Holism and Individualism in History and Social Science
Holism and Individualism in History and Social Science
HOLISM AND INDIVIDUALISM IN HISTORY AND SOCIAL SCIENCE
In most recent philosophical discussion, the contrast between holism and individualism in history and the social sciences has been presented as a methodological issue. Stated generally, the question is whether we should treat large-scale social events and conditions as mere aggregates or configurations of the actions, attitudes, relations, and circumstances of the individual men and women who participated in, enjoyed, or suffered them. Methodological individualists say we should. Methodological holists (or collectivists, as some prefer to be called) claim, rather, that social phenomena may be studied at their own autonomous, macroscopic level of analysis. Social "wholes," they say, not their human elements, are the true historical individuals.
This issue obviously bears directly upon the way we are to conceive the relations between such social sciences as psychology and sociology, and between these and historical inquiry. But it is commonly thought also to involve us in wide-ranging metaphysical problems—those of historicism and organicism, for example—and to have grave ethical and political implications as well. Sir Isaiah Berlin, in Historical Inevitability (Oxford, 1954), moves quickly from methodological to metaphysical issues when he represents holists as believing in "invisible powers and dominions," conceived as "impersonal entities at once patterns and realities, in terms of which … men and institutions must behave as they do." And May Brodbeck, in "Methodological Individualisms: Definition and Reduction," expresses a common opinion when she writes: "Culturally, holism is intimately connected with hostility toward the liberal political individualism of the Western tradition." Individualists, in their turn, have been castigated by their opponents for encouraging laissez-faire in economics and anarchy in politics, the alleged natural consequences of adopting an "atomistic" view of social life. Indeed, the threat of appropriate social consequences seems to have been regarded by some as a reason for accepting one or the other of these methodological positions. F. A. Hayek and K. R. Popper are well-known champions of the principle of methodological individualism as a bulwark against the supposed horrors of the "planned society"—or at any rate, against anything worse than "piecemeal social engineering."
It is not, in fact, entirely accurate to say that the methodological, metaphysical, and political doctrines have invariably gone together. Thomas Hobbes, for example, was in effect a methodological individualist who advocated something close to political absolutism; and Maurice Mandelbaum, as will appear below, is a contemporary methodological holist who would certainly repudiate "invisible powers" and "impersonal entities." But political or ethical argument has, in any case, a dubious place in an examination of holism and individualism as methodological prescriptions for social and historical research. Even if metaphysical questions cannot ultimately be ignored, it is worthwhile, at least at the outset, to try to consider the contending methodological doctrines in their own terms. The discussion that follows makes no attempt to trace the considerable history of the problem in Western philosophy; rather, it is a report on what some contemporary philosophers have said by way of exposition and defense of the two positions. Since it has generally been the individualists who have taken the initiative in controversy, it will be convenient to set forth their position first.
J. W. N. Watkins, one of the most prominent recent advocates of methodological individualism, has presented it as primarily a theory of sociological or historical explanation. In his "Ideal Types and Historical Explanation," Watkins stated its requirements thus: "Social processes and events should be explained by being deduced from (a ) principles governing the behaviour of the participating individuals and (b ) descriptions of their situations." The elaboration of criteria for acceptable explanation is, of course, an activity characteristic of philosophers. What has most often concerned them, however, has been the formal or structural features of explanation, that is, the logical relation that must hold between an explanans and explanandum. Watkins's criterion, by contrast, is a material one. It makes a stipulation about the content of a social or historical explanans, holding that it must be "psychological," at least in the sense of being, in Watkins's words, about "the situations, dispositions and beliefs of individuals."
In formulating their material requirement, individualists often have in mind successful patterns of explanation in other branches of science. According to Watkins, the principle of methodological individualism is a correlate of the principle of mechanism in physics, which held triumphant sway from the seventeenth to the nineteenth centuries. An especially prestigious example of the application of the mechanistic principle is the explanation of the solar system by reference to Isaac Newton's laws and the positions, masses, and momenta of its component "individuals." Another example, often cited, is the explanation of the macro properties of a gas—its temperature, for example—as a resultant of the micro properties of its molecules. The best illustration of the same explanatory procedure in social science is afforded by classical economics, which regards macro states of the market as resultants of the dispositions and consequent activities of individual producers and consumers. There are differences (some will be discussed later) between the way particles in a mechanistic system are linked with what they explain and the way psychological facts about individuals are linked with social events. Methodological individualists, however, regard the likenesses as more instructive than the differences.
