The what and the why of history: philosophical essays 9789004103085

Citation preview

THE WHAT AND THE WHY OF HISTORY Philosophwal Essays

Series Editor

Michael Krausz, Bryn Mawr College Aduisory Board

Annette Barer (University of Pittsburgh), Cora Diamonp (University of Virginia), William Dray (University of Ottawa), Nancy FRASER

(Northwestern University), Patrick Garpiner (Magdalen College, Oxford), Clifford Grrrtz (Institute for Advanced Study, Princeton), Ernest GELLNER (King’s College, Cambridge), Peter Hacker (St. John’s College, Oxford), Rom Harre (Linacre College, Oxford), Bernard Harrison (University of Sussex), Martha Nusspaum (Brown University), Leon Pompa (University of Birmingham), Joseph Raz (Balliol College, Oxford), Amélie OxseNBerG Rorty (Radcliffe College), Georg Henrik Von Wricut (University of Helsinki) VOLUME 15

SEG

* ; ¥isry, veo oe)

e2, ‘hel a GOR os (16 8—

THE WHAT AND THE WHY OF HISTORY Philosophical Essays BY

LEON J. GOLDSTEIN

x EG ip .

ae EJ. BRILL LEIDEN - NEW YORK : KOLN 1996

The paper in this book meets the guidelines for permanence and durability of the Committee on Production Guidelines for Book Longevity of the Council on Library Resources.

Library of Congress Cataloging-in-Publication Data Goldstein, Leon J. The what and the why of history : philosophical essays / by Leon J. Goldstein. p. cm. — (Philosophy of history and culture, ISSN 0922-6001 ; v. 16) Includes index. ISBN 9004103082 (cloth : alk. paper)

1. History—Philosophy. I. Title. II. Series.

901—dc20 95-15082

D18.8.GB43 1996

CIP

Deutsche Bibliothek — CIP-Einheitsaufnahme

Goldstein, Leon J.: The what and the why of history : philosophical essays / by Leon J. Goldstein. — Leiden ; New York ; Kéln : Brill, 1996 (Philosophy of history and culture ; Vol. 15) ISBN 90-04—-10308-2

NE: GT

ISSN 0922-6001 ISBN 90 04 10308 2 © Copynght 1996 by E.F. Brill, Letden, The Netherlands All rights reserved. No part of thts publication may be reproduced, translated, stored in a retrieval system, or transmitted in any form or by any means, electronic, mechanical, photocopying, recording or otherwise, without prior written permission from the publisher. Authonzation to photocopy items for internal or personal use 1s granted by E.F. Brill provided that the appropriate fees are patd directly to The Copynght Clearance Center, 222 Rosewood Drwe, Suite 910

Danvers MA 01923, USA. Fees are subject to change. _ PRINTED IN THE NETHERLANDS

For Claire

BLANK PAGE

CONTENTS [troduction o........cecccescseececeeeeeeeseeeseneececeseceneeessscessseseeesesssessessssees IX

Acknowledgements ...........ssseceesseeresseesesseeseeseeetsteresssssseessseteeeses XIV T. The Why ou... eeeeeccccccccceceeeeeseeesssnneeeeseesesesssessseeeseeeeeeeeeeeneees l

1. Evidence and Events in History ........ eee seeeeeeeeseneeeees 3

2. Disposition Concepts and History ...... eee 28 3. Theory in History oo... eee eeeeseeceeereeeeetsestersssstttesererne OD

4. Ideals of Order: History and Sociology ............ 98 5. Historical Explanation and the Close of Inquiry ....... 81 IT. The What oo... cece eeeecsseeeeeseeeeeeseneeeeseeeeesseeeseseeesesseseneens OS

6. A Note on the Status of Historical Reconstructions ....... 95

7. The “Alleged” Futurity of Yesterday... 102 —

8. Historical Realism: The Ground of Carl Becker’s

SCEPUICISM 0... eee eeeeeeeeeesssseeeeeeeetstecesestsetestseesstessseersseees LOG

