A number of social theorists have contended that the essence of historical analysis is the employment of ideal types to comprehend past goings-on. But, while acknowledging that the study of history through ideal types can yield genuine insight, we may still ask if it represents the full emancipation of historical understanding from other modes of conceiving the past. This paper follows Michael Oakeshott’s work on the philosophy of history in arguing that explaining the historical past by means of ideal types, even though offering a coherent and fruitful enterprise, nevertheless falls short of fully embodying the characteristics that differentiate historical understanding.
This dispute involves fundamental issues regarding the nature of the social sciences. Furthermore, I suggest that it is of more than purely theoretical interest, in that social theorists who have accepted the Weberian view of historical inquiry will tend to be unaware of the defects inherent in all attempts to capture the nature of past events in a net of generalizations, and thus may be lead to ignore the inherent partiality of the understanding of historical happenings provided by ideal typification.
Keywords: philosophy of history, ideal types, modal knowledge, Oakeshott, Weber.
A number of significant social scientists, including Wilhelm Dilthey, Max Weber, Ludwig von Mises, Alfred Schutz, and Fritz Machlup, have contended that the essence of historical analysis is the employment of theoretically constructed ideal types to comprehend past goings-on. But, while acknowledging that the study of history through the lens of ideal types can yield genuine insight, we may still ask if it represents the full emancipation of historical understanding from other modes of conceiving the human past, or whether there is, perhaps, an approach that must be regarded as its superior at exemplifying the unique character of history. In this paper I contend that Michael Oakeshott’s work on the philosophy of history convincingly argues that explaining the historical past by means of ideal types, even though offering a coherent and fruitful enterprise, nevertheless falls short of fully embodying the characteristics that differentiate history from other modes of knowledge.
This dispute is worth our attention because it involves fundamental issues regarding the nature of the social sciences. Furthermore, I suggest that it is of more than purely theoretical interest, in that social theorists who have accepted the Weberian view of historical inquiry will tend to be unaware of the defects inherent in all attempts to capture the complex and individual nature of past events in an abstract net of generalizations, and thus may be led to ignore the inherent partiality of the understanding of historical happenings provided by ideal typification.
Weber’s conception of the character of historical research is neatly summarized (and also endorsed) by Mises, who writes, “the difference between sociology and history is considered as only one of degree. In both, the object of cognition is identical. Both make use of the same logical method [i.e., ideal types] of forming concepts. They are different merely in the extent of their proximity to reality, their fullness of content, and the purity of their ideal-typical constructions” (Mises, 2002, http://www.mises.org/epofe/c2sec3.asp).
One of Weber’s most explicit statements of this view appears in his essay, “The Logic of the Cultural Sciences,” where he says: “Even the first steep toward an historical judgment is thus—this is to be emphasized—a process of abstraction…. Even this first step transforms the given ‘reality’ into a “mental construct” [i.e., an ideal type] in order to make it into historical fact” (1949, p. 173).
Schutz, an admirer of Weber and one of the foremost theorists of the ideal-type method, held that such abstractions are not found only in the constructs built by the social scientist, but also are constitutive of the everyday social world (as did Weber: see (Weber, 1949: 177-180). Except for the case of the "pure" Thou-orientation that occurs in face-to-face encounters, all of our relationships with and understandings of other people are mediated by the use of ideal types of varying degrees of anonymity. (The concept that one way of classifying ideal types is according to how anonymous they are can be made clear with a simple example: we might, in order of decreasing anonymity, regard some individual as an instance of the ideal type "Englishman," "early-twentieth-century Englishman," "early-twentieth-century Londoner," "early-twentieth-century, working-class Londoner," and " early-twentieth-century London chimney sweep.") Since the historian, qua historian, is never in a Thou-relationship with the historical actors whose deeds he seeks to comprehend, Schutz's view implies that he must employ one or more ideal types in his undertaking. (See Schutz, [1932] 1967, especially pp. 176-250.) Roger Koppl contends that, for Schutz, what differentiates history from a more theoretical social science such as sociology is only the degree of anonymity of the ideal types it employs: "What we call either 'history' or 'applied economics' entails the use of relatively concrete ideal types. 'Theory' uses more anonymous types" (1998 COMPLETE).
