SOURCE: "The Sociology of Pareto," in Reason and Unreason in Society: Essays in Sociology and Social Philosophy, Harvard University Press, 1948, pp. 84-103.

[In the following essay, which was first published in 1936, Ginsberg challenges the central points of Pareto's sociological theories.]

Pareto's sociology falls naturally into two parts. The first is devoted to an analysis and classification of the elementary constituents of human nature as manifested in social life. The second is concerned with the interactions of these elementary traits and the changes which occur in their distribution in the different classes of society. The method followed is inductive and comparative, that is to say, it starts with empirical facts such as beliefs actually held in different societies, maxims of conduct accepted by them and the like, and it seeks to analyse out the constant and variable elements in these forms of behaviour and to discover the laws or uniformities which determine their mutual relations. Incidentally, Pareto discusses at great length the nature and importance of what he calls the "logico-experimental" method in social science, but he hardly lives up to his own requirements. The defini tions given of fundamental terms are obscure, and they are not, as they might be expected to be, gradually clarified by "successive approximations." Further, what appears to me the most interesting portions of the treatise, namely, those devoted to the dynamics of social change, are very inadequately supported by empirical evidence, the facts given being hardly more than illustrative of the hypotheses put forward. The plan of the work is conceived on an imposing scale and it is carried out with great independence and a wealth of learning. It is therefore worthy of the serious consideration of sociologists.

The analysis of the fundamental forces of social life is carried out mainly by means of a classification of human actions into logical and non-logical, and by a more detailed account of the non-logical acts which brings out their overwhelming preponderance in human affairs. In this account Pareto pays no attention to the work of psychologists, but proceeds to put forward independent hypotheses suggested, as he thinks, by direct inspection of the facts. His neglect of psychology has resulted in an extremely vague use of such terms as "sentiments," "instincts," "interests," which has made a proper understanding of his views more difficult than it needs to be. But shed of technicalities and expressed as a first approximation, his conclusions are hardly revolutionary. They amount to this: that people perform and always have performed many acts without knowing why they do them (i. e. by habit and instinct); that the real drives of action are often quite different from the purposes which the agents consciously entertain; that in the pursuit of given conscious ends people often attain quite other ends than those aimed at, either because they adopt the wrong means, or because they do not foresee die remoter consequences of their acts; that men, having a hunger for logic or reason, will try to give a reasoned explanation or jus tification of acts they do from obscure or unconscious motives. The bulk of the treatise is devoted to an analysis of the non-rational elements in human conduct and of the fictions which are invented to give a flavour of rationality to conduct that is really the result of feeling and impulse.

The classification of acts into rational and non-rational, or in Pareto's not very happy terminology, logical and non-logical, turns upon the distinction between means and ends which he assumes without further inquiry to be applicable to all human behaviour. Briefly, acts are non-logical (i) when they serve no end subjective or objective, e. g. futile or non-adaptive instinctive acts, if such there be; (ii) when the agent thinks a particular end is being realized but nothing is in fact achieved through the act as judged in the light of wider knowledge, e. g. in magical operations; (iii) when there is an objective end but the subject is not consciously aiming at it, e. g. in theoretically "pure" instinct; (iv) when an end is actually achieved which differs from the end the subject sets to himself, whether the objective end would or would not have been acceptable to him could he have foreseen it. Briefly, acts are non-logical when the subject acts without explicit knowledge of the purpose of his action, or, having such knowledge, chooses means which in the light of better grounded information are either not likely to achieve the purpose, or to achieve something else. By contrast acts are logical when the consequences anticipated by the subject are identical with the consequences that might reasonably be anticipated in the light of wider knowledge. So far logic, or rather rational reflection, is not concerned with ends at all, save perhaps that in order to act rationally you must know what you want. Logic is concerned rather with the appropriate linking of means and ends. But this position is not consistently maintained. Unfortunately, Pareto makes no attempt to classify the ends of conduct or to relate them to the fundamental drives. These are said to consist of sentiments, tastes, proclivities, inclinations, instincts, residues, and interests. The residues are said to include neither the simple appetites or instincts nor the interests, but to "correspond to" instincts or appetites. By this, as we shall see later, appears to be meant that there is a residue when an appetite or instinct does not act itself out simply but finds expression in an indirect or disguised form. Thus sex conceived as mere union of the sexes is not a residue, but it is residual, for example, in the behaviour of people who preach virtue as a way of lingering in their thoughts on sex matters. If this is the correct interpretation, the residues are not themselves drives, but rather ways in which the fundamental drives disguise themselves, and they are thus non-logical in the sense of obscuring the nature of the impulses really at work. So far it would seem that the adequate fulfilment of any impulse, provided it is conscious and direct, is logical. But there are many passages in which it seems to be suggested that it is more logical to act in accordance with one's interests than in accordance with other drives. What, then, are the interests? They are said to consist in impulses to acquire material goods, whether "useful" or merely pleasurable, and to seek consideration and honour. It is not easy to see why these goods are singled out as rational, why, for instance, it is more logical to pursue honour and consideration than to satisfy other social impulses, or let us say, the desire for knowledge. Or is the pursuit of interests regarded as logical not because of the particular ends involved, but because in achieving them men can be shown to act with greater circumspection than in other activities in the choice of appropriate means?