9. A Note on Historical Interpretation... LLY

10. Epistemic Attitudes and History oe 130

11. History and the Primacy of Knowing oe 143 12. Toward a Logic of Historical Constitution .............. 171 13. Impediments to Epistemology in the Philosophy Of HistOry oo... eeeesesseccceesseeceesseseeeseteeeeesesseeeersssteetssseee 207

14. Historical Being oo... eeeteeeeeesssteetssssserssestsssseenee 231 15. The Past of Our Present 0.0.0... eeeeeeeeeeeeeeeeteseeeeeeee 243

16. The Sociological Historiography of Charles Tilly ...... 254 TT. Collingwood oo... eeeeseceeeeesssescsessssssssssssssesessssssseeserene 27]

17. Collingwood’s Theory of Historical Knowing ............ 273 18. Collingwood on the Constitution of the Historical Past oe. ccccccccscececceeteseeessstesesssssesessststssssenee SLD

19. The Idea of History as a Scale of Forms... 337 Tex ooeee ee eeeeceeccesssneeeeessneceeeeseaecececssssaseesevseeecesssetsstecertessssteteeeens 349

BLANK PAGE

INTRODUCTION It is more than thirty-five years since I published my first paper in philosophy of history, and I am pleased to be able to gather nineteen of the twenty-two of them together in one volume.' Otherwise, they would continue to be scattered about in all manner of publications, any number of them lost to an audience their author obviously hoped to influence. In the very early years of my career, I published three papers in the Amencan Anthropologist, and it is no exaggeration to say that virtually no one in philosophy knows they exist. | am delighted to be able to save the papers which follow from that fate. I have always been interested in philosophy of history—I know I

wrote a term paper in that area when I was an undergraduate at Brooklyn College—but I would never have expected in my student

days that it would become such a dominant interest for so many years. I was a graduate student, at Yale, in 1952, when Patrick Gardiner’s The Nature of Histoncal Explanation was published, and I

had already sufficient interest in the subject to acquire a copy. In those days, there was an Oxford philosopher on the Yale faculty, and spotting him on the street, I stopped him to ask if he knew the author or knew who he was. He didn’t recognize the name and asked if I knew to what college he belonged. The book provided that information so I told him, and the name of the college called to his mind a person concerning whom he said—TI cannot quote him exactly after all these years—he always worked on the periphery of philosophy. He quickly qualified that by adding something like: at least British philosophy.

' The omitted three are in direct response to critics, and there seems no point in re-publishing them without the criticisms toward which they are directed. The omitted papers are “Social Science, Ontology and Explanation: Some Further Reflections,” Philosophy of the Social Sciences, vol. 4, no. 4, December 1974, pp. 359-68; “Against Historical Realism,” Philosophy and Phenomenological Research, vol. 40, no. 3, March 1980, pp. 426-9; and “The Way to Historical Being and Time,” Philosophy of the Social Sciences, vol. 16, no. 1, March 1986, pp. 101-5. * “On Defining Culture,” vol. 59, no. 6, December 1957, pp. 1075-81; “A Note on Platonism in Anthropology,” vol. 60, no. 3, June 1958, pp. 575-80; and “Ontological Social Science,” vol. 61, no. 2, April 1959, pp. 290-8.

x INTRODUCTION I cannot remember what my reaction to that was, but I have thought of that conversation many times over the years always with amusement. But perhaps my amusement is premature, misled by the extraordinary interest there now is in the subject in contrast to how little there once was. If there was a significant piece of writing in the

English language dealing with our subject which preceded F. H. Bradley’s The Presuppositions of Cntical History (1874), I confess that I

do not know what it 1s, and I believe that the first significant contribution to our subject by an American is Maurice Mandlebaum’s The Problem of Historical Knowledge, which was published as recently as 1938.

