But in most cases it is possible to put matters right by means of small corrections. Thus, most economists, J.B.Say and J.S.Mill in particular, thought altogether too much of the analogy with the physical sciences, which the latter declared to be the ‘proper models’ for economic theory (Autobiography, p. 165)—a point for critics to fasten on but actually irrelevant, since no practical use was made of it. 7 J.B.Say, while correctly emphasizing that economics is an observational science, nevertheless called it ‘experimental.’ But this can be easily corrected into ‘empirical.’ Furthermore, practically all economists used the term Law or even Natural Law, the avoidance of which would have saved them much obloquy from philosophy-minded critics. But this habit was quite harmless, since what they really meant was nothing but Montesquieu’s ‘necessary relations’ between economic phenomena or Marshall’s ‘statements of tendencies.’ In view of J.S.Mill’s insistence on ‘the very limited and temporary value of the old political economy,’ there is no excuse for later critics who harped on those words. In fact, all the really valid points in the latter’s methodological credo could have been copied out from Mill. Again, Mill used the term a priori in a misleading sense 8 and also placed unnecessary emphasis upon ‘deduction.’ This was perhaps responsible for the absurd argument of later times about ‘induction vs. deduction,’ but, always remembering that he thought of the theoretical apparatus of economics when speaking of methods of political economy, we readily see that it never caused any errors in practice. 9 Finally, as regards the method of ‘isolating’ economic phenomena or motives, or of abstracting from noneconomic ones, not only the practice of the ‘classics’ but even their methodological rationalization of it was free from serious error. It is difficult to believe that any critic who raises objections on this score can economy’ in the sense of economic theory. The passages in the Logic deal with the methodology of a much wider sector of the social sciences, mainly with what is called economic sociology in this book. The epistemological situations of the two fields differ substantially, and there is no contradiction between prescribing ‘deductive’ methods for the one and ‘inductive’ (or ‘inverse deductive’) methods for the other. The chief reason for this is that economic theory, owing to its quantitative character, admits of systematic elaboration to a much greater extent than does the analytic apparatus of any other social science. 7 We have already observed that the introduction of the terms Statics and Dynamics does not involve any such use, i.e. any borrowing of a method from any physical science. Nor do economists borrow from mechanics when they employ the term equilibrium any more than does a bookkeeper who ‘balances’ an account. 8 This is all the more surprising, because in his Logic he even made geometry an empirical science and because, though not an empiricist to 100 per cent, he certainly was more of an empiricist than was Kant. 9 This, to repeat, is the real test. The literal meaning of a methodological profession of faith is of little interest except for the philosopher. Alternatively, our test may be formulated by saying that any objectionable piece of methodology is immaterial whenever it can be dropped without forcing us to drop any result of the analysis that is associated with it. History of economic analysis 512 have studied J.S.Mill. 10 Of course, this statement must be understood to refer to the principles of isolation and abstraction per se as the ‘classics’ applied them for the purpose of carving out the domain of purely economic research. Only, so far as this goes, I maintain that, in principle as well as in practice, their procedure did not differ either from that of A.Smith 11 or from that of later economists as formulated by the later methodologists, Carl Menger and John Neville Keynes (see below, Part IV, ch. 4), and as accepted, around 1900, by a large majority of non-German economists. But I do not maintain that individual ‘classic’ writers, when reasoning within that domain, always ‘isolated’ relevant factors and ‘abstracted’ from others faultlessly. This would be absurd on the face of it, because it would imply that practically all their propo sitions were faultless: any criticism that does not charge either logical error or misstatement of facts can be formulated as an objection to the manner in which the criticized author ‘isolates’ or ‘abstracts.’ The distinction that I have just tried to convey will help us greatly in understanding the methodological situation of the period, that is to say, the nature and extent of differences about ‘method’ that then existed among economists. At first sight, we have the impression that the scientific controversies of the times turned largely upon method. Thus, the two most famous and most prolonged controversies, the one on value and the one on general gluts, speedily led to the familiar situation in which, no progress being made with more concrete arguments, the parties fall back upon objections to one another’s methods. In itself, this means little beyond admission of inability to convince 10 As a matter of fact, men like Roscher, who had mastered J.S.Mill, did not raise any such objection. But then, as will be shown presently, he was really no critic of the ‘classics’ in the sense in which this implies a methodological creed, incompatible with theirs. The majority of later German critics, however—those of the genuine historical school—cannot have had much first-hand knowledge of Mill or of the ‘classics’ in general, for it is hard to believe that if they had they could have misunderstood Millian methodology as completely as they did: they argued rather against a wrong picture of it that had become fixed by the time they wrote. And, considering their absorption in a research program of their own, this is after all not so difficult to understand. But what about Ingram (see below, Part IV, ch. 4)? How was it possible for him to preach the gospel of a ‘new economics’ that took its methodology from Comte? The only answer I am able to offer—and which must suggest itself to any professional who makes a study of Ingram’s History of Political Economy—is that both his knowledge of economics and his interest in it did not go beyond general ‘philosophies’ that were inspired by generous enthusiasm for the great slogans of his day but never came to grips with real problems. His other objections to ‘classic’ economics point to this conclusion still more clearly than does his Comtist objection to an autonomous economics. 11 This is so important as to justify repetition: A.Smith’s work looks less ‘abstract’ because it includes so much factual information that the specialized later works on economic theory did not include—but left for other specialized works to provide. But where he does move within the orbit of economic theory, his reasoning is not less abstract than is, say, Ricardo’s. With the latter, ‘abstractness’ shows more because he confines himself to topics of an ‘abstract’ nature, and does not provide illustrative foliage, but this is all. General economics 513 the other man, coupled with a declaration of being unconvinced by him—or, in a word, deadlock. For instance, when in the controversy on gluts, Malthus (and Sismondi) objected to Ricardo’s procedure as too abstract and Ricardo himself emphasized the abstract nature of the argument, 12 they simply verbalized a sigh of despair. It would be quite wrong to infer that Malthus and Sismondi really objected to Ricardo’s ‘method’ in the sense in which this word came to be used in the later Battle of Methods (see Part IV, ch. 4). That this was not the case can be established by analysis of their own modes of reasoning: these were ‘theoretical’ in the same sense as was Ricardo’s—just as Lord Keynes’s theory is theory in the same (logical) sense as is Marshall’s. In other words, Malthus and Sismondi theorized in a different way and partly with an eye to different sets of facts, but their practice proves that they had no objection to theorizing per se, such as had, for a time, the later Schmoller school or the American institutionalists. But was not this objection (against theory per se) raised by other people? It was, but only in isolated instances that had no significant influence upon the work of a large majority of economists. 13 One such root-and-branch objector was Comte. But as we have seen, he exerted no perceptible influence within the period and on economists: J.S.Mill, so far as technical economics is concerned, yielded not an inch. Another was Le Play. He initiated an important program of research but, for the rest, he was hardly known among the economists of the period. I doubt whether R.Jones and B.Hildebrand can be called root-and-branch objectors. But even if they can, they were no more than forerunners. Cliffe Leslie did not declare for a distinctive historical method in economics until 1876. Ingram did not raise the flag of his New Economics until 1878. Knies, 14 it cannot be too often repeated, was primarily an economic theorist and, for the rest, an able general economist without any 12 For, though ‘abstractness’ is usually urged against an argument, it may also serve in its defense. Marxists, in particular, frequently—and in some cases with justice—save a perilous situation by yielding the doctrine (in controversy) on one level of abstraction with the reservation that the same doctrine is quite right on a higher level of abstraction. 13 I am not speaking now of extra-scientific objectors such as Carlyle. Nor am I speaking of those objections that arise from the general public’s dislike of anything that looks like a complicated argument, nor, finally, of those objections that simply express everyone’s natural preference for his own chosen type of work. 14 As already stated, Knies’s chief work and also the better part of his activity as a teacher belong to the next period. But the only one of his writings that connects him with the historical school appeared during the period under survey: Die politische Ökonomie vom Standpunkte der geschichtlichen Methode (1853; greatly enlarged, 1883). This book presents an interesting problem. It expresses not only a sense of the general flux of social institutions and of the impossibility of framing universally valid ‘policies’—and other things that the author could have just as well taken from J.S.Mill—but actually