Forwarding a similar view, Mises writes: "Although unique and unrepeatable, historical events have one common feature: they are human action…. What counts for history is always the meaning of the men concerned: the meaning that they attach to the state of affairs they want to alter, the meaning they attach to their actions, and the meaning they attach to the effects produced by the actions" (Mises, 1966: 59).
Since the subject matter of history is composed of these human meanings, Mises argues that the historian cannot avoid employing some conceptual framework that enables him to abstract categories of meaning from the myriad of specific meanings that have composed the circumstances and actions of historical individuals: "The aspect from which history arranges and assorts the infinite multiplicity of events is their meaning. The only principle which it applies for the systemization of its objects--men, ideas, institutions, social entities, and artifacts-- is meaning affinity. According to meaning affinity it arranges the elements into ideal types… Ideal types are specific notions employed in historical research and in the representation of its results" (Mises, 1966: 59).
Finally, here is Machlup in the same vein: “Analysis of human action with aid of constructs isolating, idealizing, and exaggerating some human trait or function is needed both for historical and theoretical investigation. That is to say, the ‘breaking up’ of ‘whole man’ is needed for understanding the actions of historical persons…” (Machlup, 1978: 268).
Having presented above a brief, but hopefully sufficient, summary of the ideal-type theory of history, I want to differentiate a trivial sense in which the proposition that historians necessarily employ such types is obviously true from a stronger thesis that I understand the afore-mentioned thinkers to be presenting. Historians always use human languages to convey their findings, and, as such, they must continually employ words that denote general classes of objects, events, or activities. Those words, in the Weber-Schutz view of ideal types, might sensibly be regarded as referencing very fundamental ideal types. No practical language can do without such generalized denotators. Any “language” in which each signifier stood for only one, specific object or event in the world would be next-to-useless for communication, as there would be no way, short of a speaker bringing his audience into the immediate presence of the aspect of reality he had “named” and then pointing at it, by which his audience could hope to grasp his intended meaning (See O’Neill, 1996: 40, and Joseph, 2002:19).
The very nature of human language makes it impossible for the historian to compose a sensible narrative about the past that is not replete with generic terms. If an historian writes, “A messenger ran from Marathon to Athens to carry news of the Athenian victory to the city,” she is employing general concepts like “messenger,” “running,” “news,” and “victory.”
However, our “ideal-type historians” clearly are making a less trivial contention. An historian who undertakes an analysis of the High Middle Ages using concepts like “the nature of serfdom,” “the role of the peasantry,” “monasticism,” “feudal obligations,” and “the rise of the town,” is invoking a level of abstraction going far beyond that needed merely to speak intelligibly. The question at hand is whether constructing such higher-level ideal types is the essential method of historical understanding.
An Alternative View of Historical Understanding
So is the analysis of past events in terms of their conformity to various ideal types the ultimate method of history? Michael Oakeshott, who, among his varied intellectual achievements, is regarded as one of the more important philosophers of history of the twentieth century, rejected that view. He did not dismiss the effort to understand the past with the aid of ideal types as incoherent or fruitless, but argued that it could not achieve a fully historical account of the situations it sought to explain. Furthermore, he suggested that a method for achieving such an account does exist.
While I find the argument Oakeshott offers against regarding ideal typification as the fundamental tool of historiography to be convincing as it stands, I believe it is worthwhile to review his case here for three reasons:
1) Oakeshott presents his argument in a dense and, at least to those unversed in it, rather obscure prose style. As Paul Franco, author of a recent introduction to Oakeshott's thought, puts it, "Reading the late Oakeshott, like reading the late Henry James, can be a vertiginous experience" (2004, p. 142). I suspect that a satisfactory understanding of some of his points requires of the reader a more than passing familiarity with much of his later work. So it seems to me that an exegesis of his case, employing a less individualized vocabulary, could be of use to anyone who is interested in historical methodology but is not prepared to undertake an extensive study of Oakeshott's mature thought.