The difficulty may be illustrated by reference to the frequent description of economic activity as typical of logical behaviour. It is easy to see that economic activity contains a logical element in so far as the means chosen are technically appropriate to given ends. But economic behaviour is clearly non-rational in so far as men acting economically are not aware of the motives which impel them. A man's choice of profession may be as an intention clearly envisaged. But the motives of the choice are often obscure and even unconscious. It may be influenced by all Pareto's residues, for example, by authority and prestige, by sociability, by the persistence of abstractions and what not. Further, it is clear that men in seeking economic satisfaction attain ends which they did not foresee and do not want. Men do not want war, but thenbehaviour leads to it. Business men and workers do not want unemployment, but the outcome of their linked activities is to produce it. If success in the fulfilment of impulses is the criterion of logical behaviour, economic activity must be largely non-logical, since it fails to secure for the masses of men the conditions of a purposeful life. In short, without an examination of the ends of human endeavour and of their relations to each orner as well as to the means available for their realization, it is impossible to throw much light on the rational elements in behaviour, or on the relation of economic to other activities in human life.

There is a further complication which must now be considered. In the later portions of the treatise a person is said to act logically in so far as he tries to secure a maximum of individual utility. This means action in accordance with what is "advantageous" or "beneficial" to him and involves a comparison of different satisfactions in accordance with some norm. Presumably a logic is required for making these comparisons, but this line of thought is not pursued. In economics it would seem the individual is assumed to be the best judge of his own interests or utility and to act rationally in regard to it. But outside economics there may be an infinite number of norms and therefore an infinite number of possible maxima of utility. The choice of the norm itself is arbitrary and non-logical. Thus we cannot say whether it is to the advantage of the individual to suffer physically for the sake of a moral satisfaction, or whether it is better for him to seek wealth or to apply himself to some other pursuits. Despite the elaborate discussion of utility, there is extraordinarily little to be gained from it. Since the norm is arbitrarily chosen we can only determine the maximum utility for an individual from the line of conduct that he actually adopts. That is the maximum which in fact appears to him to be such and that appears to be the maximum which in fact he pursues. Or is it possible for the individual to make mistakes regarding what is to his advantage apart from the mistakes he may make in the choice of means? If so, a logic of ends is required which would enable the individual to distinguish apparent and real advantage, and that would soon lead to an ethics of the teleological kind which Pareto despises along with all other brands of ethics. In brief, Pareto's treatment of the logic of behaviour leaves out of consideration what to most people will appear essential to it. Rational behaviour no doubt requires us to know what we want and to choose means in a manner which will stand the test of empirical verification. But a logic of behaviour would also have to discover whether the norms that individuals adopt in relation to the ends that they pursue are self-consistent, and whether they form part, or can be made to form part, of a systematic and ordered whole. Such a logic obviously could not be confined to the norms governing the acts of particular individuals, since it is equally or more important to inquire how far the norms of different individuals or groups are or can be made to be compatible and perhaps harmonious. Pareto makes no attempt whatever to deal with these problems and asserts as a self-evident dogma that norms are just the expression of "sentiments." There is, for example, no criterion save sentiment for choosing between a society based on large inequalities of income and one based on approximately equal incomes. If we admire supermen we will assign zero utility to the lower classes; if we love equality we will prefer the type of society which secures to the lower classes an equal share in the goods of life. Is reason really helpless in the face of such a problem?