The rival thesis of methodological holism is that explanations in history and social science may (some would say "must") employ holistic societal laws or dispositions. Social dispositions are envisaged as being holistic, not only in the sense of being macroscopic relative to individual behavior but as being irreducibly so. Except in extreme versions of the theory (usually framed by opponents for polemical purposes), psychological elements are not actually excluded from a social explanans; they are merely regarded as insufficient. Thus, in their most usual form, the two methodological doctrines are not contraries but contradictories.
In elaborating their position, holists often match paradigm cases with the individualists. In economics, for example, they point to the Keynesian theory, which relates such variables as national income and savings, as showing the need to supplement the classical approach with a macroscopic one. In physics they note the decline of mechanism with the development of wave and field notions. And methodological holists do not limit their claims to cases in which social phenomena are explained by other societal factors. The explanation of individual actions themselves, they insist, may often have to be given partly in societal terms, employing laws that link individual behavior with types of social conditions. They deny, however, that this commits them either to organicism or to historicism. For sui generis societal laws can be of various logical types. They need not be organic, in the sense of relating the parts of the social system in a way that makes society self-regulating or self-maintaining, nor need they be developmental. There is thus no necessary connection between methodological holism and the dismal conclusion that men are caught up in some inexorable process that possesses something like a life of its own.
Refinements of Individualism
The basic response of methodological holism to the individualist claim is that the procedures of history and social science are in fact largely holistic, and that attempts to apply the principle of individualism do not work. The theory of the social sciences should accept the consequences. To methodological individualists, on the other hand, failures of application simply indicate a need for further analysis and research. Yet the discrepancy between fact and theory has induced individualists to make a few concessions, which are often represented as "refinements" or "clarifications" of the original thesis. A brief look at four of these may help to sharpen the issue.
levels of explanation
Individualists generally concede, first, that macro explanations may sometimes be both true and informative. The temperature of a gas, for example, may be explained by referring to a heat source that was applied to it, or to such simultaneous macro conditions as its volume and pressure; the outbreak of a revolution may be similarly explained by referring to economic or social trends in the society as a whole. According to Watkins, all the methodological individualist claims is that until we manage to reduce such explanations to terms of the molecular theory of gases or the psychology of individuals, we fail to achieve a full understanding of what has occurred. Thus, what the individualist seems to offer is not a criterion of being an explanation at all (for this, the satisfaction of formal criteria may be enough), but of being an ultimately satisfactory one. Yet the acceptability of "half-way explanations" (to use Watkins's term) is said to depend on the possibility of eventually reducing them to "rock-bottom explanations." The concession, in other words, is only with regard to "practice"; nothing is yielded at the level of "principle."
A second refinement arises out of the suspicion that what is actually possible in social science, even "in principle," is seldom an explanation in terms of the dispositions of the specific individuals involved. We might explain the rise in a stock's value, for example, by pointing out that the individual dispositions that most stockholders may be presumed to share lead them to be willing to pay a higher price under the circumstances; but we could hardly hope to ground our conclusion in knowledge of the detailed motives and beliefs of all the individuals involved. Methodological individualists consequently limit their prescription, even for "rock-bottom explanations," to typical dispositions of anonymous individuals. Such explanations, they will point out, still follow the model of mechanistic physics, in which information about specific particles is not required. Unlike physical particles, it cannot, of course, be presumed that human beings are all alike, or even that they are similar in all respects relevant to the social resultant that is being studied. This is particularly the case in historical inquiry, with its concern for unique rather than recurring circumstances and events. Thus, it will often be impossible to give adequate historical explanations without taking at least some named individuals into account. Even in the field of history, however, there is considerable scope for the anonymous.