It is not too much to suggest that what started the outpouring of interest in philosophy of history—I refer, of course, to philosophy of history only in the critical sense—in the English-speaking world was the appearance, in 1942, of GC. G. Hempel’s classic paper “The Function of General Laws in History.”’ Given the character of philoso-

phy of science as it was practiced at that time, Hempel’s position made perfectly good sense. To begin with, the focus of attention in philosophy of science was on explanation, a problematic which was confined to what was called the context of justification. ‘Those who thought in that way were quite explicit that the context of discovery, the way in which one’s thesis came into being, was of no philosophical interest. And so Hempel simply ignores the context within which

the historical past comes to be known, and attends only to what is presupposed by historians when they try to explain a swath of the historical past no matter how it comes to be known. Virtually the entire thrust of the philosophical interest in history—surely, in the English-speaking world—was on the problem of explanation. It does not mean, of course, that everyone followed in Hempel’s direction and sought to talk about the general laws which justify the move from what is explained to what explains it. Some supposing that in point of fact historians did not actually explain by means of covering laws, turned to historians’ own writings to discover what they really did. And that led some to focus on the way in which historical events or actions seemed to be the outcome of what historical actors chose to do, thus making philosophy of history a branch of action theory.

3 Fournal of Philosophy, vol. 39, 1942, pp. 35-48, and reprinted so many times as to defy an attempt to list them all. One might mention, however, Patrick Gardiner, ed. Theones of History, Glencoe, IL, 1959, and a collection of Hempel’s own papers, Aspects of Scientific Explanation, New York 1965.

INTRODUCTION xl And others, noticing the way in which historians wrote up the results

of their investigations as narrative, began to attend to history as a mode of discourse making it subject to literary criticism and analysis.

The point of that orientation, of course, is that one explains what one explains by showing how it is required by the narrative. In sum, the thrust of philosophical reflections on history has been almost entirely on the why of history, almost entirely on the way in

which historical facts are to be explained. When I began to think about the problems of history, I, too, inclined to think in terms of the why, although it may be worth noting that my very first paper on the subject, “A Note on the Status of Historical Reconstructions,”

clearly anticipates my later interests. ‘To begin with, I had not yet thought my way out of the view—an utterly absurd view I now incline to think—that the context of discovery cannot be of interest to philosophy. That I had the good fortune to have been a student of Hempel in graduate school no doubt contributed to the formation of the attitude I brought with me to my earliest work on history. That interest is reflected in the papers which constituted the first part of this volume. But not all of the material included in that part are exclusively concerned with that. In going over this material in preparation for the present volume, I was amazed to discover how much of “Evidence and Events in History” anticipates what would become my veritable preoccupation with the what of history, and “Ideals of Order: History and Sociology” is likewise, not exclusively concerned with the why of history.

As I continued to think about the problems of history I came increasingly to reject the view that the context of discovery has no philosophical relevance. On the contrary, I came increasingly to think

that the nature of historical explanation is not all that interesting, and that it raised no special questions that were not raised in other areas in which matters had to be explained. The covering-law orien-

tation simply extended to history a point of view that had been developed in reflections on explanation in the natural sciences. Indeed, if one sees lurking behind the covering-law point of view the epistemology of David Hume and its particular sense of causality, one might argue that what is being extended to history is the sort of explanation we find in everyday common sense. And the action-theory orientation extends to historical figures the sort of explanation that is used wherever human behavior is rendered intelligible. It would thus appear that there is really nothing special about history as a discipline.

xi INTRODUCTION But such a conclusion would be mistaken. The trouble with philosophy’s preoccupation with the why of history is that it is precluded from seeing what that discipline is all about and how it carries out its business. The focus of attention is on the finished product of the historian’s work, the report, as it were, of the outcome of his

research. For the most part, but not entirely, that report takes the form of a narrative, an account which purports to tell us what took place in some swath of the human historical past. But how does one

come to know what happened in the historical past? In my judgment, the most interesting and exciting questions which confront a philosophy of history in the critical sense of the term have to do with how historians can claim—can claim responsibly—to know a past they can never experience. What 1s the nature of the discipline that makes such knowledge possible, and what can factuality, reference, truth and objectivity be in a discipline having such a nature? That I have come to believe that these are the important questions is clearly reflected in the fact that the section devoted to the what of history, the focus on history as the discipline which tells us what the past was like, is so much larger than the section devoted to the why. Collingwood’s The Idea of History appeared posthumously in 1946, four years after Hempel’s paper made historical explanation virtually