It may be suggested that before dealing with the ultimate problems of valuation here involved there is a good deal that reason can do by way of clarifying the issues and settling questions of fact. Pareto, together with other antiegalitarians, assigns a meaning to the principle of equality which egalitarians are not concerned to defend. The principle does not assert either that men are equal in endowment or that they should be treated equally. It is concerned, negatively, to exclude arbitrary assignments and, positively, to base distribution on a general rule impartially applied. If for the sake of argument it be agreed that this rule is that distribution should be in proportion to the needs of individuals with a view to the realization of such capacity in them as they have, it will be seen that this does not involve equality of treatment. Certain questions of fact then become very important. Firstly, what is the extent of the differences in capacity between individuals, and are these so great as to justify us in regarding some of them as supermen and large numbers of the masses of men as having zero value? Secondly, how great are the differences in external conditions which are really required in order to enable the alleged supermen to fulfil their capacities, and can these differences in conditions only be assured them by a system of private property, involving the amount of inequality that now prevails? Thirdly, we need to know what effects upon the total available for distribution will be produced by adopting the principle of equality, and this raises questions not only of economics but of psychology also; since we need to know how incentives work in different economic systems. Ultimately, no doubt, when these questions of fact have been answered, value judgments will have to be discussed, but it may be doubted whether they would then loom so large in the minds of the disputants. Pareto, at any rate, does not discuss the nature of value judgments, but merely asserts dogmatically that they express nothing but "sentiments." He is impressed by the fact mat in moral judgments, for example, people are swayed by superstitions and prejudices which deceive themselves and others. But this applies to all human thought and action and if seriously pressed would lead inevitably to the conclusion that there can be no logical thought or action at all. Pareto also makes much of the argument that if ethical judgments permitted of rational examination, ethics would have made greater progress than it appears to have made since the days of Aristotle. This, however, is not substantiated by any examination of ethical systems. Moreover, it would apply with equal force to, say, economics right up to the eighteenth or nineteenth century, since it is by no means certain that any great advances were made in it in the interval between these periods and Aristotle's discussion of economic problems. Curiously enough, Pareto thinks that ethical discussions, though logically futile, have had great influence on social life: "they are forever shaking the foundations of the social order." A philosopher might say that this was no mean achievement for mere "derivations."

The difficulties in Pareto's theory of non-logical actions are due ultimately to his failure to inquire more fully into the nature of logical, or as I should prefer to say, rational action. The function of reason is, according to him, exhausted in linking means and ends appropriately. But even Hume, who held a similar view, admitted that thought can influence action by disclosing the hollowness of objects of desire which before reflection excited lively passions, and it is clear further that many of our most passionate devotions are only possible on the reflective level. Thought and impulse cannot, in fact, be sharply dissevered and the ends of life cannot therefore be relegated to the sphere of impulse alone. Ends and means again profoundly affect one another, and it is impossible to deal logically with means without clarification of the nature of the ends. Reason, too, is concerned with the relations of the various ends to each other, with the possibility of their mutual consistency or harmony, and in cases of conflict with the grounds of preference. An element of generality in preferences cannot surely be denied. We prefer not only particular things to other particular things, but kinds of things to other kinds, and our orders of preference have a certain constancy; the business of reason is to reflect on the standards which are implicit in these intuitive judgments. If action can be rational at all, such reflection on values and standards of values must be able to claim validity. If, on the other hand, our choices and preferences are utterly arbitrary, there can be no sense in speaking of any action as rational or as logically justifiable. All that we could then do in a theory of conduct would be to describe and classify human actions as sheer matters of fact, and at most to inquire into the relations which men subjectively set themselves and the ends which are in fact attained by them. In such a theory of human conduct the belief that some acts are "logical" would only be one fact among others, and to deal with it "logico-experimentally" would mean to inquire whether it in fact satisfies the queer hunger for logic that men appear to have, or whether it is useful as a means to other ends. Its power to satisfy the demands of logic at any rate does not seem to be very great.

The theory of non-logical actions is further elaborated by Pareto in his doctrine of the residues and derivations. Formal definition of the residues is lacking and we can only rely upon an analysis of the very numerous examples given and the classification offered of the principal types. To begin with, the residues are not identical with what psychologists call instincts. They are expressly said not to reflect all the instincts, and to include neither the simple appetites, tastes or inclinations, nor what he understands by interests. Yet the residues "correspond to" the instincts, and it is pointed out that there may be residues corresponding also to other impulses, though these are not further dealt with. The meaning seems to be this. In so far as the fundamental impulses are realized directly without diversion or substitution of object they do not give rise to residues. Animals who are supposed to act on pure instinct can have no residues, and in human beings the simple satisfaction of food or sex impulses is not residual. The sex residue becomes important when we recognize its influence in such phenomena as asceticism. Only creatures capable of theorizing and therefore of deceiving themselves can have residues. A classification of the residues would thus be a classification of the different ways in which the fundamental impulses realize themselves in human behaviour, excluding, on the one hand, fully conscious and experimentally directed behaviour, and on the other, behaviour which is based on simple and direct impulse. If this is the right interpretation, the ultimate dynamic elements in human nature are not to be found in the residues but rather in what Pareto calls the sentiments. The residues are the patterns or principles in accordance with which the sentiments work, and they can only be discovered by an analytic and comparative study of complex acts, in which the influence of the sentiments may not at first sight be at all obvious. In studying them Pareto is thus trying to discover the different ways in which the "sentiments" unconsciously affect belief and action.