2) While it is clear to me that Oakeshott is addressing ideal-type methodology in the passages I cite below, he neither explicitly says so, nor does he cite the theorists of history whose methodology he is critiquing. (The paucity of references throughout Oakeshott's works is a general problem plaguing any scholar attempting to locate their place in the broader intellectual currents of his time.) As a result, a search of the literature on ideal types is not likely to even turn up Oakeshott's analysis. I hope that an exposition of Oakeshott's argument that is more transparent in regards to what theory of history is being examined will aid others doing research in the area.
3) Although, as I mentioned above, I consider Oakeshott's critique of ideal-type history to be adequate to his purpose, it is a rather terse consideration of that methodology, appearing in a book addressing many other facets of historical understanding. Therefore, an expanded version of it, particularly one that fleshes out the bare bones presented by Oakeshott with some concrete examples, may serve to make his argument more accessible.
The case I present here has been anticipated, to a degree, and in a specifically Austrian context, in Kaplan (1987). However, while the focus of Kaplan’s analysis of the ideal-type method is on demonstrating that it is incapable of arriving at general sociological laws, this paper asserts that it does not even represent a full realization of the unique character of historical inquiry.
So, let us examine Oakeshott’s thesis in some detail. He launches his consideration of whether ideal typification is the primary method of history by acknowledging that it is an historical method: "A past composed of carefully anatomized situations of various magnitudes, durations and constitutions, themselves composed of mutually and conceptually related occurrences, is certainly a past which has been endowed with a certain level of historical intelligibility…. [It] cannot be denied the character of an historical enquiry" (Oakeshott, 1999: 65).
The term "ideal type" does not appear in the passage just quoted -- indeed, as I mentioned above, Oakeshott never uses it anywhere in the argument under consideration here. Therefore, it may seem unclear that he even is referring to ideal types. However, he goes on to offer some examples of the approach he is contemplating that, I suggest, demonstrate that he is doing just that. The type of historical inquiry he is considering is exhibited, he declares, by "an historian… who spells out the character of Elizabethan Puritanism or of a doctrine identified as 'civic humanism,' who unfolds 'the structure of English politics on the accession of George III,' or who (like Fernand Braudel) specifies the 'energy resources' of Europe in the late eighteenth century…" (Oakeshott, 1999: 65-66). Those are all clearly examples of historical ideal types a la Weber.
While, as noted above, Oakeshott acknowledges that this approach to the past is genuinely historical—unlike, for example, the legendary past or the didactic past, which is mined as a source of stories providing guidance to those facing practical choices in the present—he contends that, nevertheless, it suffers from only partially realizing the ideal character of distinctively historical inquiry: "But although it has been called the most sophisticated understanding of the past, it is, I think, an unstable level of historical understanding. It recognizes (or half-recognizes) what it cannot itself accommodate, and it cannot defend itself against being superseded by what is a genuine competitor, critical of it in its own terms, and thus capable of superseding it" (Oakeshott, 1999: 65).
Oakeshott holds "an historical past may be regarded as a passage of historical change" (Oakeshott, 1999: 121). But in any explanation of past events as instances of ideal types the consideration of change can play at best a supporting role. Because an ideal type is constructed by abstracting out patterns seen as common to a number of concrete historical situations, it focuses the theorist's attention on an unchanging constellation of properties present in each episode, and directs it away from the unique and contingent events that led to each exemplar, and also away from the particular and ever-shifting characteristics by which each instance of the type differs from the idealization. The historian engaged in ideal type analysis "purports to be anatomizing a bygone present situational identity in terms of its constituent occurrences. No doubt he recognizes himself to be concerned with a passage of time which contains genuine change; but his enquiry, centred upon the articulation of a situational identity, cannot properly accommodate this recognition" (1999, p. 65).
Furthermore, the very nature of an ideal type as a static constellation of intelligibly related patterns of social life renders it ill-suited for discovering the specific historical events that explain the appearance of each unique instance of the type at a particular time and in a particular place: "And further, an engagement to anatomize an historical situation, in specifying its duration, recognizes it as an emergence and admits its evanescence; but the enquiry is not concerned to abate the mystery of its appearance upon the scene, to investigate the mediation of its appearance or to trace the vicissitudes of its evanescence. It is concerned only with correctly inferring an intelligible structure composed of notionally contemporaneous mutually related constituent occurrences" (Oakeshott, 1999: 65-66).