Advocates of individualism often emphasize that if explanation need not be in terms of the actions and dispositions of specific human beings, still less need it show that social phenomena are brought about deliberately, or even knowingly, by individuals. Methodological individualists do not question the contention, constantly reiterated by holists, that social phenomena are largely the unintended results of the behavior of hosts of interacting human beings. The individualist principle is thus to be distinguished from what K. R. Popper, in The Open Society and Its Enemies (London, 1945), has called the "conspiracy theory of society": the view that for every social effect there is a manipulator (hero or villain) to be found. Not that individualists doubt that public affairs are controllable through the knowledgeable intervention of people; they hold, rather, that even when events are not so controlled, they can be explained individualistically. The individualist principle is also to be distinguished from a second doctrine with which Watkins felt it is sometimes confused, namely, the view that social phenomena "reflect" the dispositions of component individuals. Social characteristics are often, in fact, quite different qualitatively from the characteristics of the individuals referred to in explaining them. But there is likewise no qualitative similarity between the thermodynamic properties of a gas and the mechanical properties of its elements.
Some individualists are willing to make a concession that leads to still a fourth refinement of the original doctrine. They allow that there are some social phenomena, at least, that may not be open to individualistic explanation at all, although they usually add that these exceptions are not very important for the theory of the social sciences—certainly not important enough to justify the acceptance of methodological holism as a general principle for explanation in these fields. The exceptions fall into two classes. The first contains phenomena that can be treated only statistically. The second consists of occasional instances of what may be genuinely organic "social" behavior: Watkins mentioned the physical union of mating couples, the ecstatic singing of revival meetings, the rioting of panicking crowds. But, individualists argue, we cannot extrapolate from such cases to the nature of "higher-grade" forms of social organization. The latter, even when unplanned, are related by "ideas" and involve people widely separated in space and time.
Arguments for Methodological Individualism
Clarified and refined, then, the principle of methodological individualism asserts that ultimate or final explanation of the more significant social phenomena must be given in terms of at least typical dispositions (including beliefs, attitudes, and volitions) of anonymous individuals involved. Individualists often seem to present this principle as self-evident. Yet arguments for it have been offered, among the most characteristic in contemporary literature being the five that are considered below. No separate presentation of the case for methodological holism will be given, since holists are generally content to offer rebuttals of what their opponents claim.
One common argument appeals directly to ontological considerations. According to Watkins, "the ontological basis of methodological individualism is the assumption that society … really consists only of people." Social "things" may even be said to be "created" by individuals, by their attitudes as well as by their actions. "Remove the attitudes of food officials, shop-keepers, housewives, etc., towards ration books," Watkins observed, "and they shrivel into bits of cardboard." To a methodological individualist it seems paradoxical to suggest that social objects, thus constituted, could be explained other than individualistically. To try to explain individual actions in social terms seems to involve referring what really exists to a mere "construction." Yet although ontological individualism offers the methodological doctrine a "basis," Watkins conceded that the former does not actually entail the latter. It might still be true that what is constituted by individual actions and attitudes is governed by autonomous social law, although the ontology of individualism makes this difficult to believe.
Today, few holists would argue directly from a corresponding ontological thesis, which would rest upon some such principle as "a whole is not equal to the sum of its parts," the social whole thus being conceived as free to operate in accordance with laws which hold true at its own "level of existence." Typical of objections to this are Ernest Nagel's observation, in The Structure of Science, that wholes are recognized in physical science, too, apparently without presenting special problems for individualistic explanation; and Popper's jibe, in The Poverty of Historicism, that the metaphysical principle of holism, although "trivially true," applies even to three apples on a plate. However, most methodological holists (for example, Maurice Mandelbaum in "Societal Facts") prefer to argue that although social phenomena can be said to be ontologically dependent upon the actions and attitudes of individuals, the two are not simply identical. They point out, too, that their doctrine does not commit them to claiming that societies could exist without people, this being an absurdity eschewed even by full-blooded ontological holists like G. W. F. Hegel. The frequent use, in this connection, of the epiphenomenalist account of the mind-brain relation to show what might be meant by ontological dependence without identity is rather unfortunate. For, whereas a mind with no brain may be conceivable, few, if any, methodological holists would allow that society was conceivable without individuals. Many methodological holists, in fact, profess complete ontological individualism. What they demand of individualists is a willingness to try to find out whether there are any irreducible societal laws.