the only issue for philosophy of history of the critical sort. And so it was simply taken for granted that that was what Collingwood was dealing with as well, particularly in his well-known doctrine that historians re-think or re-enact past thought. Collingwood was simply absorbed into the dominant problematic notwithstanding the fact that his account of re-enactment actually begins with the words: “How, or on what conditions, can the historian know the past?”* Knowing is not explaining, and it is clear that Collingwood was working on the what not the why of history. I no longer remember when I first read The Idea of History, but my recollection is that my first understanding of what he was doing was along the lines of everybody else. At the beginning of “Evidence and Events in History,” which was published in 1962, I talk about giving some view of Collingwood a methodological interpretation, which suggests a suspicion that there was more to what he had been doing than the explanation-reading of him, but I had not yet abandoned

* R. G. Collingwood, The Idea of History, Oxford 1942, p. 282; italics added.

INTRODUCTION xiii the view that re-enactment was intended to be a species of explanation. Some years later, while writing my Historical Knowing, a book

that is entirely concerned with the what of history and seeks to understand factuality, truth, objectivity and reference in history by taking history seriously as a way of knowing, I remember stopping suddenly and saying to myself: maybe this is what Collingwood was trying to do all the time. I re-read the /dea of History, and discovered that that was, indeed, the case. I then proceeded to read Collingwood’s Speculum Mentis, the penultimate chapter of which deals with history, and then proceeded to read Collingwood’s occasional pieces on history which had been brought together conveniently in a volume edited

by William Debbins.’ With all that read, I proceeded to write the first of my Collingwood papers, “Collingwood’s ‘Theory of Historical Knowing,” which was published in 1970. “Collingwood on the Constitution of the Historical Past,” continues further my interest in what

Collingwood says about and actually does in trying to know the historical past, and “The Idea of History as a Scale for Forms,” deals with Collingwood’s sense of the discipline of history as itself having a history. *

In closing, I should hke to express my gratitude to Professor Michael Krausz, editor of Brill’s series on Philosophy of History and Culture,

for his enthusiastic encouragement of this project; to Brill editors Elisabeth Erdman-Visser and Gert Jager for the things they have done to facilitate publication; and, last but definitely not least, to Jeanne S. Constable, secretary of our department, without whom nothing gets done.

> R. G. Collingwood, Essays on the Philosophy of History, edited with an Introduction

by William Debbins, Austin, 1965.

ACKNOWLEDGEMENTS

The papers which follow were all published previously, and I want to express my appreciation and gratitude to editors and publishers for permission to reprint them here. ‘The papers and their original places of publication are listed as follows and in the order of their appearance in this volume. “Evidence and Events in History,” Philosophy of Science, vol. 29, no. 2, April 1962;

“Disposition Concepts and History,” Mind, vol. 72, no. 288, October 1963;

“Theory in History,” Philosophy of Science, vol. 34, no. 1, March 1967,

and in Boston Studies in the Philosophy of Science, vol. 4, Reidel, Dordrecht 1968; “Ideals of Order: History and Sociology,” Philosophy of the Social Sciences, vol. 4, no. 4, December, 1974; “Historical Explanation and the Close of Inquiry,” Jnternational Studies in Philosophy, vol. 9, 1977; “A Note on the status of Historical Reconstructions,” Journal of Philos-

ophy, vol. 55, no. 11, May 22, 1958; “The Alleged Futurity of Yesterday”, Philosophy < Phenomenological Re-

search, vol. 24, no. 3, March 1964; “Historical Realism: The Ground of Carl Becker’s Scepticism,” Philosophy of the Social Sciences, vol. 2, no. 2, June 1972; “A Note on Historical Interpretation,” Philosophy of Science, vol. 42,

no. 3, September 1975; “Epistemic Attitudes and History,” Philosophy © Phenomenological Re-

search, vol. 37, no. 2, December 1976; “History and the Primacy of Knowing,” History and Theory, Betheft 16,

vol. 16, no. 4, 1977; “Toward a Logic of Historical Constitution,” Boston Studies in the Philosophy of Science, vol. 71, Kluwer, Dordrecht, 1983; “Impediments to Epistemology in the Philosophy of History,” History and Theory, Betheft 25, vol. 25, no. 4, 1986; “Historical Being,” The Monist, vol. 74, no. 2, April 1991; “The Past of Our Present,” The Gallatin Review, vol. 12, no. 1, Win-

ter 1992-1993;

