International Studies in the Philosophy of Science; Oct99, Vol. 13 Issue 3, p275, 13p

What is historicism?

Abstract "Historicism" has become a ubiquitous and equivocal term. A classification is given here of five separate uses of the term currently in vogue, each provided with a unique qualifying adjective to help keep them distinct. I then offer a few objections to some of the more radical conclusions which have been drawn by proponents of a specific version of historicism, one associated with "postmodernism". The positions of Rorty and Putnam are contrasted as examples of strong and weak degrees of historicism, respectively.

Historicism is a label that gets applied to a confiisingly wide array of theses. One motivation for writing this paper was to try to get a clearer grasp on what people might mean when they use the term "historicism" and its cognates. For many people the term is likely to bring to mind the position criticized by Karl Popper in his books The Poverty of Historicism (1961) and The Open Society and its Enemies (1971). Popper's employment of the term has added to the confusion, since the position he identifies by that label is almost diametrically opposed to another usage already well established much in advance of his own interest in the subject. Popper used the term to refer to the thesis that an important object of the historical and social sciences is to make predictions about future developments in political and social trends. But "historicism" had already been associated with the ideas of the 18th-century thinkers Giambattista Vico (1668-1744) and Johann Herder (1744-1803). This earlier version of historicism arose in opposition to the Enlightenment ideal of an ahistorical and universal rationality, the inspiration for which was drawn from the growing successes of the natural sciences. Vico and Herder emphasized the unique and particular over the universal, and the unpredictable over the predictable. They were also less optimistic than their "modernist" counterparts about the prospects for a progressive improvement in human circumstances by a simple application of the principles of reason to the problems affecting individuals and nations. It is clear that an element of this same skepticism remains a strong motivating force in the current popularity of one particular form of historicism to be discussed here. A second motivation of this paper was to address this skepticism, and to offer reasons why some of the skeptical theses about science, objectivity, and knowledge in general that are typically associated with "postmodernist" trends are unfounded.

[276]

This section lays out a tentative classification of some commonly met varieties of historicist positions. The classification itself is not intended to capture the historical details of just who maintained what thesis and when, nor does it claim to be a proper partitioning of the possible versions of historicism into mutually exclusive and exhaustive categories. I do hope though that it will be helpful in allowing us to identify what people might mean when they use the term "historicism" and "historicist". It is also intended to display the varying strengths or degrees of radicalness of the different versions in question. Once that is done, I will look more closely at the type of support each version rests upon.

The first version maintains simply that:

(1) To be understood properly things must be considered within their historical contexts. This is just the methodological rule required for all good history of philosophy and, as Kuhn argued, for good philosophy of science. I take it that it is really quite non-contentious, relying only upon the observation that no thinker or writer operates within a vacuum. Philosophers and scientists are always trained in a tradition of some kind and are thinking in response to other traditions and thinkers. Because traditions and the problems with which they deal change throughout time, one must become familiar with them so as to better understand the thinker or thinkers in question. I propose to call this "mundane historicism".

The next thesis is also a methodological one and states that,

(2) History has its own methods which are distinct from those of the natural sciences. This thesis is associated with the 18th-century Italian Vico, the 19th-century German Dilthey, and in our own time with the Oxford philosopher R. G. Collmgwood. The claim here is that whereas the natural sciences are concerned with very general trends and repeatable phenomena, which can be explained by deduction from general laws, history deals with highly contingent, unrepeatable and particular events. Moreover, since the historian's object is to come to understand specifically human events, the natural procedure is to try to understand the intentions of the agents involved and to discern the significance of their actions for future events. This I will refer to as "methodological historicism".

Philosophy of history since about the middle of this century was consumed with a debate largely inspired by Hempel and his attempt to apply the deductive-nomological model of scientific explanation to the social sciences and history (cf. Hempel, 1959). According to that program explanations have the syntactic structure of argument forms (either deductively valid or probabilistic), knowledge of the premises of which would permit the prediction of the conclusion. Explanation is prediction (or predictability in hindsight). Collingwood, among others, was deeply critical of this proposal. History is just too complex and riddled with contingency for general laws to be of much utility, they argued. Moreover, explanation in history requires a sympathetic understanding of the motives of human agents; it is a hermeneutic activity. And while a foreknowledge of an agent's motives might permit prediction of the agent's behavior, it is a well-known adage that each particular act results in a multitude of unforeseeable reactions. History, as it has become popular to emphasize of late, is a complex and chaotic system.