This takes us within range of a second metaphysical argument. According to Watkins, it is a "metaphysical commonplace that social events are brought about by people." He interpreted this "commonplace" to imply that individual men and women "together with their material resources" are the "only moving agents," indeed the "only causal factors," in history. Social wholes, whether or not they can be said really to exist, cannot do anything; in particular, they cannot affect the behavior of the concrete human beings who constitute them. Methodological individualists therefore disagree with economists who regard long-term cyclical waves in economic activity as, in Watkins's words, "self-propelling, uncontrollable and inexplicable in terms of human activities." They similarly oppose historical materialism, which, in its more uncompromising forms, at any rate, asserts a one-way causal relation between certain social conditions (the economic substructure) and the thoughts and actions of those who live under them. It is a "central assumption of the individualist position," Watkins declared, that "no social tendency is somehow imposed on human beings 'from above' (or 'from below')." Actually, even the more modest (and more usual) thesis of "interaction" between the social and the individual spheres is often deemed unacceptable by individualists.
The usual response of the holists to this line of argument is to ridicule the implied denial of social conditioning—as if people were not born into social situations in the first place. The "real oddity" of methodological individualism, wrote Ernest Gellner in "Explanations in History," is that "it seems to preclude a priori the possibility of human dispositions being the dependent variable in an historical explanation—when in fact this is what they often or always are." An associated peculiarity is that it precludes "the possibility of causes … being a complex fact which is not describable in terms of the characteristics of its constituent parts alone—which again seems often to be the case." Individualists, of course, would regard this charge as a misunderstanding of their doctrine. They would hold that the social conditioning of individuals, although real, is simply their conditioning by other individuals, referred to compendiously by holistic terms. And they would accept this claim that causes may be complex facts as long as the complexity of the cause is regarded as "resultant" from individual actions in the way indicated by the ontological argument. (Some individualists, however, would find it less easy to counter the argument that to speak of causes as "moving agents" at all is tacitly to accept an "activity" view of causation that has been suspect since David Hume.)
Alan Donagan, in The Later Philosophy of R. G. Collingwood, provided a version of the individualist's causal argument that turns on a conception of human action made familiar by idealist philosophers. The only way men's actions can be explained, Donagan maintained, is through their "thoughts"; it is not men's actual situations which explain what they do, but their conception of the situations (although it may be necessary to refer to the actual situation in explaining a man's success or failure in translating his intentions into action). Thus, if physical causes, like climate, operate in history, they must operate indirectly; and the same is true of such social events and conditions as an economic depression or a military victory. Unless we are to challenge the common assumption that the causal relation is transitive, however, methodological holists may well feel that such considerations, even if they are acceptable in themselves, do little to establish Watkins's original contention. For to say that social causes require the mediation of individual thoughts and responses is not to establish the latter as the only "moving forces" in history. On the contrary, to cause individuals to cause is still to cause.
The theory of action thus indicated has a bearing on a third general argument that is sometimes used by methodological individualists. This argument develops Watkins's contention that even if we learned to describe, predict, and control social events and conditions holistically, we still could not properly claim to understand them without treating them as a collection of individual responses. For "understanding," Watkins seemed to insist, requires the explanation of what happened in terms of intelligible human dispositions. What he appears to have had in mind is the discerning of the participants' reasons for doing what they did, which allows us the intellectual satisfaction of seeing why they thought their responses were appropriate. As Gellner has pointed out, there is a dual thesis here: first, that social or historical explanation must be couched in terms of the dispositions of individual human beings; second, that these dispositions must be of a special kind. For those who would claim, on general philosophical grounds, that explanation by reference to an agent's reason or motive is logically different from subsuming an occurrence under a law (or even under a "disposition" properly so called), the present claim opens up the possibility of giving individualistic explanations of social phenomena without reference even to psychological laws.