Pareto does not undertake, as might have been expected, an analysis of such processes as repression, projection, aim inhibition, substitution or sublimation, symbolization, dramatization, and the like. Of the work done in this connection he appears to have no knowledge. Yet despite his repudiation of psychological methods, what he here attempts to do is psychological and not sociological. He does not endeavour to study the social influences affecting belief and behaviour, but on the contrary finds the explanation of social behaviour in the permanent underlying psychological elements and their varying combinations in different societies, and his conclusions must be therefore tested from the point of view of their adequacy in the light of psychology. Thus regarded, his account is not very impressive. He gives six classes of residues with numerous subdivisions, namely, combinations, persistent aggregates, sociability, activity, the integrity of the individual, and sex.

The residue of combinations is of such wide scope that it really includes the whole synthetic activity of the mind, the operations of science and of constructive imagination, and, indeed, all forms of association. Behind all these there is apparently a single drive to combine elements into aggregates. That the mind has a tendency to combine or synthesize is true, though it is equally true that it has a tendency to break up or analyse, and no account of mental activity can be given unless both these tendencies are taken into consideration. But in any case the resort to such general tendencies takes us but a little way, and it is important to discover the principles in accordance with which the various forms of analysis and synthesis are effected. As far as the underlying motive is concerned it cannot be assumed that it is just an urge to combine. This is certainly not the case either in purely theoretical or practical activity. When Pareto comes to distinguish the different types of combination he is far too ready to rely on his assumed general tendency just to combine. This leads him to stress unduly the arbitrariness of the combinations, as in his account of magical operations, or to adopt familiar principles of association such as of similars or opposites which permit of more refined psychological analysis. Magical practices, for example, do not rest upon a general tendency to combine anything and everything, but upon a readiness, under the stress of practical needs and in the absence of a critical method, to rely on coincidences. There is always an element of experience behind them, though this is too readily generalized and no adequate means are available for disentangling the subjective and objective factors. The tendency to generalize on a slender basis has a much better claim to be called a residue than the tendency to arbitrary combinations. No doubt the experiences underlying a particular belief may be difficult to detect, and Pareto is undoubtedly right in stressing the difficulty of tracing the historical origins of ancient or primitive magical beliefs. Yet occasionally what appears to be an arbitrary association can be shown by historical analysis to be based on intelligible, though of course not scientifically founded, associations. To say "five" in order to avert the evil eye may seem hopelessly arbitrary. Yet in Morocco, according to Westermarck, this is a remnant of the ancient practice of throwing the hand forward with outspread fingers and saying "five in your eyes"; which has now become attenuated to just saying "five" or even "Thursday." Here the original practice requires examination in accordance with the psychology of the magic of gestures, and is in line with much else of the pantomimic or dramatic in magic. In all cases an analysis of the objective and subjective conditions determining belief in particular connections is necessary. It is mere evasion of the issue to appeal to purely general tendencies capable of explaining all connections, and therefore not specially helpful in dealing with any of them.

Under the heading of the persistence of aggregates Pareto brings together a number of interesting facts, but here again the analysis is not very illuminating from the psychological point of view. At least two rather different things are here confused. One is the tendency for sets of psychological dispositions which have grown up between a person and other persons or things to cohere and persist in time. This requires analysis in terms of Shand's doctrine of me sentiments and the theory of complexes. The other is the tendency to individualize or to regard as single entities groups of experiences in relation to which sentiments have grown up, and to attribute to these entities, real or imaginary, any further attributes which our emotional attitude to them requires. The phenomena here included have usually been studied under the headings of animism, animatism, personification, and the like, and to their elucidation, I should say, Pareto makes very little contribution, except perhaps in the stress he lays on the influence of personified or reified abstractions on social life.

Under the residue of activity Pareto discusses facts which are usually treated by psychologists under the heading of the expression of the emotions and other drives and the pleasure taken in the exercise of faculty. He rightly stresses the part played by fantasy or imagination in providing symbolic expression of the emotions, but does not further analyse symbolism, nor does he inquire into the reasons why symbolic substitution is needed. The principal examples that he uses are taken from the phenomena of religious exaltation, such as revivals, mystical ecstasies, and the like. But his interpretations of these phenomena are of very doubtful value. There is much to be said for the view that ecstatic manifestations are due not so much, as he thinks, to a sheer need for activity as to the need for relaxation from the strain and monotony of ordinary life and for release from repression and conflict. In political agitation, which is another of his examples, the feeling that "something must be done" is hardly due to a desire for activity as such. On the contrary, in the case of leaders and agitators it is rooted in deep conflicts, and in the masses the readiness to yield to leaders who claim to get things done is a reflection of their own apathetic anxiety and the disinclination or inability to do anything effective themselves. The whole discussion is extraordinarily vague. It is not at all clear whether the residue of activity is a specific tendency to act, or whether it is a collective term for the need of expressing all the emotions and impulses in outward acts. In any event, to find the residual, that is to say, the constant and invariable elements in religious manifestations in bare activity without any attention to the nature of the emotions and needs which are at work can hardly be said to constitute a profound contribution to the psychology of religion.