Our third thesis is also methodological in nature, but will be more familiar to those acquainted with Popper's writings on the social sciences. It is the more metaphysical-sounding claim that,

[277]

(3) There are to be found in history general laws, rhythms, or patterns. And with these the social sciences can make predictions about the future. It should be immediately obvious that this version of historicism is at odds with the last version. For what I have called "methodological historicism" consists in the denial that one can make use of general laws for prediction-making in history. In fact Popper drew a fair amount of flak from historians and philosophers of history for using the label historicism to refer to a position so at odds with what had already been known by that name for a hundred years or more. Yet despite this I suspect that for the sake of having a convenient handle that will bring readily to mind all the important associations, no better title can be found than "Popperian historicism".

Of course, the position Popper had in mind was that of Hegel and Marx, whom he characterized, not altogether unjustly of course, as being obsessed with the idea of discerning in history general laws of development or "evolution". In fact, the 19th century was rife with this kind of speculation about historical evolution, Herbert Spencer being perhaps the most popular writer on the subject. A common element in such speculations is an organismic model of societies. Each nation, race or state is perceived as a distinct unit playing its part in an overall pattern of unfolding or development, just as individual cells differentiate themselves during embryological development for the sake of the resulting organism. For Hegel, all former nations were building up to the Prussian state in which the idea of freedom had achieved its final and fully developed form.

We might also mention the founder of positivism, Auguste Comte, in this regard. For his law of three stages of social-intellectual development is just such a kind of general historical pattern. According to the law of three stages, humankind has passed through three stages in its attempts to explain the natural world: first a theological stage, in which events are ascribed to the divine will of supernatural agents, second, a metaphysical stage which introduces unverifiable forces operating beneath the level of observable phenomena; and lastly the positive, or "scientific", stage in which explanations are properly restricted to the prediction and control of events by the observation of correlations and regularities among surface appearances.

It would not be entirely strange to place Kuhn's thesis of scientific revolutions in this category either. For what else does his dialectical theory of normal science – crisis-revolution-new normal science represent if not a general pattern of historical development. More will be said about the type of evolutionary theory it is most evocative of in the final section.

While all the historicist theses we have seen so far were (chiefly) methodological, the next two are less concerned with method than with epistemology and value theory. The fourth thesis maintains that:

(4) Standards of rationality are not fixed and eternal, but change over time. We might place Comte's theory of the three stages here also, even if he probably believed that the only "correct" standard of rationality was to be found in the final positivistic stage. But two more contemporary figures which I would argue fit this category are lan Hacking and Hilary Putnam. Hacking has drawn upon the research of the late historian of science A. C. Crombie (1994) to point out that over time what he calls "styles of reasoning" have changed. Some styles of reasoning, such as the symbology and numerology so important to astrology and alchemy, have gradually died out (among scientists at least). Others, such as statistical reasoning, are of much more recent invention and have truly revolutionized how we understand and deal with the world about us. The lessons to be

[278]

drawn from this extend beyond historiography to epistemology and our own self-understanding. For one it shows that we cannot rule out that currently acceptable forms of reasoning will not be replaced by new forms. Secondly, it gives the lie to much of the Enlightenment faith in a universal canon of rationality constraining all peoples at all times. Modem attempts to fulfill the Leibnizian project of developing a calculus ratiocinator, that is a precise and formal logic adequate for both deductive and inductive reasoning, have failed. This failure, one of the most notable including that of the logical positivist Rudolph Camap to formulate a precise logic of induction, has led Putnam to conclude that "There is no algorithm or mechanical procedure, no set of fixed ahistorical 'canons of scientific method,' which will lead us to the truth in every area, or in any ...," although he ends on the optimistic note that "there is the imperfect but necessary 'path' of struggling for and testing one's ideals in practice, while conceding to others the right to do the same" (Putnam, 1994, p. 195).¹ Due to the nature of its subject I will refer to this as the thesis of "epistemic historicism".

Other contemporary thinkers who might seem to belong to this category are Kuhn, Feyerabend, and Foucault (whose notion of episteme is the motivation for my choice of label for this particular version of historicism). It should be said that unlike Kuhn and Feyerabend neither Hacking nor Putnam has concluded that shifts in conceptions of rationality over time lead to an incommensurability between the successive conceptions. The same would seem to be true for Foucault as well.