Many methodological holists would agree that to accept the additional thesis would make their position quite untenable, for it might be claimed that "intelligible" dispositions could be sought at the level of social wholes only on the assumption either of an immanent group mind or of an external historical providence: in other words, methodological holism now would require ontological holism. Something just short of this nevertheless sometimes appears to be entertained. Thus Morris Ginsberg, in On the Diversity of Morals (London, 1956), while denying for ontological reasons that society is itself a mind, conceded that it has a "mental organization" or "inner side" that is not identical with the mentality of any of its component individuals. Most methodological holists, however, simply deny the necessity of the additional criterion of explanation. They would hold that the essential claim of methodological individualism could be achieved without reference to intelligible dispositions if appropriate psychological laws could be found. And they would similarly claim for their own position that subsumption under autonomous societal laws (if such laws could be found) would yield understanding in the only sense significant to "scientific" inquiry.
A fourth argument makes the even more basic epistemological claim that whereas we can observe human individuals, we cannot similarly obtain knowledge of the macro features of social groups. As Watkins put it: "The social scientist and historian have no 'direct access' to the overall structure and behaviour of a system of interacting individuals (in the sense that a chemist does have 'direct access' to such overall properties of a gas as its volume and pressure and temperature, which he can measure and relate without any knowledge of gas molecules)." Since reliable knowledge of the dispositions and situations of individuals is readily available, Watkins continued, and since these individuals constitute the group, "a theoretical understanding of an abstract social structure should be derived from more empirical beliefs about concrete individuals." How else can what is said about social wholes be verified? Such an appeal to "hardheaded" empiricism is a shrewd blow. For the contenders here are (with a caveat, perhaps, about history) two theories of "scientific" inquiry. It would be odd if they did not both claim to be "empiricist."
Faced with this argument, many methodological holists insist that some social phenomena, at least—for example, parades, trials at law, battles—are directly observable. It is true that no one will notice such things if he lacks certain interpretative ideas or concepts. But this is not a peculiarity of social observation. Individual human actions themselves will not be "observed" unless we are able to discern the intentions and motives of the actors; and it may be questioned whether these can be known "directly" in a sense in which group phenomena cannot. The epistemological criterion of the individualists, in other words, either allows some social phenomena to be counted as observable or excludes the most interesting individual phenomena. Many holists nevertheless concede that the social wholes of most significance for history and social science cannot, in any ordinary sense, be directly observed. They reject instead the implication that this puts them at odds with accepted procedures in natural science. Not all physical theorizing proceeds, as in chemistry, from observables to what explains them: astronomy, for example, "constructs" its wholes as surely as sociology does. They admit, too, that assertions about societies must be verified by discovery of what individuals do. But they deny that this undermines their claim to possess knowledge, not just of the individuals but of the social wholes of which they are elements.
The fifth argument for methodological individualism, although it obviously has some affinity with both ontological and epistemological ones, is presented as a point of logic or semantics. As L. J. Goldstein stated it, in his "The Two Theses of Methodological Individualism," individualists require, as a condition of their being meaningful, "that all of the concepts used in social science theory be exhaustively analyzable in terms of the interests, activities, volitions and so forth of individual human beings." If this condition were met, the apparent holism of explanations employing societal laws would be tolerable, because it would be eliminable "by translation." Watkins himself denied that this conceptual thesis actually belongs to the central position of methodological individualism. The latter, he maintained, is a theory of explanation, not of concept-formation or description. Yet in arguing for the explanatory thesis, he characteristically slipped into the conceptual one. He maintained, for example, that to an individualist, the statement "The Jewish race is cohesive," if it is to be empirically meaningful, must mean such things as "Jews usually marry Jews"—a statement about anonymous individuals. And he commended Max Weber for insisting that the only way to make the meaning of social terms precise is to define them individualistically—as if such concepts appear holistic only when they remain vague or undefined.