The residues of sociability include a number of tendencies, principally the desire for uniformity, the desire to impose uniformity on others, the hatred of the new, counterbalanced by interest in novel combinations, the tendency to pity balanced by cruelty, the tendency to share with others, to suffer for them even to the extent of self-sacrifice, the need for the approval of others, the compound of submission, fear, respect, pride, and domination which constitutes the psychological basis of hierarchical organization, and others. The account given of these tendencies, and especially the discussion of asceticism, is of great interest, but it is not very precise or systematic, and there is too great a readiness to invent instincts ad hoc. The desire for uniformity, for example, is hardly to be accounted for in terms of a general instinct to imitate. Psychologists are not agreed that such an instinct exists, and, in any case, the respect for rules qua rules is very complex. There is a rational element in it based on the recognition that for societies to cohere there must be a readiness on the part of individuals to conform to rules without insisting on a reasoned justification on every occasion of their application. Such recognition may not be very clearly present to the minds of all members, but there is always present a feeling that order must be maintained and that there must be rules. Whether rational or not, the feeling of respect for accepted rules does not rest on sheer imitativeness, but on deeper social bonds. The purest form of the tendency to imitate Pareto sees in fashion; but here again the analysis strikes me as superficial. Fashion is not based on a tendency to imitate anything and everything, but rather upon an identification with those who have social prestige, and thus involves at least as much desire to be distinguished from others as to be like them. In regard to his account of the other social tendencies, it may be noted that Pareto owes much of his recent popularity to the cynical account he gives of humanitarianism. Anything more remote from logico-experimental evidence can hardly be imagined. He imputes all sorts of motives to humanitarians without the slightest attempt at proof, and indulges in vast historical generalizations without anything like adequate inductive verification. It is one of his favourite generalizations that repugnance to suffering and the tendency to pacifism are characteristic of élites in decadence. One might counter this by formulating ad hoc the parallel generalization that brutality and war-mongering are characteristic of élites uncertain of their power. For neither generalization is there adequate evidence of the "logico-experimental" kind that Pareto considers essential for a scientific sociology.

The treatment of asceticism as a residue of sociability is striking. Pareto interprets ascetic behaviour as in the main due to a hypertrophy or perversion of the social instincts, or, as it might perhaps be better put, as an exaggeration of the need to control and master the self-assertive impulses. The interpretation is worked out with much insight, but perhaps insufficient attention is paid, especially in the elaborate discussion of flagellation and allied phenomena, to the sado-masochistic elements in asceticism.

The residue of the integrity of the individual broadly includes all reactions tending to maintain equilibrium or to restore a violated equilibrium. It is not at all clear whether this is a specific tendency, how it is related to what psychologists call the self-regarding sentiment, and whether it is not merely a collective term for a group of reactions. In a sense all responses whatever may be brought under it, since they may all be interpreted as the result of a disturbance due to inner or outer stimuli. The examples that Pareto gives here are mainly derived from ritual. Thus purificatory rites are regarded as efforts to restore the integrity of the individual which has been disturbed by pollution. But here, as in the case of the residue of combinations, the tendency appealed to is so general that it would explain all ritual whatever and therefore throws but little light on any. Why is the integrity of the individual endangered by contact with blood, and why is the malaise produced by this pollution got rid of just by this or that form of purificatory ritual? To say that these are just arbitrary combinations is surely to abandon the problem. There is, I think, a somewhat similar difficulty in Lévy-Bruhl's treatment of what he calls "transgressions," with which Pareto's discussion has some affinity. For example, it is not an explanation but only a restatement of the problem to say that the horror of incest is due to the fact that it is treated as a transgression.

