond the gold points. Similarly the marginal utility principle applies to the demand and the supply sides of the value problem in any case, both in the long and in the short run. Cost of production is not an independent principle taking charge in the long run. But the marginal utility principle, acting upon the data of the situation, will in the long run (granting a number of assumptions) so operate as to equate exchange value to costs. Accomplished Marshallians will think this note superfluous. But I have heard the misinterpretation in question so often—and so often with reference to Marshall, even from competent economists such as Bortkiewicz—that I cannot myself think it so. General economics 889 shall never made it: that Jevons knew not enough mathematics and the Austrians none at all. 7 The only other point that must be attended to in this connection is Marshall’s Real Cost. If the Austrians had used this term, they would have meant by it the consumers’ goods (as distinguished from the satisfactions afforded by their consumption) that we ‘sacrifice’ when we decide to produce others. Marshall meant ‘The exertions of all the different kinds of labour that are directly or indirectly involved in making it [a commodity]; together with the abstinences or rather the waitings required for saving the capital used in making it’ (Principles, p. 418). This point turns up here because, at the time, it was discussed within the general controversy on the nature, cause, or ‘ultimate standard’ of value or cost. The quarrel was exclusively a Mengerian one. For most of the other sponsors of the ‘new’ theory of value, such as Gossen, Jevons, Auspitz, Lieben, and Clark—not Walras though—experienced no qualms about admitting both disutility of labor (Jevons’ term) and abstinence into their analytic structures. Since all these authors were in no mind to rehabilitate the ‘classics,’ by pegging the independent role of costs or in any other way, this should suffice to show that recognition of disutility and abstinence does not impair the marginal utility position or constitute the adoption of another one. But the Austrians were of a different opinion. Böhm-Bawerk fought hard indeed in order to minimize the importance of both, evidently believing that allegiance to marginal utility theory compelled him to do so. Let us take the measure of the problems involved. Abstinence is, of course, very important for every economist who holds an abstinence theory of interest. But, though holding an abstinence theory of interest of course involves introducing abstinence into the general theory of value, the question has always been dealt with primarily in connection with the theory of interest and we shall do the same (sec. 5). As regards disutility of labor, we have our choice: either we can take (with a given population) the quantity of labor hours available as a datum, for example, as institutionally fixed; or we can make it a variable to be determined, in which case our system contains one more ‘unknown’ and one more independent equation (which says that, for every workman, the marginal disutility of labor must in equilibrium equal the marginal utility of his wage income). Which we choose will depend on considerations of realism and considerations of analytic convenience. 8 But the point is that our choice does not make any great difference to our theoretical pattern. For the element of disutility acts upon the value of 7 Marshall’s severity was the more uncalled for because he occasionally fell into the same kind of error himself. On p. 440 of the Principles we read: ‘In a stationary state then the plain rule would be that cost of production governs value.’ 8 As the reader presumably knows, the disutility equation was thrown out by Lord Keynes on grounds of realism. Böhm-Bawerk also threw it out on a ground of realism though a different one, viz. that the individual worker must accept the regulation working day and cannot vary the quantity of labor he is willing to offer. But whereas Böhm-Bawerk could have inserted the disutility equation if he had wanted to do so, Lord Keynes’s analytic set-up made it imperative to throw it out. History of economic analysis 890 products only through its (possible) influence upon the quantity of labor offered, and leaves the principle of opportunity costs untouched for purposes of allocation of the quantity that is being offered. It is always the latter that matters in the first place, whereas disutility matters, if at all, at one remove. Moreover, if we attach enough importance to having our value theory based upon utility only, all we have to do is to replace disutility of labor by the utility of leisure. 9 Hence, Böhm-Bawerk did not gain much by the limited success of his attempt to minimize the importance of disutility. But neither did Marshall gain much by introducing Real Cost: always excepting the special service that abstinence rendered in his theory of interest, it can be left out without being missed—not to speak of the difficulties inherent in the concept of a sum of all disutilities and abstinences ‘directly or indirectly involved’ in producing a commodity. Thus, we return from this excursion with the same result that we always get when inquiring into the nature and importance of Marshall’s deviations, in what purport to be fundamentals, from the Jevons-Menger- Walras analysis: they are negligible. 10 5. INTEREST, RENT, WAGES Any pure theory needs facts in order to produce concrete results. This platitude must be repeated because economists are in the habit of including certain particular facts in what nevertheless they call a pure theory. Thus they speak of a minimum-of-existence ‘theory’ of wages although the minimum-of-existence theorem may be deduced from any general theory of wages, provided we introduce the appropriate factual assumption about the behavior of workmen. But the marginal utility theory not only needs to be complemented by a supply of particular facts if it is to apply to concrete cases; it also needs to be complemented by additional material in order to produce general theoretical propositions. As has been stated at the end of Section 3, it does not, of itself, yield any general theory of interest though it does supply, of itself, adequate explanations of rents and wages. Since it is interest that causes the trouble, we shall begin by discussion of the interest theories of the period under survey. (a) Interest. We know already that the economists of that period sharpened the distinction between entrepreneurial gains and interest. But most of them still took the view that we have traced to Nicholas Barbon (Part II, ch. 6, sec. 7b), namely, that interest constitutes the bulk of business gains—the part of business gains that results from the application of physical capital and is a return to physical capital in the same sense in which rent is a return to land and wages are a return to labor. In this respect, it is highly significant that 9 This is to be recommended in any case. The particular problem in which disutility plays a role, e.g. the problem of explaining why increase in wage rates sometimes results in a decrease in the amount of work done, becomes even easier to solve when we use the concept of leisure (and put leisure on one of the axes of an indifference diagram, the other axis representing wages, monetary or real). 10 In addition, it is difficult to understand how Marshall can have believed (if he did) that introducing real cost of this type would help Ricardo’s position in any way. General economics 891 . at all. The history of analytic effort in this field is the history of a growing awareness, partial at first, ever more general later on, of the presence of a logically coherent economic process,. History of economic analysis 890 products only through its (possible) influence upon the quantity of labor offered, and leaves the principle of opportunity costs untouched for purposes of. different parts of the analyst’s task. Yet the ‘issue’ was accepted as real and gravely discussed by all scientific parties alike. History of economic analysis 886 Moreover, economic life