Methodological holists have denied both that such analysis, definition, or translation is possible and that the conceptual thesis, even if sound, would establish the explanatory one. In arguing for the first of these positions in "Methodological Individualisms: Definition and Reduction," Brodbeck allowed that there are no insuperable problems for conceptual individualism so long as we are dealing with group concepts that are basically statistical—as in "He got his votes from the poor." Less straightforwardly statistical locutions like "Boom follows slump" might present problems for individualistic translation only because their implied statistical reference is vague. The real difficulty arises over such terms as "renaissance" or "the government." There seems to be no finite list of individual actions and attitudes that would count as their exemplifications; yet the problem does not seem to be one simply of vagueness. Appropriate exemplifications, furthermore, seem to vary from culture to culture, without our being able to say that the relevant terms are ambiguous—which suggests, perhaps, an evaluative element in their meanings. Terms within this range of social description appear to be logically holistic. As Mandelbaum has observed, the problem here is analogous to one which phenomenalists have paid great attention to in contemporary theory of perception. The full meaning of a material object statement, it is generally agreed, cannot be given by any finite set of sensation reports alone; we always need reference to "conditions of appearance," which are stated in the material object language. Attempts to translate societal statements into psychological terms founder on the similar need to specify the social conditions under which an action must be performed in order for it to count as an exemplification.
Mandelbaum himself actually wavered on this point. He conceded that partial translatability, at least, is required; otherwise there would be no way of verifying societal statements (an echo of the epistemological argument). He was sufficiently moved, too, by metaphysical considerations (both ontological and causal) to believe that full translation may be possible "in principle," even if this cannot be made the basis for a "practical" methodology. Many methodological holists have claimed, however, that full translatability would still not warrant the acceptance of the individualist thesis as it is most commonly understood, namely, that explanations should be (or should be capable of being) limited to psychological terms, with psychological laws as the only permissible kind of connecting generalizations.
Consideration of the claim that full translatability would not warrant acceptance of the individualist thesis calls attention to three important obscurities that have dogged much contemporary discussion. The first concerns the sense of "explain" in which a methodological individualist asserts that ultimately all explanations must be individualistic. The conceptual claim, it should be noted, has been stated not only as an argument for the explanatory one but also as though it were itself a theory of explanation—and so it is, in one important sense of the term. According to Watkins, every complex social situation or event is "the result of a particular configuration of individuals, their dispositions, beliefs, and physical resources and environments." This is often, and plausibly, read as meaning that we must be able to explain large-scale social phenomena as configurations or resultants of individual ones. But to explain something as something else is to explicate its nature: It is to explain it in the sense of showing what it really is. Although this kind of explanation is common in history and social science, however, it does not seem to be what methodological individualists have usually had in mind when advancing their explanatory thesis. What they envisage is explanation in the sense of showing how or why something came to be what it is: explanation that goes on to give causes, for example. It remains to be seen whether the conceptual thesis has any bearing on individualist claims about such explanations.
It may be objected that this ambiguity underlying the notion of being a "resultant" need not destroy the support given by the conceptual thesis to the full explanatory thesis, since what is specified in the productive sense may itself be treated, in its turn, as a resultant in the constitutive sense, thus achieving full "reduction." But this directs attention to a second obscurity in the individualist thesis, the question whether a "why" or "how" explanation in which all societal terms were replaced by psychological ones would involve the reduction of societal to psychological laws. Holists such as Nagel and Brodbeck have contended that it would not, at any rate, achieve reduction in the sense most familiar to the philosophy of science. For the derivability of macro laws from micro laws, no matter what the field of inquiry, is at least partly an empirical matter. Even in such exemplary cases as the reduction of chemistry to physics, they have pointed out, composition laws, which specify the way individual behavior changes as groups increase in size, must be added to the ordinary laws of the micro discipline; and these, however "self-evident" they often seem, have an empirical status. If laws of individual psychology are to be related "reductively" to laws of group phenomena, empirical composition laws would similarly have to be found. The reduction could never be just a matter of definition.