Once we allow that reason is not some Platonic essence we can formalize into a rigorous system of axioms and rules of inference binding on all rational agents for all times, then some people are quick to point out that perhaps the "modernist" passion for logical thought is just that, a fancy for one species of behavior among many possible alternatives. In many ways this was what the Romantic reaction to modem science and the Enlightenment, by people like Herder, was all about. The prospects of a pluralism concerning rational behavior strikes many people as essentially more "democratic", and leads us to the next and final version of historicism.

The fifth version of historicism is I believe the most interesting, largely because it happens to be the most radical and currently the most popular among scholars of a wide array of disciplines. It insists that, just as there are no eternal standards of rationality,

(5) There are no absolute ahistorical values of any kind, rather all ideals are local and relative to a particular historical culture and period. An important question we will want to ask of this claim is whether it is supposed to be a purely descriptive one, based upon inspection of the historical record and anthropology, or a deep conceptual discovery revealed by a philosophical analysis of some kind.

An immediate corollary from this claim is that all that we are inclined to identify in our own culture as "knowledge" is really just a reflection of local culture, not an accurate and objective representation of independent reality. As historian Michael Stanford explains, historicism makes the claim that, "all social and cultural phenomena are historically determined, and therefore have to be understood in terms of their own age" (1998, p. 155). It is, of course, a position currently identified in general culture with "postmodernism" and with "social constructivism" in philosophy of science. This form of historicism is typically called "perspectivalist" or "relativist". Whereas epistemic historicism still allowed hope for a convergence to objective knowledge and styles of reasoning (cf. Hegel and Comte), this thesis draws the more radical conclusion that the very concepts of "truth", "objectivity", "reason", "scientific knowledge", etc. are merely social constructions favored by a particular culture at a particular time in history.

[279]

According to it we are cut off from all anchoring notions of objective reality and progress and are left to float free in the vast, dark expanses of historical space. Because it purports to apply to the entirety of a culture and its ideals I will call this "total historidsm".

This is the thesis that the historian Aileen Kelly has in mind (1998a, p. 407), when she describes historicism as "the denial of all suprahistorical absolutes and universal historical goals". The history of humankind therefore is not the story of an ultimate progress toward eternal truths. Rather is it the case that the ideals of the Enlightenment philosophy reflect merely contingent and historically local values of a particular culture bound by time and circumstance. "Total historicism", then, amounts to the relativiza-tion or localization of truth to particular historical epochs (cf. Iggers (1969) on "historicism" as ethical nihilism).

Postmodernism, as a project or attitude, is a rejection of the Enlightenment goal of arranging society in accordance with its ideal of objective rationality. What it proposes is a new, and supposedly "better", conception of human flourishing, one not so polluted with the ideology of liberal capitalism. That it is a predominantly political program is apparent, as it must be if its own claims about the essentially ideological basis of all systems of thought are to be followed through with consistency. It is worth noting too that this skeptical thesis of total historicism is also incompatible with Popper's version of historicism. Postmodernism as expressed by Lyotard is an "incredulity toward metanarratives", of which Marxism and liberalism are two prime examples.

Upon what basis of support do these various historicist theses stand? As with any claim there are two different sources of possible support. One is empirical or factual, the other is conceptual. We shall see that of the various historicist claims, some draw their support from appeals to history, while others rely upon conceptual arguments. The latter are chiefly arguments about the nature of language and its ability (or inability) to represent objects and events existing independently and externally to language.

"Mundane historicism" draws perhaps on both empirical and conceptual support. One may contrast the results of ahistorical studies with those which take the historical context of an idea or person into account, and show that the latter are to be preferred for greater accuracy, depth, or a number of other virtues. But one may also claim that the historically sensitive study is to be preferred on the basis of a plausibility argument, in the absence of any actual comparison of results. (They just intuitively seem better.)

"Methodological historicism" almost surely rests upon a conceptual argument, to the effect that given the distinct nature of their respective subject matters the social sciences and natural sciences must follow distinct methods. History, for instance, is surely not an experimental science; nor has it been unquestionably improved by the adoption of quantitative methods. This does not, however, rule out the marshaling of empirical evidence as well.

The search for general laws of history which lies at the heart of "Popperian historicism" has traditionally relied upon both a purportedly empirical inspection of the historical record and abstract arguments. Hegel's philosophy of history with its theory of a dialectical logic driving events onward toward the final goal of the absolute is one example of how both types of consideration can be combined.