Oakeshott sums up his critique as follows: "These, then, are what I take to be the historical defects of an enquiry concerned to infer from record a past composed of situational identities: transitory passages of human engagement represented as patterned situations composed of mutually related occurrences which come and go but are here halted and made to gyrate in a notional interval between coming and going" (Oakeshott, 1999: 66). He goes on to suggest the solution to the difficulties he has pointed out: "And the remedy for the shortcomings of this level of historical understanding is not, I think, in doubt. It lies in an enquiry designed to assemble a past, not of anatomized situational identities composed of mutually related occurrences, but of historical events and conjunctions of historical events" (Oakeshott, 1999: 67-68).
The critique of ideal type history presented above bears some non-superficial similarity to Onora O’Neill’s distinction between abstraction and idealization. For O’Neill:
“Abstraction, taken straightforwardly, is a matter of bracketing, but not of denying, predicates that are true of the matter under discussion. Abstraction in this strict sense is theoretically and practically unavoidable… Idealization is another matter: it can easily lead to falsehood. An assumption, and derivatively a theory, idealizes when it ascribes predicates—often seen as enhanced, ‘ideal’ predicates—that are false of the case in hand…” (O’Neill: 1996: 39-44)
However, the analogy is only partial. While the theoreticians of history certainly will use idealizations in O’Neill’s sense—see, for instance, the quote from Machlup above—I believe that Oakeshott’s critique also applies to the use of abstractions beyond the level that would be required for an ordinary, natural language description of events.
Thomas Kuhn, In The Structure of Scientific Revolutions, describes the history of science as it typically had been practiced up to the time of his writing (Kuhn, 1962) as an attempt to comb the past for evidence of what has been termed “the Whig view” of scientific progress, wherein science makes slow, steady progress from ignorance and error toward knowledge and truth:
“If science is the constellation of facts, theories, and methods collected in current texts, then… [s]cientific development becomes the piecemeal process by which these items have been added, singly or in combination, to the ever growing stockpile that constitutes scientific technique and knowledge. And history of science becomes the discipline that chronicles both these successive increments and the obstacles that have inhibited their accumulation. Concerned with scientific development, the historian then appears to have two main tasks. On the one hand, he must determine by what man and at what point in time each contemporary scientific fact, law, and theory was discovered or invented. On the other, he must describe and explain the congeries of error, myth, and superstition that have inhibited the more rapid accumulation of the constituents of the modern science text.” (Kuhn, 1962: 1-2)
What Kuhn is describing is a view of history based on the ideal type, “the steady progress of scientific knowledge.” (I have no evidence that Kuhn ever heard of the notion of “ideal types,” but that is irrelevant as to whether that is what he is -- perhaps unwittingly -- discussing. As I will attempt to demonstrate, he employs them despite his apparent unfamiliarity with the concept.) But Kuhn does not recommend replacing the history of science he criticizes with one arrived at via Oakeshott’s method of examining the concrete series of differences through which historical research might render more intelligible some sequence of events. Instead, he reconstructs the history of science through the lens of a small number of other ideal types: “paradigms,” “pre-paradigmatic science,” “normal science,” “novel discoveries,” and “paradigm shifts.”
That Kuhn’s notion of a paradigm falls with the category of ideal types can be seen by examining the examples of paradigms he lists: “’Ptolemaic astronomy’ (or ‘Copernican’), ‘Aristotelian dynamics’ (or ‘Newtonian’), ‘corpuscular optics’ (or ‘wave optics’), and so on” (Kuhn, 1962: 10). Such examples clearly parallel Oakeshott’s “the character of Elizabethan Puritanism,” “civic humanism,” or “the structure of English politics on the accession of George III” (cited above). And we can find the same sort of categories in, for example, Mises:
“No historical problem can be treated without the aid of ideal types. Even when the historian deals with an individual person or with a single event, he cannot avoid referring to ideal types. If he speaks of Napoleon, he must refer to such ideal types as commander, dictator, revolutionary leader; and if he deals with the French Revolution he must refer to ideal types such as revolution, disintegration of an established regime, anarchy” (1966, p. 60).