Individualists may complain that this involves too restricted an interpretation of their demand that sociological and historical explanations be reduced to "psychological" terms. Thus, when Mandelbaum, a methodological holist, attacked the conceptual thesis on the assumption that no statement will be counted as psychological if it employs any societal term, his argument was rebutted by Donagan, an individualist, on the ground that hardly any human dispositions would be psychological on such a criterion. Watkins made it clear that, for him, a psychological disposition is simply one which specifies a way of acting and thinking that is open to an individual. Thus, he argued against an anthropological holist that marriage rules are widespread dispositions of anonymous individuals in a society to behave in a certain way, and against a historical holist that the "Calvinistic outlook" of seventeenth-century Huguenot traders was similarly individualistic. Behind the uncertainty about what is to count as "psychological," there in fact appear to lie two different interpretations of the conceptual thesis itself. The first, which imposes a limitation upon the ways of behaving that may be cited in a "rock-bottom" explanans, is the translatability thesis. The second, to which many methodological individualists appear to retreat under pressure, is the much weaker demand that an acceptable explanation employ concepts which can be attributed to an individual, or jointly to a group of them.
Willingness to move in the latter direction suggests that contemporary methodological individualists and holists are not really as far apart as they often seem. In spite of their insistence that what they put forward is a methodological doctrine which is merely supported by metaphysical considerations, it seems clear that what interests methodological individualists most is the related ontological claim that human beings are the "ultimate constituents" of the social world. By contrast, although methodological holists find themselves continually under fire for allegedly flirting with dangerous metaphysical notions, it appears that what they are most concerned to uphold is the logical respectability of using holistic collective concepts and macroscopic laws, if need be. As was indicated, many methodological holists protest their allegiance to ontological individualism, and this appears to be a perfectly coherent claim. Some would even accord the corresponding methodological principle of individualism "regulative status" as formulating an ideal to be striven for. What they resist is the conclusion a priori that we can realize the ideal, and the associated temptation to refuse anything less.
Brodbeck, May. "Methodological Individualisms: Definition and Reduction." Philosophy of Science 25 (1) (1958): 1–22.
Danto, A. C. "Methodological Individualism and Methodological Socialism." Filosofia 13 (1) (1962): 3–24.
Donagan, Alan. The Later Philosophy of R. G. Collingwood. Oxford: Clarendon Press, 1962.
Gellner, Ernest. "Explanations in History." PAS, Supp., 30 (1956): 157–176. Reprinted under the title "Holism versus Individualism in History and Sociology," in Theories of History, edited by Patrick Gardiner, 489–503. Glencoe, IL: Free Press, 1959.
Goldstein, Leon J. "The Two Theses of Methodological Individualism." British Journal for the Philosophy of Science 9 (33) (1958): 1–11.
Hayek, F. A. The Counter-Revolution of Science. Glencoe, IL: Free Press, 1952.
Mandelbaum, Maurice. "Societal Facts." British Journal of Sociology 6 (4) (1955): 305–317. Reprinted in Theories of History, edited by Patrick Gardiner, 476–488. Glencoe, IL: Free Press, 1959.
Mandelbaum, Maurice. "Societal Laws." British Journal for the Philosophy of Science 8 (31) (1957): 211–224. Has full bibliographical reference to much of the current literature.
Nagel, Ernest. The Structure of Science, 336–397, 536–546. New York: Harcourt Brace, 1961. Like Donagan, he treats the problem of holism versus individualism in the context of a broader discussion of science.
Popper, K. R. The Poverty of Historicism. London: Routledge, 1957.
Watkins, J. W. N. "Historical Explanation in the Social Sciences." British Journal for the Philosophy of Science 8 (30) (1957): 104–117. Reprinted in Theories of History, edited by Patrick Gardiner, 503–514. Glencoe, IL: Free Press, 1959.
Watkins, J. W. N. "Ideal Types and Historical Explanations." In Readings in the Philosophy of Science, edited by Herbert Feigl and May Brodbeck, 723–743. New York, 1953.
W. H. Dray (1967)