"Epistemic historicism" relies largely upon historical (and anthropological) research showing how styles of reasoning and ideas of rationality have undergone a rise and fall in popularity and influence throughout the development of (chiefly European) civiliza-

[280]

tion. At least that appears to be the motivation for Backing's adoption of what I am calling epistemic historicism. Putnam, on the other hand, seems to be guided by further considerations arising from the failure of attempts to codify or to naturalize reason. The prime example of the failure to formalize reason is Camap's project of inductive logic. Putnam has also argued against the plausibility of reducing the epistemic to non-epistemic physicalist concepts for some time now.² Our notions of rationality, he has argued, are inextricably tied up with notions of value and of "human nourishing". And if this is correct then different values and conceptions of the "good life" will lead to different notions of rational conduct. For instance, game-theoretic attempts to illuminate rational decision-making enshrine the particular ideal of "utility maximization".

While very likely motivated by disillusionment with the outcome of events in the 20th century (for instance, the horror of two world wars, ethnic cleansing, threats to traditional social units imposed by global capitalism, the widening gulf between rich and poor), "total historicism" is often defended on the basis of arguments about the nature or essence of linguistic representation and/or linguistic practices purported to be central to the activities of science, philosophy, and history (and in its most extreme form, to the entire range of human activities–placed under the rubric of "discourse").

What are these arguments concerning the nature (or essence) of language? The best-known claims are perhaps those associated with the French poststructuralist philosopher Jacques Derrida, who insists that all language is self-referential, and consequently that there is "nothing outside of the text". Despite what the label itself may suggest, poststructuralism retains an important element of structuralist theory, namely Saussure's theory of language and signs. According to this structuralist account of language, all signs obtain their meaning or significance through a relationship of "differences", that is through being differentiated from other signs. The upshot of this is that no sign points beyond the linguistic structure to objects external to the structure, so there is no reference to or representation of the external world. In Saussure's words, "There are only differences". What ^(structuralism rejects is the additional structuralist claim that different sign systems share a common underlying structure which in turn allows for a kind of universal theorizing about disparate topics. Poststructuralists prefer to emphasize the relationships of power that sustain systems and practices. This penchant for pointing to the corrupting ideologies underlying many powerful and deeply entrenched practices and institutions is a common bond shared with postmodernist writers.

The target of my criticisms here will not be the total historicist claim that we alone are the source of all our values and ideas, and that they are therefore subject to change over time, but rather two separate claims that are often associated with this idea about the historicity of our values and ideas. The first is the claim that language is inherently self-referential, and so cannot refer to objects or things in the external world. The second is that all knowledge claims are tainted by ideology, and hence that there can be no objective knowledge.

How good is the poststructuralist argument against the capacity for language to point beyond itself to an external world of objects and events? Let us begin by noting that it will only be as convincing as the linguistic model upon which it is based. That model assumes that meaning as encapsulated in thought is entirely determined by the

[281]

speaker with no input from the external world of objects and events. Were this the only theory of language or signs, a study known as semiotics or semiology, we might be more inclined to accept it, at least provisionally while attempting to improve upon it. But as it happens it is not the only semiotic theory available. A much richer one, in my opinion, is that of the 19th-century scientist and logician, Charles Peirce. In Peirce's model signs can and do refer to external objects and events. I do not wish to go into a discussion of the quite intricate details of Peirce's semiotic theory here, but suffice it to say that it combines the kind of representational capacity we desire with some of the genuinely important insights of both structuralism and poststructuralism. One of these insights is that meanings are not static, eternal platonic objects established once and for all time. The meaning of a sign in Peirce's theory involves the form of conduct it inspires, and this is always open to change and improvement. Meaning too then acquires a history (and a future), and–what is the same thing–loses an essence. (Compare this to the Fregean theory of the "museum of meanings" so central to much of analytical philosophy.) To paraphrase Peirce, "Meaning is always virtual", and yet refers to an external world of publically observable objects and behaviors.