Kuhn recognizes that an ideal type must be abstracted from an historical situation: “Why is the concrete scientific achievement, as a locus of professional commitment, prior to the various laws, theories, and points of view that may be abstracted from it?” (Kuhn, 1962: 11). His adherence to the ideal type approach is further displayed in passages such as: “Examining selected discoveries in the rest of this section, we shall quickly find that they are not isolated events but extended episodes with a regularly recurrent structure” (Kuhn, 1962: 52, emphasis mine).
Kuhn sees that he must carefully reconstruct the historical evidence in order to fit it into his ideal types: “In a third area, however, the existence of significant parallels between the discoveries of oxygen and of X-rays is far less apparent” (Kuhn, 1962: 58). He even acknowledges that his ideal-type reconstruction will necessarily be incomplete: “Breakdown of the normal technical puzzle-solving activity is not, of course, the only ingredient of the astronomical crisis that faced Copernicus. An extended treatment would also discuss the social pressure for calendar reform… medieval criticism of Aristotle, the rise of Renaissance Neoplatonism…” However, while he sees the inadequacy of his method, he can only conceive of amending it by incorporating more ideal types – “the social pressure for calendar reform,” “medieval criticism of Aristotle,” and “Renaissance Neoplatonism” – into a more complete account.
Imre Lakatos notes this shortcoming in Kuhn’s “rational reconstruction” of the history of science. (By rational reconstruction, Lakatos clearly means the interpretation of history according to particular ideal types of how science adavances.) For example, he complains, “let us imagine that in spite of the objectively progressing astronomical research programmes, astronomers are suddenly gripped by a feeling of Kuhnian ‘crisis’; and they are all converted… to astrology…. All [the Kuhnian] sees is a’crisis’ followed by a mass conversion effect in the scientific community: an ordinary revolution. Nothing is left problematic and unexplained” (Lakatos, 1978: 135-136).
However, Lakatos himself is unable to conceive of an historical method superior to ideal type analysis, and so he can only suggest replacing Kuhn’s preferred ideal types with his own, the “progressive research programme.” Indeed, echoing Mises, Lakatos declares “History without some theoretical ‘bias’ is impossible” (Lakatos, 1978: 120, emphasis in original). But Lakatos offers no more justification for that claim than that he can detect a theoretical bias even in histories where they are “obscured by an eclectic variation of theories or by theoretical confusion” (Lakatos, 1978: 120). However, the fact that the histories Lakatos has examined all do show some theoretical bias is nothing like a demonstration that all history must do so. Even if he could show the impossibility of any real historian operating without such a bias – which he does not even attempt to do -- he still would not have demonstrated that ideal history would not be free of such bias, and that, as a corollary, actual history should try to eliminate it. (Analogously, a survey revealing that all works of history examined contain factual errors certainly would not indicate that making mistakes about facts is an intrinsic aspect of the historian’s research.)
Contrary to the claim Lakatos makes, there exist many works on the history of science that, while not pristinely exemplifying Oakeshott’s superior style of historical inquiry, nevertheless consist primarily of efforts to understand episodes from the past in terms of the specific conditions from which they arose. To cite just one example, Richard Westfall’s biography of Isaac Newton, Never at Rest, does not explain Newton’s achievements by invoking a Kuhnian scientific revolutionary or a Lakatosian founder of a research program, but as the results of Newton’s particular abilities and circumstances. I offer just one typical, but I think sufficient, example as an illustration:
“Newtown’s early notes on analysis investigated the relation of particular axes to the equation of a curve. He tried various transformations of axes in order to simplify equations, seeking a regularized procedure whereby one of the new axes would also be an axis of the curve. He was not yet beyond fairly simply mistakes. Like the early analysts, he did not at first comprehend the significance of negative roots. He drew the cubic parabola, x3 = a2y, as though it were it were symmetrical with respect to the y axis, and somewhat later he can find his diagram of Descartes’s folium, the curve x3 – axy + y3 = 0, to the first quadrant, tacitly assuming that its shape is identical in all four quadrants. He learned quickly, however. By October, he comprehended negative roots clearly enough to set down the rule that when the x axis is the diameter of a curve such that it bisects all the ordinates, then y cannot appear in the equation in odd powers because there is a negative root of equal absolute value corresponding to every positive one. Above all, he seized and made his own the central insight of analytic geometry. Already in September he proposed a problem to himself in the following terms: ‘Haveing ye nature of a crooked line expressed in Algebr: termes to find its axes, to determin it & describe it geometrically &c.’” (Westfall, 1980: 107)
As noted above, no historical analysis presented in a natural human language can avoid the use of absract terms, and the passage from Westfall cited above is no exception. Nevertheless, I contend that the difference in the level of abstraction displayed by the work of Kuhn and Lakatos, on the one hand, and Westfall, on the other, amounts to a difference in kind and not merely degree. Westfall does not seek to explain Newton’s advances in terms of highly generic concepts like “the revolutionary scientist,” “the paradigm shift,” or “a progressive research programme.” Instead, Newton’s progress is made more intelligible by examining the specific mathematical topics he addressed and the particular approaches that enabled him to achieve breakthroughs in those areas.