Somewhat unexpectedly, Pareto brings under the residue of the integrity of the individual the demand for equality by inferiors. He interprets this demand as really a hidden desire for another kind of inequality or selfish privilege. Perhaps this is not as pessimistic a view as that of Freud, who suggests that social justice means that we are ready to deny ourselves many things so that others may have to do without them as well. But what direct evidence is there of the real motives which inspired the leaders of humanitarianism or the mass of their followers? In general, Pareto's attack on humanitarian ethics hardly calls for detailed analysis here. His arguments are very far from being presented with the detachment which he considers so necessary in a logico-experimental sociology and they abound in value judgments. Since such judgments are in Pareto's view nothing but the expression of "sentiments," his discussion has merely biographical interest in so far as it throws light on Pareto's own mentality.

In his treatment of the residue of sex Pareto brings out with great gusto the vagaries and inconsistencies of sexual morality and he stresses the well-known fact that behind the condemnation of sex there is often hidden an excessive preoccupation with it. He might have generalized this and shown how in the case of other impulses the repression of self takes revenge in the reprobation of others. In this discussion even more than elsewhere the fundamental weakness of his method is revealed. In his anxiety to stress the constant and invariable elements in sex he fails to come to grips with the medley of social forces affecting the morals of sex-relationships and to deal with the variations that have been observed in them. One almost gets the impression that the rules regulating the relations between the sexes and the respect for chastity are based on nothing but disguised sexual greed and jealousy. There is no study of the need to canalize and control the sexual impulses in view of the manifold derangements of which they are susceptible, no examination of the relation of sex to tenderness and affection and the social impulses, or of the problems connected with precocious sexuality, or of the influence on sex relationships of the institutions of property and the family. In short, there is no treatment of the numerous factors, sociological and psychological, which must be taken into consideration in a just estimate of sex regarded as a constant and invariable drive.

The theory of the derivations is intended to furnish a psychology rather than a logic of error, that is, to reveal the hidden forces which lead to error and make it acceptable rather than to disclose the logical structure of erroneous reasoning. Clearly the derivations must be rooted in the residues. There is, in fact, a double connection between them. Firstly, men have a strange hankering after logic and they are not satisfied unless they can give reasons for their actions and beliefs. This Pareto regards as being one form of the residue of combinations which supplies the drive both for logical and non-logical reasoning. But, secondly, particular derivations owe their-strength and influence to other residues and the "sentiments" underlying them. That the ultimate driving power lies in the sentiments or their residual manifestations and not in the theories which are offered to account for behaviour can be seen, Pareto argues, from the fact that the feelings or sentiments remain essentially unaltered despite changes in derivations and theories. He is fond of using in this connection examples derived from the history of morals. "A Chinese, a Moslem, a Calvinist, a Catholic, a Kantian, a Hegelian, a Materialist, all refrain from stealing; but each gives a different explanation for his conduct." Strangely enough, Pareto claims greater constancy for moral rules than is needed for a rationalist ethics, and he makes no attempt whatever to account ior the variability of the moral judgment. The residue of the integrity of the individual will account for the laws of theft, but only if you are content to neglect the enormous variations that are found in the laws of property and consequently in what is regarded as theft, and so with other institutions. While both the varying and constant elements in morals contain both rational and non-rational elements, I do not think that Pareto provides any method for estimating their relative strength, for determining, for example, the ràle of reason in the history of law or indeed of any social or political movement; though nowadays no one would be concerned to deny the importance of the irrational or even unconscious factors in human life.

The interest of Pareto's treatment of the derivations lies largely in his acute and penetrating criticism of many famous social theories, for example of Benthamism or of the General Will, for which it provides an occasion. It is not particularly successful as a systematic exposition of the sources of prejudice and modes of sophistication. The derivations are grouped under four headings: affirmation, authority, accord with sentiments, and verbal arguments. Under the first are included assertions claiming authority simply as assertions. The examples that he gives are maxims such as "Silence is an ornament to all women"; "Neither do nor learn aught that is shameful." These correspond, I think, to what Mill calls fallacies a priori, mere assertions claiming to be self-evident. It is not easy to see where derivation comes in here, since by definition no reason is given for the assertion, unless what is meant is that if challenged the answer will be just their indisputability. Pareto does not discuss the psychological factors which produce the feeling of self-evidence, nor why assertions that are regarded as self-evident in one age are considered nonsensical or false in another. Occasionally the examples chosen beg important questions of theory, as when aesthetic judgments are interpreted as unconscious conversions of subjective likings into assertions of objective fact. I doubt whether the derivations of affirmation form a distinct class, and in most instances they pass readily into those resting on authority or verbal argument. The derivations of authority have long been familiar, and among the writers whom Pareto quotes in other connections, he might here have referred to Bentham who has given an elaborate discussion of them. The derivations of accord with feeling present a good opportunity for a consideration of the subjective factors of belief, and Pareto has much of interest to say on the influence of the self-assertive and the social tendencies upon belief. Here his analysis would have been greatly improved had he paid attention to the work of modern psychology and especially the psychology of the unconscious. In his discussion of the derivations of verbal argument the logical aspect is not kept very distinct from the psychological. Perhaps the most valuable part of his exposition is his insistence on the tendency of abstractions to persist and to become the nuclei of powerful emotional dispositions.