Were it in fact the case that the model of language shared by structuralism and poststructuralism was correct, two things it seems would follow. First, the situation that all of us would be in would be akin to that described by Putnam in his "brains in a vat" thought experiment. Recall that the scenario envisaged there is that we are all supposed to be disembodied brains sustained in vats and led to believe that our experiences of the world and of one another are genuine, when in fact they are the manipulations of an evil scientist who has hooked our brains up to very complex computers which only simulate these experiences. Although we would each seem to be talking with one another and referring to objects (trees and cats) in our shared external environment, in actuality our experiences would be entirely solipsistic, since our words would not refer to trees and cats and one another, but only to our brain images of trees, cats and other people. This it seems to me is indistinguishable from the scenario in which the structuralist theory of language places us. If we cannot refer to objects in a shared environment occupied by us all, if our words only refer to other words or mental images, then we are little different from brains in vats. And that seems absurd enough to suspect that the model of language from which it follows is inadequate.³

The second thing that follows if the poststructuralists are right, is that since they have managed somehow to obtain supposedly objective knowledge about the very essence of language, then the possibility is opened up that we can exploit this possibility further to obtain more objective knowledge. It does not really matter that the objective knowledge in question here is only "internal" to the structure. The point is that if objective knowledge is possible at all, if ideology can be overcome in this instance then it cannot be a universal impediment to knowledge in general, contrary to what many people have been claiming under the banners of poststructuralism and postmodernism. The poststructuralist must either give up the claim that ideology is a universal impediment to knowledge or that the structuralist model of language is correct. In either case the way is opened up for knowledge about the world.

Another error that seems widespread in postmodernist thinking is what I might call (with tongue in cheek) the Marshall McLuhan fallacy: this is to assume that the medium is the message. Some people have concluded, that is, that since all theorizing is "textual" the world itself must be a text. If the medium of knowledge is linguistic or social then so too must be the signal. A parody of this argument would go as follows: since all my experience of the world is mediated through flesh of some sort (eyes, ears, fingertips,

[282]

etc.), then the world itself must be nothing but flesh too. Alan Musgrave (1999) makes very persuasive objections to such lines of argument.

One of the chief problems with many postmodern critiques of science and knowledge in general is that they focus exclusively on the representational aspects of what are called knowledge practices. But as others have insisted before, Dewey at the start of this century with his criticisms of the "spectator theory of knowledge" during the heyday of logical empiricism, and more recently lan Hacking, the trick is to pay attention to experiment, to doing, to what Peirce called the "outer clash" of experience–and this consequently will show why talk of people with different conceptual schemes living in different "worlds" is so silly (or at least terribly one sided). For although we can say of a modem capitalist industrialist and a contemporary South American Indian living in the Amazon rainforest that they live in different worlds, no amount of such talk is going to protect either of them from the other should they choose to fling their respective weapons at one another. Words and beliefs may make available a plurality of conceptual worlds, but the world of causes is one.

Which brings us to some of the wilder claims of social constructivism. Let us concede to the social constructivist that theories are social constructs, and that even the very notion of a "fact" is as well, for without humans and social activities the notion of a fact makes little sense; and so it goes, for example, for the notion of temperature, the Fahrenheit and Celsius scales are both social constructs too–and which we choose to use is a convention. But once we have these things in place, it does not follow at all that on the scale we choose to use there is no fact of the matter about what the scale reads–and yes how to read a scale, etc. are all social conventions too. But there is a fact of the matter that the mercury is at one level rather than another. To repeat: the very idea of a fact may be a social construction, but what the facts are–once we have our language and practices in place–are not, at least not wholly.

Where exactly do the current versions of total historicism (social constructivism, postmodernism) go astray? While I agree entirely with the insight about the origins of our values (that they are created by humans, not by God or Nature or Reality), I believe that many have confused this with a conclusion about the justification of those values. It was typical for Enlightenment thinkers to suppose that origin and justification were one and the same thing: to show that X is derived from God, or pure reason, or the senses, or scientific method was to have shown it to be duly justified. The idea appears to be that if a belief or idea could be shown to have the right lineage, a noble birth, then it was shown to have a noble status. But once we admit that God is dead or that "reason", "nature", or "scientific method" are not the monolithic eternal verities we once thought them to be, the inclination seems to be to slide into a universal skepticism or nihilism. In that sense the postmodernists are merely working with an uncritical assumption of the modernist thinkers whose ideas they find so untenable. But so what if our notions of method or rationality were not handed down to us etched in stone tablets? They can still prove their worth instrumentally. The notions of science, truth, and reason, etc. are not to be discarded tout court simply because their pedigrees are not divine.