Ideal-type analysis yields, by its very nature, an understanding of an event in terms of its similarity to other events. It may very well succeed in rendering the past more intelligible. But, for Oakeshott and this author, it fails to produce a fully historical past. That can only arise from a contemplation of the past as a passage of differences whose continuity makes a subsequent event explicable. In his view, it is the unique task of history to comprehend the past not in terms of whatever regularities it may exhibit, what typical situations may be abstracted from it, or any general precepts it is held to illustrate, but as an assemblage of events each of which is unique and unrepeatable. Such a view does not disparage other approaches to understanding the past; it only proposes that they "do not mix with and cannot take the place of an historical understanding concerned with what was actually the case, there and then, in terms of situations composed entirely of mutually related occurrences inferred from the record" (Oakeshott, 1999: 65).
Franco, P. (2004) Michael Oakeshott: An Introduction, New Haven and London: Yale University Press.
Joseph, Sister Miriam (2002) The Trivium: The Liberal Arts of Grammar, Logic, and Rhetoric, ed. Marguerite McGlinn, Philadelphia: Paul Dry Books.
Kaplan, E.H. (1987) book review of Writing History: Essay on Epistemology, Review of Austrian Economics, Vol. 1, No. 1.
Koppl, R. (1998) "Alfred Schutz and F. A. Hayek as Misesian Methodologists," paper delivered to the Workshop on “Spontaneous Orders: Austrian Economics, Philosophy, and Aesthetics” sponsored by Center for Interdisciplinary Business Research in Economics and Organization Copenhagen of Comparative Literature of the University of Copenhagen, 3 November 1998, available at http://www.hayekcenter.org/hayek-l/koppl_seminar.html.
Kuhn, T.S. (1996) The Structure of Scientific Revolutions, 3rd edition, Chicago and London: University of Chicago Press.
Lakatos, I. (1978) The Methodology of Scientific Research Programmes: Philosophical Papers Volume I, Cambridge: Cambridge University Press.
Machlup, F. (1978) Methodology of Economics and Other Social Sciences, New York: Academic Press.
Mises, L. von (1966) Human Action: A Treatise on Economics, online edition published by Ludwig von Mises Institute, available at: http://www.mises.org/humanaction.asp.
-- (2002) The Epistemological Problems of Economics, online version Auburn, Alabama: Ludwig von Mises Institute, based on the 1976 print version, New York: New York University Press, http://www.mises.org/epofe.asp.
Oakeshott, M. (1999) On History and Other Essays, Indianapolis, Indiana: Liberty Fund.
O’Neill, O. (1996) Towards Justice and Virtue: A Constructive Account of Practical Reasoning, Cambridge: Cambridge University Press.
Schutz, A. ([1932] 1967) The Phenomenology of the Social World, trans. G. Walsh and F. Lehnert, Evanston, Illinois: Northwestern University Press.
Weber, M. (1949) The Methodology of the Social Sciences, New York: The Free Press.
Westfall, Richard S. (1980) Never at Rest: A Biography of Isaac Newton, Cambridge, New York, and Melbourne: Cambridge University Press.