outlines essential parts of the Schmoller program. But the modes of procedure actually employed in all his subsequent publications are so many disavowals of these principles. As a teacher, too, Knies was very far from inculcating economic historicism. History of economic analysis 514 distinctive bent as to method. As for Roscher, who described Ricardo and Malthus as ‘political economists and discoverers of the first rank’ and who went out of his way to express agreement with J.S.Mill’s methodology, 15 a comparison of his Grundlagen with J.S.Mill’s Principles fails to reveal any fundamental difference in procedure—he even speaks of natural laws. It is true, nevertheless, that he laid claim to having used a historical or ‘physiological’ method. But, as is evident from Chapter 3 of the Introduction to that work, all he meant by this was to dissociate himself from what he called the ‘idealistic’ method that prescribes norms for an ideal state of society, whereas he wished to describe things as they are ‘after the manner of the investigator of nature’ 16 (op. cit. vol. I, p. 111). We therefore emerge with the result that, barring isolated rumbles, the Battle of Methods had not been engaged as yet and that—taking the word method in the sense that is relevant here—methodological peace substantially prevailed. This was also Cairnes’s opinion. (c) The Science and the Art. Most of the writers of standing who paid serious attention to the fundamental questions of methodology clearly saw, and strongly emphasized, the distinction between arguments about what is and arguments about what ought to be: the distinction between the ‘science’ of economics and the ‘art’ of policy. 17 But it would be a great mistake to read into their statements the meaning which this distinction acquired later, when the question of ‘value judgments’ was raised. Senior, who was more explicit on the point than was anyone else, said indeed that the economist’s conclusions ‘do not authorize him in adding a single syllable of advice.’ But by this he did not mean that the economist as a scientific worker is debarred from presenting practical advice because such advice presupposes ultimate valuations that are extra-scientific by nature—preferences that are beyond the range of scientific proof. This is the point of view that was taken by Cairnes (who did not, however, adhere to it in practice) and later on, more explicitly, by Sidgwick and by M.Weber. Senior and Mill and their contemporaries did not mean this at all. They merely meant that questions of economic policy always involve so many noneconomic elements that they should not be dealt with on the basis of purely economic considerations— which, by the way, in itself suffices to show how little there is in the common indictment that the 15 W.Roscher, Grundlagen (English trans., Principles of Political Economy, 1878, vol. I, p. 106n.). 16 This does not apply, without qualification, to L.Wolowski’s ‘Preliminary Essay’ that Roscher’s translator, J.J.Lalor, prefixed to the English translation (1878). In 1878 things looked different and, by a not unnatural confusion, Lalor spoke of Roscher as the ‘founder’ of what was then the Historical School. This gives a completely wrong picture. 17 Confusions between the two and even denials of the validity of the distinction were of course frequent; but not more so than they have been at any later time, including the present. In the large majority of cases the distinction was firmly kept in view. For France, Charles Coquelin’s article on political economy in the Dictionnaire de l’économie politique is typical in this as it is in other respects. Roscher’s emphasis upon the distinction between what is and what ought to be deserves particular notice. General economics 515 English ‘classics’ never saw anything but the economic aspects or, still worse, the wealth or even the profit aspects of things. But none of them really questioned the validity of value judgments that were based on ‘philosophical’ grounds and took proper account of the noneconomic as well as the economic elements of a given case—the value judgments of ‘the writer or statesman who has considered all the causes which may promote or impede the general welfare’ as distinguished from ‘the theorist who has considered only one, though among the most important of those causes’ (Senior, Outline of Political Economy, p. 3). This, as we have seen, was also J.S.Mill’s opinion and in fact practically everyone’s. There is, of course, sound sense in it: one could only wish that the economists of that (or any) period had never forgotten this piece of wisdom—had never been guilty of the Ricardian Vice. 18 However, it still remains true that the real problem of value judgments never occurred to them. To the end of the period, economists considered their recommendations concerning policy as scientific results which followed from scientific, though not purely economic, analysis. In this sense they were after all, as later critics sneeringly remarked, purveyors of recipes. Fortunately they also were more than that. 3. WHAT MILL’S READERS ACTUALLY GOT Mill’s readers got, in the first place, factual information to the extent of about one-sixth of the book. On the face of it, this is a smaller proportion of space than either A.Smith or Roscher allotted to the presentation of facts, and what there is of this factual presentation is extremely ill-balanced, facts about ‘peasant proprietors,’ for example, taking a share more commensurate with Mill’s own interest in the subject than with its probable interest for his readers. But this may be a wrong view to take. As Mill’s preface emphasized, his treatise abounds in practical ‘applications.’ And these have reference to factual material that Mill often failed to present, perhaps because he assumed that his readers could easily supply the deficiency from universally accessible sources—such as the work of Babbage. 1 If we count as ‘factual’ all discussions 18 See above, ch. 4, sec. 2. 1 Charles Babbage, On the Economy of Machinery and Manufactures (1832). This work, which was widely used (also by Marx), is a remarkable performance of a remarkable man. Babbage (1792– 1871), who was one of Newton’s successors in the Lucasian chair of mathematics (Cambridge), one of the founders of the British Association for the Advancement of Science (1831) and of the Statistical Society, and a versatile writer on many subjects, was also an economist of note. His chief merit was that he combined a command of simple but sound economic theory with a thorough first- hand knowledge of industrial technology and of the business processes relevant thereto. This almost unique combination of acquirements enabled him to provide not only a large quantity of well-known facts but also, unlike other writers who did the same thing, interpretations. He excelled, among other things, in conceptualization: his definitions of a machine and his conception of invention are deservedly famous. It is interesting to note that in some points he recognized the priority of Gioja, whose name should in fact be coupled with his. By way of contrast to his sound well-balanced treatment, I mention A.Ure’s (Philosophy of Manufacture, 1835), who also presented interesting facts, but was not Babbage’s equal as an analyst. History of economic analysis 516 in the book that in this sense presuppose factual information, although the information is not actually presented, then, if I have estimated with anything like accuracy, the ‘factual’ portion of the book increases to a little more than two-thirds of the whole, a little less than one-third being left for exposition of the analytic apparatus. In the second place, his readers got a fairly thorough—but none too thorough—grounding in ‘theory.’ However, as has been pointed out already, no contact whatever is made with any statistical method. From another standpoint, we can illustrate Mill’s range of topics by means of the following list of headings that, as the reader can easily satisfy himself, could be lengthened both by including further minor items and by dividing up some of the major ones: prices, price fixing, competition, custom, monopoly; wages and employment, wage policy, trade unions, poor laws, and other items of the Sozialpolitik of that age; socialism, with special attention to Saint-Simonism and Fourierism; producers’ and consumers’ co- operatives; future of the working class; education; population; enterprise and forms of enterprise, capital, profit, interest; saving and investment; technological advance; money and banking, central banking, foreign exchanges, government paper money; crises; foreign trade; colonies; private property, inheritance; partnerships, companies, bankruptcy legislation; rent, ownership of land, primogeniture, peasant proprietorship, metayage, cotter tenancy, slavery; ‘progress,’ ‘maturity’ (stationary state); government policy and government control; grounds for and limits of laissez-faire; public finance, especially taxation and public debts. I do not think that this list is conspicuous for either narrowness of range or remoteness from the practical issues of the day. It should be observed particularly that all that interested later generations could have been hung on hooks presented by Mill without upsetting his system. For instance, later institutionalists could have inserted into Mill’s niches all the additional material of a specifically institutional nature that they might have wished to insert, without thereby destroying the general character of the treatise: there was room for everything within its spacious folds; and everything could have come to it as a development of existing points, nothing need have come to it as a revolution. J.S.Mill arranged his material in five books: ‘Production,’ ‘Distribution,’ ‘Exchange,’ ‘Influence of the Progress of Society on Production and Distribution,’ and ‘On the Influence of Government.’ The last also contains things other than public finance but nevertheless mainly corresponds to A.Smith’s Fifth Book. In Book IV, the shortest, Mill concentrated what he had to say on the subject of economic evolution—a happy innovation in exposition. The titles of the first three books suggest the influence of Say’s arrangement or, rather, a not very felicitous attempt to improve upon it. The central theory of value, which should come first on logical grounds (and does come first with Ricardo and Marx), is presented in Book III as if it had to do only with the ‘circulation’ of goods and as if production and distribution could be understood without it. This is worth mentioning because it points to a fundamental weakness of the ‘classic’ construction. I do not accuse the ‘classics’ of having failed to sense the pivotal importance of the analysis of value (choice) which is, if I may say so, the specifically economic element about the economic process. But there is some truth in Professor Knight’s indictment that the ‘classics’ had ‘no clear or definite conception of the meaning of economy as a process of maximizing a value return’ and that ‘the problem of General economics 517 distribution …was not approached as a problem of valuation at all.’ 2 To this extent, we must qualify our recognition of Ricardo’s chief merit. He and all the ‘classics,’ including Mill, did indeed make progress toward the acquisition of an analytic apparatus that would unify all purely economic problems; but, partly owing to the shortcomings of their groundwork, they never realized its possibilities to the full. They still divorced production from distribution—J.S. Mill even took credit for doing so—as if they were governed by different ‘laws.’ The first to point this out was Ferrara. 3 But Say’s and Mill’s combined authorities kept this plan of exposition alive for many decades to come. It is not worth our while to discuss its variants. Roscher, for example, has: Production, Circulation, Distribution, Consumption, Population—including Credit with Production. The five books are preceded by ‘Preliminary Remarks,’ which, among less interesting things, contain a short sketch of what we should call the evolution of economic society— a universal economic history in a nutshell. This, of course, is not in itself surprising in a work that aimed at doing again what A.Smith had done. But much beyond what tradition leads us to expect from Mill is his handling of the factors to which he attributes causal roles in the shaping of the fortunes of a society or country. Environment, race (racially differentiated quality of human material), class structure, habits or propensities, combine to make a colorful and, what is more, a very realistic picture. There are no intellectualist and in particular no utilitarian errors about it: ‘knowledge’ is considered as a consequence as well as a cause ‘of the state of the production and distribution of wealth,’ and objective conditions receive more emphasis than ideas or principles. Such prefatory sketches of economic history—though of course not always of this quality—became more and more popular as the century wore on: Marshall’s is the peak performance of this type. 4. THE INSTITUTIONAL FRAME OF THE ECONOMIC PROCESS (a) The Institutions of Capitalist Society. Economic sociology covers, first, the facts of economic behavior from which economists forge certain assumptions and, second, the institutions that characterize the economic organization of the societies to be studied. ‘Classic’ practice as regards the former will be more conveniently discussed in the next chapter. As regards the latter we must distinguish three questions. Many writers, primarily the English theorists—such as Ricardo, James Mill, and Senior—did not bother to specify the details of the institutional frame they visualized, but took them for granted. Is it true, as has been averred so often, that they believed in the permanence of the capitalist order of things or even that laissez- 2 F.H.Knight, ‘Ricardian Theory…’ op. cit. p. 6. The second part of that statement goes perhaps too far. But Professor Knight supports it by a highly significant passage from a letter of Ricardo to McCulloch, where the former avers in so many words that the relative distributive shares ‘are not essentially connected with the doctrine of value.’ Though this cannot be taken literally—for it could be refuted from Ricardo’s own argument—it does show that the full implications of the fact that capitalist distribution is a value phenomenon were not clearly seen even by Ricardo. They were seen by Marx. 3 In his Prefazione to Say. History of economic analysis 518 faire capitalism is the only possible form of civilized society? What were the institutions they took for granted? When they did discuss them, what methods did they use? The first question must, I think, be answered in the negative. It is true that Ricardo, for example, by virtue of the very fact that he failed to specify his institutional assumptions, creates an impression to the effect that the problems of social change were beyond his range of vision. But this does not follow. All that follows from his practice is that they were beyond his chosen field of inquiry. There is no reason to believe that, had he offered a description of the institutional frame, this description would have differed significantly (though his value judgments might have) from that of J.S.Mill, who, aiming as he did at systematic completeness, was more explicit. But as we know already (see sec. 1, above) there cannot be any doubt whatever about the latter’s awareness of the historical relativity of social institutions and also of some at least of his ‘economic laws.’ To this extent, the current belief that this awareness was confined to isolated forerunners of later historicism, such as R.Jones and Sismondi, is certainly erroneous. It would be nearer to the truth to say that explicit belief in the permanence or in the unsurpassable excellencies, for all times, of capitalism occurred only in isolated instances. The reader should observe, however, that this does not amount to crediting the ‘classic’ theorists with the idea that the capitalist order is only a historical phase and bound to develop, by virtue of its own inherent logic, into something else. This idea belongs to Marx alone. Even J.S.Mill held only that men could, should, and would change capitalist institutions through a rational perception of what he considered to be their defects. He did not hold that institutions would change of themselves or even that they would have to be changed because they would become objectively untenable. He saw that ‘opinions…are not a matter of chance’ 1 but the product of social conditions, and we might feel tempted to develop this in the Marxist direction. But this would hardly be justified: we must, I think, leave him in the citadel of eighteenth-century faith in intellectual progress, from which he occasionally sallied forth, no doubt, but to which he always returned. Practically, this matters but little. Scientifically, it makes a great deal of difference. The second question is an easy one. Economists, wishing to serve their time and countries, took for granted—and reasoned in terms of—the institutions of their time and countries. Since conditions differed as between different countries, this accounts for certain differences in outlook that, at the time and later, were erroneously construed as differences of analytic principle. The traits of the picture that were selected by the English ‘classics’ stand out very clearly. They envisaged the legal institutions (if we disregard the historical survivals that suggest another type of society) of a private- property economy that left so much room for free contracting as almost to justify the practice of economists to leave limitations out of consideration. This only means, of course, that no account was taken of limitations explicitly and consciously. As a matter of fact, English economists always reasoned with reference to the actual extent of the sphere that English law and administrative practice left for private decision and to the actual use that was being made of this freedom within that sphere—subject to the prevailing moral 1 Last but one paragraph of Book II, ch. 1, 4. General economics 519 habits. Failure to be specific about all this exposed the English ‘classics’ to much mistaken criticism concerning their apparent neglect of ethical aspects. The unit of that private-property economy was the firm of medium size. Its typical legal form was the private partnership. Barring the ‘sleeping’ partner, it was typically managed by the owner or owners, a fact that it is important to keep in mind in any effort to understand ‘classic’ economics. The facts and problems of large-scale production and, in connection with them, those of joint stock companies were recognized by economists after everybody else had recognized them. They received textbook status at the hands of J.S.Mill, who duly blamed A.Smith for his narrow views on corporate business— forgetting only the detail that there was little merit in realizing its importance in 1848 and that actually he was doing no more than what A.Smith had done, namely, describing with sober and somewhat platitudinous common sense that which stood before his eyes. Two further points deserve notice. In the normal case, these firms were supposed to work under what the ‘classics’ called Free Competition. With them, this competition was an institutional assumption rather than the result of certain market conditions. And so firmly were they convinced that the competitive case was the obvious thing, familiar to all, that they did not bother to analyze its logical content. In fact, the concept was usually not even defined. 2 It just meant the absence of mo- nopoly—which was considered as abnormal and was vigorously condemned, 3 but was not properly defined either—and of public price fixing. J.S.Mill took credit, not without justification, for two important steps. First, he emphasized the importance of customary prices, mainly for earlier civilizations and for the Continent, but in certain cases, such as rent and professional fees, also for England. And, second, he emphasized the fact, though custom was the only reason he gave for its existence, that competition often ‘falls short of the maximum’ and that in this case a general correction must be applied, ‘whether expressly mentioned or not,’ to all conclusions arrived at on the hypothesis of perfect competition (Book II, ch. 4, 3). Into such a picture co- operative price setting could enter only, if at all, as another deviation from normal practice like straight monopoly and as a conspiracy against public welfare exactly as it does now. There was, however, an exception to this: in J.S.Mill’s scheme of things, trade unions were a normal element of the institutional pattern and laws against them ‘exhibit the infernal spirit of the slave master’ (Book V, ch. 10, 5). 