Pareto has no doubt that the residues remain constant or undergo only slight and slow change even over long periods. But though this may be true in a sense the proof offered is not very convincing. The residues are so vaguely defined that it is easy to find what is alleged to be the same residue in what are apparently very different social movements. In this way it is argued, for example, that behind Ancestor worship, Polytheism, Catholicism, Protestantism, Nationalism, Socialism, Humanitarianism there is the same residue of group-persistence. So again the residue of individual integrity is regarded as constant, because though it is not so strong in the modern plutocracy as it was, say, in the feudal nobility, the loss is made up by the growth of self-respect on the part of the lower classes and in the recognition that even criminals have a personality deserving of consideration. A humanitarian may be pardoned for thinking this "compensation" a matter of some importance. Even more surprising is the claim that the residues of combination have not changed much if the class is considered as a whole; on the ground that territory formerly occupied by magic, theology, and metaphysics is now increasingly occupied by experimental science which is also a product of the residue of combinations. Units of comparison so pliable and interchangeable are hardly what one would expect to find in a logico-experimental sociology.

The most interesting and suggestive part of Pareto's treatise is that concerned with the dynamics of social change and the factors determining social equilibrium at any one time. The social system is conceived as made up of the elements which have hitherto been considered in abstraction but which in fact are in a relation of mutual dependence. The elements in question are the residues, the derivations and the interests, and since these are differently distributed in the population, account has to be taken of individual differences and of the amount of circulation or movement from one group to another that occurs in given societies. The important influences he thinks are those exerted by the interests, that is, broadly, of economic factors on the residues and upon their distribution in the different social classes and the converse influence of the changing distribution of the residues on the interests. On the other hand, the influence of theories or derivations on the residues is slight, if not negligible. The interaction of these elements is such as to result in undulations or oscillations, movement in one direction usually setting up compensatory movements in the opposite direction, with the result that change is not in a straight line but is cyclical in character.

The individual differences that Pareto considers at length are those in the intensity or strength of the residues of combinations and persistent aggregates. He lays special stress on one particular classification. In both the ruling classes or élite and in the masses, though in different proportions, two types are to be found. There are, on the one hand, individuals of the speculator type, enterprising, eager for new experiences, imaginative, expansive, fertile in new ideas. Contrasted with them are people of the rentier type, timid, conservative, anxious to preserve what has been won, averse to anything new. The differing relative proportions in which these two types are combined in the governing class and the extent to which recruitment from below is permitted determine the different types of social structure and civilization. In the political sphere, for example, if the governing class consists of individuals in whom Class II residues predominate over Class I residues, we find types of government which rely chiefly on physical force and on religious and similar sentiments; on the other hand, if the ruling class is chiefly of the speculator type, we find types of government relying chiefly on intelligence and cunning, and appealing either to the sentiments of the multitude, as in the theocratic forms of government, or else playing upon the interests, as in the demagogic plutocracies of modern times. The changes that occur as a result of the predominance of one or other of these types are not, however, in a continuous direction. Compensatory movements occur, whether as a result of internal changes or of war, and oscillations of varying length are thus produced. It is to be noted that the distinction between the speculator and rentier types does not quite correspond to that between liberal and conservative in the political sense, since the speculators will ally themselves with or make use of liberals and conservatives alike, and even of anarchists if it suits their purpose. Revolutions occur mainly when the ruling class, relying too much on the combination residues, develops an enervating humanitarianism and is disinclined to use force, especially if it cultivates a policy of exclusiveness and does not find ways of assimilating the exceptional individuals who come to the front in the subject classes. On the other hand, a governing class may also encompass its own ruin by accepting, for their economic value, individuals who are well endowed with Class I residues, and this may end in the government passing from the lions to the foxes. History shows, Pareto thinks, that changes in the proportions between Class I and Class II residues in the élite do not continue indefinitely in one direction, but are sooner or later checked by movements in a counter-direction. In this way the modifications in the élite are shown to be among the major factors determining the undulatory form of social change. They are correlated, it is claimed, not only with political transformations but also with economic cycles and with oscillations in thought and culture. Thus in periods of rapidly increasing economic prosperity the governing class comes to contain greater numbers of individuals of the speculator type, rich in Class I residues, and fewer of the opposite type; while the converse is the case in periods of economic depression or retrogression. With these alternations are connected also the oscillations that Pareto traces in the history of thought, expressed roughly in the conflict between "reason" and "superstition," scepticism and faith.