ACKNOWLEDGEMENTS XV “The Sociological Historiography of Charles Tilly,” Developments in Modern Histonography, ed. H. Kozicki, St. Martin’s Press, New York, 1993;

“Collingwood’s ‘Theory of Historical Knowing,” History and Theory,

vol. 9, no. 1, 1970; “Collingwood on the Constitution of the Historical Past,” Critical Essays on the Philosophy of R. G. Collingwood, ed. M. Krausz, Clarendon

Press, Oxford, 1972; “The Ideal of History as a Scale of Forms,” History and Theory, Betheft

29, vol. 29, no. 4, 1990.

BLANK PAGE

PART ONE

THE WHY

BLANK PAGE

EVIDENCE AND EVENTS IN HISTORY I. Historical Past and Real Past

There is an air of paradox about hearing, as we do from Collingwood, that history is present history. ‘The events that are described in the writings of historians, we protest, did not just happen—today or the day before yesterday. It was long before anyone now can remember that Caesar crossed the Rubicon, Pope Urban preached his crusade, and Nelson died at Trafalgar. The whole point of history is the past,

when those events described in history books were really present. Thus, in history idealism is finally routed by realism (cf. [17]). Yet there zs a certain plausibility to the idealist view, one which rests upon the distinction between history as a course of events that really happened when the past was present and history as a discipline the purpose of which is to descnbe and explain the events in question. The latter is clearly an affair of the present. Presumably the realistic critic thinks that by making the distinction and giving the Collingwood thesis a methodological interpretation he has drawn

its sting and rendered it harmless. No doubt, it is much improved and no longer paradoxical when so interpreted. But what has the realist left for himself? We shall see that his distinction not only makes the idealist’s thesis entirely reasonable, but, at the same time, renders his own entirely beside the point so far as history as a human enterprise 1s concerned. The starting point of historical inquiry is a body of evidence. To

be sure, historians will tell us varying things about what brought them to their subject, and not least among them is interest in or curiosity about the past. They want to know just what it was that happened at this or that point in the past, both because of any intrinsic interest it may be said to have and because of a belief that knowledge of the past may be put to practical use in dealing with problems of our own day. But whatever the personal motives of historians, these are not to be confused with the actual starting point of history as a discipline, and curiosity about the past—in a realistic commonsensical sense—does not establish that past as the starting point, or even the end point, of historical investigation. Inquiry, then,

4 EVIDENCE AND EVENTS IN HISTORY begins with a body of evidence. And the problem of the historian 1s to explain the evidence. It is often thought that the historian’s primary problem is to explain why something or other happened or why someone did this or that. Actually, there 1s a previous step and that is, in commonsense language, to determine that something happened or that someone or other did something. It is often thought that these important facts about the real past are discovered by historians through the means of their evidence. Whatever is intended by knowledge of the past, presumably all would agree that it depends upon historical evidence, yet so widespread is the commonsense realistic view of the nature of history that most philosophical writers simply assume that in one

way or another the real past can be revealed and that the major problem is to account for the methods and laws which the historian uses in explaining it.' The philosophical problems of history, however, are by no means limited to the explanation of events, and we shall want to consider just how events are established.