The error being made here is, I would claim, a conflation of contingency with arbitrariness. Opponents believe that having shown a particular belief X to be contingent, that it was not necessarily fated from the outset of history that we would end up believing X, is to have shown that the fact that we do believe X must be for entirely arbitrary reasons. This is an all-or-nothing conception of epistemic legitimacy. It is as if having recognized the bankruptcy of divine right justifications for absolute monarchies, we were to judge all forms of government arbitrary and hence illegitimate. And perhaps

[283]

this anarchist analogy is not really that far off the mark as an analysis of the postmodern psyche. (It certainly captures important features of Feyeraband's thinking for instance.) But anarchism at least has some vision of how we are to proceed with our lives after the dissolution of coercive institutions. What is unclear about a lot of what is said under the banner of postmodernism is whether these critics of science and the Western tradition in philosophy have any constructive suggestions for how we are to make progress within the postmodern condition.

One figure who is professedly postmodern in sympathies and offers some positive suggestions about how we are to get on after we give up the notions of representing the world is Richard Rorty. And it is to Rorty's credit that he has, at least on occasion, recognized that one should not connate contingency with arbitrariness. Rather than concentrate on the correspondence between individual sentences and the world, Rorty focuses his attention on what he calls "vocabularies", which are roughly equivalent to Wittgensteinian language games. Aristotelian physics was one vocabulary popular for a long time, but then was replaced by a Galilean-Newtonian vocabulary (itself modified, although not wholly replaced by, an Einsteinean one). There are similar alternative vocabularies in use for political and social discourse (e.g. natural rights talk of early liberalism, class warfare talk of Marxism). Rorty's chief point is that the world itself does not tell us which vocabulary to use. "The world does not speak. Only we do" (Rorty, 1989, p. 6). However, he goes on to say, "The realization that the world does not tell us what language games to play should not... lead us to say that a decision about which to play is arbitrary ..." (Rorty, 1989). Why? Because the choice of which vocabulary to use, according to Rorty, is not really a decision at all. Following Kuhn he sees the transition from one vocabulary to another as something akin to a Gestalt switch, and these are not governed by any application of criteria or rule-following at all. They just happen, because one vocabulary simply strikes us as more attractive, more useful, or has better propagandists.

But having said all this Rorty does not give up on the idea of making progress in philosophy or in science or in culture in general. To be sure progress in science will be purely instrumental, there will be no accumulative progression to the one true theory about how the world really is. In this respect Rorty's philosophy draws a great deal from the instrumentalist ideas of people like Kuhn and Dewey. Making progress is not about approaching some absolute goal that exists independently of the particular vocabularies we employ. It is just about making local progress, moving away from where we presently are in a direction that we deem better relative to our current position.

But another essential feature of Rorty's writing is outwardly historicist. This historicism is intimately linked to what he calls his "ironist" position. As he explains in his own words, "Roughly, the ironist is a nominalist and historicist who strives to retain a sense that the vocabulary of moral deliberation she uses is a product of history and chance–of her having been born at a certain time in a certain place" (Rorty, 1998, p. 307, footnote 2). It might be objected that although this ironist position may make sense when applied at the level of the individual, it does not follow that it must likewise be applicable to the vocabulary as a whole. For although we might grant that it is a matter of history and chance (in some sense) that any one of us finds ourself born into the particular society and time we do, it does not follow that the things Rorty calls vocabularies are born or created in the particular time and places they do wholly as a

[284]

matter of time and chance. It was no chance accident, for instance, that Einstein's theory of special relativity took on the shape it did. Einstein was working within the prior tradition of Maxwellian electrodynamics and attempting to render that very successful theory consistent with certain recalcitrant phenomena and results. While it may be true, in other words, that history as we know it was not destined to fall out the way it did, it is certainly not true that history is just one accident after another.

Surely Rorty would not disagree with this. And yet consider the following statement: "The history of philosophy is the history of Gestalt-switches, not of the painstaking carrying-out of research programs" (Rorty, 1998, p. 11). Granted, he is talking about the history of philosophy here, but can its past be so easily disentangled from that of science? The emphasis and universality which he places on Kuhnian-type Gestalt switches in the histories of personal psychologies, therefore, is in deep tension with the continuity of the history of ideas, as considered apart from the historical circumstances of individual persons. So while Rorty may not equate contingency with arbitrariness, neither, in my opinion, does he give full recognition to the amount of criterion-based activity that goes on during periods of intellectual revolution.