It will be noted that in this theory Pareto is not merely replacing the "Marxist" conception of a struggle between bourgeoisie and proletariat by that of a struggle between speculators and rentiers. It is essential to his thesis that the residues are differently distributed in the ruling and ruled groups, and it is on the balance of the residues in both groups that social equilibrium depends. Further, it is in this part of his inquiry that Pareto makes the transition from individual psychology to sociology proper, that is to say to a study of interactions between individuals. His view does not imply that the course of events is determined by the schemes of individual speculators who rule the world by deliberate and concerted stratagem. Their policy is the resultant of a complex set of forces and an infinite number of acts each initiated by the particular circumstances of the time, but leading collectively to results which individually they do not foresee, despite the fact that they may have a clearer conception of their own interests than the masses have of theirs. Here as elsewhere in the Treatise Pareto insists on the great complexity of social interactions and on the need for replacing the notion of onesided causality by that of mutual dependence of the factors involved.

How far the theory of the circulation of the élites is to be interpreted in biological or genetic terms is not very clear. Pareto undoubtedly thinks that the residues are determined by inherited constitution. Further, it would seem that in his view "aristocracies" tend to die out in the sense of leaving no descendants: "History is the grave-yard of aristocracies." On the other hand, one gets the impression that according to him the residues are remarkably constant in a given society taken as a whole apart from infiltration of individuals from other societies. The changes that occur are rather in the distribution of the residues in the different portions of the population and the opportunities offered for their manifestation. Such changes might well occur without involving any genetic changes in the stock and be largely socially conditioned. Pareto refers now and again to the work of the Anthropo-sociologists, e. g. Lapouge and Ammon, but he seems to have paid little attention to modern studies of individual differences and the effects of differential fertility, and one cannot be sure of his attitude to them. As the residues clearly involve temperamental traits in addition to cognitive ones, and as the evidence of individual differences in temperamental traits is very slight, perhaps he would not have been able to get much help from these studies.

Pareto supports his theory of social change by numerous examples derived from the history of Graeco-Roman civilization and of modern Europe. Brilliant as the exposition is, it is hardly adequate to establish the periodicity of social and political movements as a regular law, or the correlations alleged between these movements and the history of thought and culture. The proof of such a law would require a much more exact social morphology than he provides and an extension of the inquiry to non-European civilizations. It would also require independent evidence of the mental make-up of the different social groups, especially of the individuals directly concerned in social movements, and a more exact determination of the nature and extent of what he calls the circulation of élites. It may be remarked that he makes very little use of the work of others. Occasionally he might have found support for some of his theories. It is worth mentioning that Pirenne's later study of European capitalism, and the explanation that he gives of the alternations traced by him between periods of innovation and periods of stabilization, appear to be in line with Pareto's views.

If the occurrence of undulatory movements in history be granted, there remains the important problem of their significance from the point of view of long-range trends. Pareto himself grants that in economic production and in the arts and sciences there has been on the whole a movement forward, or as he expresses it, Class I residues and the conclusions of logico-experimental science have forced a retreat on group persistences. But he insists that this growth in the power of reason has not affected political and social activities to any great extent, and that in any case there is no ground for the belief in continuous progress. The notion of progress is never mentioned by him without bitter derision. But it will be noticed that, though according to him there can be no reliable criteria of progress he does not hesitate to speak of decadence which requires criteria of the same kind. To me it is clear that Pareto has developed no adequate method for estimating the ràle of reason in law, morals, and politics, and that he vastly underestimates what has on the whole been achieved in these directions. The growing interconnection between economic and social and political movements which he himself stresses is an important phenomenon and one which may compel humanity to make increasing use of rational agencies. The fact also that the notion of conscious control of social change in its application to humanity as a whole is relatively new must be taken into consideration in estimating future trends. No one nowadays believes in automatic progress or in indefinite and unlimited perfectibility. What is asserted is that it is theoretically possible to formulate a coherent ideal of human endeavour, and that from a study of the failures as well as the successes of mankind in dealing with its problems, there is ground for the belief that such an ideal permits of realization if men are prepared to work for it. Pareto's denial of human progress rests upon (a) his disbelief in any rational ethics; (b) his view that history so far has disclosed no significant changes but only oscillations. As to (a), I do not find that he provides any reasoned justification for his disbelief. As to (b), it seems to me that he greatly exaggerates the constant elements in human history, and that if there is no law of human progress neither is there any law of cyclical recurrence. From the point of view of policy, in any event, if a choice is to be made between persistent aggregates and combinations, I see no reason for not choosing combinations.