2 It is interesting to note that J.S.Mill in Book II, ch. 4, of his Principles, which deals with ‘Competition and Custom,’ while expressing a conviction to the effect ‘that only through the principle of competition has political economy any pretension to the character of a science’—which presumably means that there is more of determinateness about prices and quantities of commodities in the case of competition than there is in other cases—he does not think it necessary to state what competition is. The only author to handle perfect or pure competition correctly was Cournot, who also indicated, though he did not formulate in so many words, a correct definition (see below, Part IV, ch. 7, sec. 4). 3 For Roscher’s unconditional eulogy of competition and his uncritical condemnation of monopoly, see op. cit., Book II, ch. 1, 97. History of economic analysis 520 The other point to remember is this. Many English economists were severely critical of the English land system. 4 But when not engaged in criticizing it or in discussing alternatives, they also took it for granted in the sense that they reasoned with reference to it and to the English type of landlords who owned but did not operate large estates. In this particular case, however, reasoning in terms of existing institutions carried an advantage that may be made to stand out by contrast with the reasoning in terms of the owner- managed firm: landlords and farmers being different persons, it was easy for the theorist to keep their economic ‘functions’ distinct; owners of firms—‘capitalists’—being then mostly the same people who also operated these firms, it proved less easy for the theorist to recognize this distinction of ‘functions.’ This teaches us an interesting lesson: it may happen, though we have no right to see in this occurrence more than a stroke of luck, that a particular historical pattern, which taken as a whole has nothing permanent about it, reveals facts and relations that are of general analytic importance. Usually, of course, it is the other way round, and we must always look out, on the one hand, for limitations that their particular institutional assumptions may impose upon the results of the ‘classics’ and, on the other hand, for possible justifications of these same results that may be found occasionally in the peculiarities of the social pattern they envisaged. 4 So much so as to be unable to see at all its strong points. I am not referring now to the political society and the cultural values it produced: a bourgeois radical cannot in fairness be expected to see these. But many economists also had no eye for the advantages of a system that separated the administration of land from the operation of it and thus e.g. eliminated the most serious of the problems of agricultural credit. Rational socialism could do worse than copy the system, replacing, of course, the landlords by a public agency. Our third question—concerning the method used by the ‘classics’ in discussing social institutions—will be addressed, for brevity’s sake, to J.S.Mill alone. 5 As an example, consider his views on inheritance. 6 The discussion culminates in the recommendations (a) that freedom of bequest be the general rule, except for a modest compulsory provision for descendants and for a provision to the effect that no person should ‘be permitted to acquire by inheritance more than the amount of a moderate independence’; and (b) that ‘in the case of intestacy, the whole property escheat to the State,’ also with a proviso in favor of ‘just and reasonable’ provision for descendants. In themselves these recommendations and also the particular ideas of ‘justice’ that enter into them are only interesting from standpoints other than ours: to the historian of civilization they reveal part of the schema of cultural values that was harbored by a leading intellectual who 5 This carries the disadvantage that our answer will be lopsided. Let me therefore state explicitly that more history-minded economists and, a fortiori, contemporaneous students who specialized in the history of institutions are not as a rule oblivious to the indictment I am going to level at Mill. His case is, however, important for general economics because he set a textbook fashion that prevailed far beyond the period under survey. The discussion of property and inheritance, e.g., that we find in the deservedly successful textbook of von Philippovich is exactly on Millian lines as far as its methodology is concerned. 6 Principles, Book II, ch. 2 and Book V, ch. 9. This treatment of the same subject in two different parts of the treatise, which is highly inconvenient for the reader and a bar to any well-rounded exposition, is one of the many symptoms of the haste with which the work was produced. General economics 521 . objectionable piece of methodology is immaterial whenever it can be dropped without forcing us to drop any result of the analysis that is associated with it. History of economic analysis 512. below, Part IV, ch. 7, sec. 4). 3 For Roscher’s unconditional eulogy of competition and his uncritical condemnation of monopoly, see op. cit., Book II, ch. 1, 97. History of economic analysis. the absurd argument of later times about ‘induction vs. deduction,’ but, always remembering that he thought of the theoretical apparatus of economics when speaking of methods of political economy,