Rorty advocates that we stop privileging activities which purport to picture reality as it really is, namely science and the analytical variety of philosophy which attempts to emulate it, and pay greater respect to the imaginative activities of poets and writers of fiction. For, he urges, the best hope for us citizens of liberal democracies to make further progress (moral and philosophical) is not through more closely reasoned argumentation, but through the use of creative (literary) imagination. It is the creative literary imagination which will provide us with the novel visions of how we might improve as individuals and as societies. That is why Rorty finds the postmodern writings of people like Derrida more interesting and valuable than the analytical philosophy still largely practiced in most English-speaking countries. Rorty's suggestion relies upon the assumption that the truth of statements and beliefs cannot consist in any kind of meaningful correspondence with the way the world is (indeed this is supposed to be an incoherent notion). This leaves us without any grounds for saying that certain statements are "objectively" true. The most we can say is that certain of them command a consensual agreement among members of specific communities. The goal of scientific activity, as with any other human activity, is not "objectivity" but solidarity. And once truth goes "immanent" in this way, Rorty urges that there is little reason to privilege one form of human activity (science, analytical philosophy) over any other activity that strives to create better communities and individuals.

Rorty is surely right to insist that we ought to value the humanities and literature just as much as we do the "hard" sciences. But does the argument for this laudable conclusion work? Other people have criticized Rorty's arguments about truth and language–Putnam, for example–so I will not go over that ground here.⁴ There is in fact quite a bit of common ground between Rorty and Putnam. But one way of stating their difference is to say that where Rorty adopts a full-blown total historicism, Putnam is willing to go no farther than epistemic historicism. It seems that what makes Putnam most uncomfortable about Rorty's "ironist" position is its apparent relativist implications. To avoid a similar implication of relativism in his own internalist position Putnam has attempted to develop a workable notion of "idealized warranted assertabil-ity". Such a grenzbegnff or "limiting notion" would place some boundaries on the thesis of epistemic historicism. In a recent article Rorty faults Putnam for refusing to follow the logic of his own anti-foundationalism and adopt a "consistently atheistic" position shaved of such absolutes (Rorty, 1998, p. 62). Rorty urges us not to look to eternity but to the

[285]

future (p. 174); Putnam, he suggests (pp. 62-63), in seeking a stand-in for universal validity is doing just the opposite.

One of the things which now worries Putnam about the Peircean-type accounts of truth and warranted assertability (those reliant upon the notion of convergence to some future-situated limit), is that they seem destined to be anti-realist when it comes to statements about the past. Rorty's position is that history, like any other inquiry, is valuable to the extent that it allows us to deal with problems and questions in an entirely instrumental fashion. And in fact this instrumentalist attitude toward history is enjoying some popularity with historians and philosophers of history today.⁵ Which brings us to the subject of speculative philosophy of history.

5.1 Rorty's speculative philosophy of history

In the introduction to her recent collection of essays, Aileen Kelly (1998b, p. 4) writes that despite various attempts to define the "postmodern" condition, "there is a consensus that its basic component is an acceptance that we can no longer credibly anchor our values in any universal ground, whether God, Reason, or History ..." Rorty is then mentioned as one of the most notable of those espousing this new historicism. Accompanying these developments within analytical or critical philosophy of history is a speculative picture of history. As Kelly writes, "The view that reality is inherently fragmentary, history a directionless flux, is the common ground of new theoretical developments across the range of intellectual disciplines". She then goes on to mention (p. 11) the paleontologist Stephen J. Gould's emphasis of contingency and the lack of ultimate directedness in natural history. This is an interesting connection Kelly notes between the new picture of history and recent developments in science and other disciplines. Nor has the connection escaped the notice of Gould himself, as the following quote from a paper co-authored with Niles Eldredge in the journal Nature bears witness:

[C]ontemporary science has massively substituted notions of indeterminacy, historical contingency, chaos and punctuation for previous convictions about gradual, progressive, predictable determinism. These transitions have occurred in field after field; Kuhn's celebrated notion of scientific revolutions is, for example, a punctuational theory for the history of scientific ideas. Punctuated equilibrium, in this light, is only paleontology's contribution to a Zeitgeist, and Zeitgeists, as (literally) transient ghosts of time, should never be trusted. Thus, in developing punctuated equilibrium, we have either been toadies and pander -ers to fashion, and therefore destined for history's ashheap, or we had a spark of insight about nature's constitution. Only the punctuational and unpredictable future can tell. (Gould & Eldredge, 1993, p. 227)

One way of understanding what the Rortian historicist-ironist is up to is to see her as posing for cultural evolution the same question Gould has posed for biological evolution (cf. Gould, 1989), i.e. Were we to play the tape of human history over again, how many of our current social ideas and institutions would we see re-develop? How much, if any, convergent evolution would there be in repeated trials of this experiment in human historical evolution? Would we have science? Would it look the same? Would we still get the atomic hypothesis, would mathematics look the same? Would our social and political institutions look familiar? Would they even be conducive to scientific understanding and research?

To suspect that the results would be wildly different each time is to take very seriously

[286]

the idea that we really are just children of chance and history. History on this picture begins to look like a random walk through time. One way of placing some subtle constraints on this picture would be to see the walk as taking place on a slope. This would allow for a wide range of play in the development of ideas and theories; yet by permitting a subtle drift in the direction of development, it retains an important role for a reality independent of and external to our social and epistemic systems. It would permit some hope that at least some of our ideas and institutions, while perhaps not being inevitable results of human history, are not quite freak accidents either.⁶

Now all of this is, of course, very speculative, and were it not for the fact that it appears to be a picture at least implicitly endorsed by a growing number of people, it would be merely idle speculation and of dubious importance.⁷ In closing, it bears repeating that care should be taken not to confuse the plausible-sounding claim about the contingency of social-cultural institutions with the very different pronouncement that they owe their existence to entirely arbitrary factors.⁸ Nor should we conclude from the contingency of our history that there is no objective and independent reality involved in its winding and weaving path. Any hope for improvements in the future direction of history would seem to require a commitment to belief in just such an objective reality, independent–in some crucial respects–of mind and language.

I would like to express my thanks to both the coordinators and participants of the 25th Annual Philosophy of Science Conference at Dubrovnik, especially to Arun Balasubramaniam, Dave Davies, and Dave Spurrett for helpful comments leading to improvements to the paper. Christine Levecq also provided helpful comments and suggestions. The Social Sciences and Humanities Research Council of Canada (SSHRC) is gratefully acknowledged for its financial assistance.

1. A word of caution is required, however, for Putnam writes: "Thus I agree with the subjectivist philosophers that there is no fixed, ahistorical organon which defines what it is to be rational; but I don't conclude from the fact that our conceptions of reason evolve in history, that reason itself can be (or evolve into) anything, nor do I end up in some fancy mixture of cultural relativism and 'structuralism' like some French philosophers. The dichotomy: either ahistorical unchanging canons of rationality or cultural relativism is a dichotomy that I regard as outdated" (Putnam, 1981, p. x).

2. See, for example "Why reason can't be naturalized" in Putnam (1983).

3. It should be noted that Putnam himself does not read Derrida this way, cf. Putnam (1994, p. 341). But despite this Derrida is often taken to have provided support for this position by enthusiastic readers.

4. Pumam's objection is that Rorty too quickly gives up on the idea of representation just because we cannot step "outside" of our language to compare it with the objects we are attempting to describe. We need not be able to do that just to refer to things outside of the language Putnam responds. See "The question of realism" and "A comparison of something with something else" in Putnam (1994).

5. See, for instance, Jenkins (1995, 1997) for an application of Rorty's ideas to historiography. Southgate

(1996) also adopts a postmodernist attitude toward history, but without an explicit reliance upon Rorty's ideas. Lorenz (1994) and Appleby et al. (1994) opt for Pumam's pragmatic realism.

6. James McAllister and Paul Thompson both suggested to me that this is one implication of Stuart Kauffman's work on self-organizing systems, cf. Kaufflnan (1995).

7. Just as interest in analytical philosophy of history has dropped off, interest in speculative philosophy of history has been increasing among many in the areas of literary theory and cultural studies. See Jenkins

(1997) for a sample.

8. Let me add that while I agree with the thesis that we alone are the source of our values, beliefs, etc., and that these have changed throughout history, I do not accept the relativist thesis that these values must be

[287]

restricted in application to the particular historical periods or epochs whose interests they are said to reflect. To insist that would require an argument that one had cut up the past into its proper historical "chunks" or periodizarions, and that this preferred classification was no more "arbitrary" than any other. (In fact it is a genuine question whether the past can be legitimately divided up into distinct periods.) Nor do I accept the claim that many or all of our most cherished beliefs are held for "arbitrary" reasons. I do not even know what it would mean to believe something for "arbitrary" reasons. The particular use of the notion of arbitrariness by structuralists and poststructuralists is one I have difficulty understanding. Unless by arbitrary they just mean contingent, and then it would seem that the points I have made here still stand.