When we say that the starting point is the evidence, we mean only that the suspicion that there were events is suggested by the fact that there are present certain things which seem not to fit into the present context of culture and life: writings which most of us cannot read, coins which will buy nothing at the grocery, ruins of buildings and of entire cities, and so on. Some writers (e.g., [28]: 99ff. and elsewhere) see in these and similar phenomena traces of the past, which when properly interpreted give us some knowledge of how things once were. If taken literally, this view seems to be that the method of history enables us to get to the past, that the traces provide some more or less direct contact between the present and the past. The properly trained historian, then, is one who knows how to find the past in things which are present. Historical inquiry, to be sure, is of the present, but it results in uncovering before our eyes some aspects of the real past. Presumably, this is the point of the distinction to which reference has been made above. ' Though the books and papers on history of Hempel, Popper, Gardiner and Dray are by general agreement among the most important discussions of the subject by contemporary philosophers, their readers will not find in any of them an account of the problem of historical knowledge and, particularly, of the relation of evidence and events. Now that Teggart’s Theory and Processes of History [32] is once again in print, it may be hoped that more philosophers will take advantage of the opportunity to read its excellent account of precisely this problem.

EVIDENCE AND EVENTS IN HISTORY 5 But surely it is odd to talk of having contact with the past in any literal sense. No matter how diligently he has carried out his investigation, the historian never gets any nearer to the real past than he was when he began. The real past, whatever it was, has no more to do with history—the discipline—than Hylas’ material substance has to do with the experienced objects that Philonous wants to talk about.

In both instances there may be reasons for wanting to believe in substance or in the real past, for one does not want, presumably, to say that either the reality of objects or of human experience 1s dependent for its existence merely upon the thought of an inquirer or a perceiver. But however one wishes to effect the avoidance of solipsism, neither the thing as it exists apart from our experience of it nor the real past figures in any human activity.’ There is no unequivocal evocation of the past by the evidence, and calling present evidence traces of the past does not affect this at all. It sometimes happens that not-withstanding general agreement on what the evidence is, conceptions of what actually happened differ.

The difference is over what the past would have to be in order to best make sense of the evidence. ‘The past that the historian evokes is not a real past as it was when it was present, but rather a construction of his own: not, to be sure, a free creation as in the writing of a novel, or a tendentious creation contrived for the purpose of propaganda,’ but a construction devised as the best explanation of the evidence he has. The historical event—the only historical event that figures in the work of historians—is an hypothetical construct. The historian does not look for evidence in order to explain the event, as if the event is clearly before him and he is required to

make sense of it, but, rather, he calls it forth for the purpose of explaining his evidence. And while one might want to say that the * Cf. Mead: “The outcome of what I have said is that the estimate and import of all histories lies in the interpretation and control of the present; that as ideational structures they always arise from change, which is as essential a part of reality as the permanent, and from the problems which change entails; and that the metaphysical demand for a set of events which is unalterably there in an irrevocable past, to which these histories seek a constantly approaching agreement, comes back to motives other than those at work in the most exact scientific research” ({20]: 28). The entire chapter from which this is taken, especially the note on pp. 28-31, may be read with profit. ° I fear that this is a distinction which is lost upon Mr. MaclIver, who seems to think that Collingwood’s view that history is present history is of a piece with and philosophical justification for the treatment of Trotsky’s role in the Russian Revolu- _ tion by Stalinist historians ({17]: 200f.).

6 EVIDENCE AND EVENTS IN HISTORY historical construction which most nearly describes the real past 1s the best or the truest account, how can we ever know? How can we ever test the event except in terms of our evidence? We can never compare conflicting accounts of what “really happened” then and there with anything but our evidence. I am not, to be sure, denying the metaphysical proposition that the past is or was real, but only observing that, it is irrelevant to history. Not even an eye-witness account provides an exception to this. From the point of view of historical research, one should refer to it as “what purports to be an eye-witness account,” for that it is such

an account is an hypothesis that may explain it. It is the simplest hypothesis for dealing with the account itself, but it may well have to be rejected if there are reasons which make an alternative hypothesis more reasonable.* The simplest hypothesis concerning the date of the Old Testament book of Ezekiel—which would take as

relevant evidence the data upon which our knowledge of ancient Israelite chronologies are based and the second verse of the book’s first Chapter—would place it early in the sixth pre-Christian century. If one broadens the basis for judgment, that is, if one concludes that an hypothesis concerning the date of the book of Ezekiel must take account of rather more evidence than the two just mentioned, then

one may feel the need to offer a somewhat different hypothesis. Presumably one factor in back of scholarly disagreements in the matter is disagreement over what configuration of evidence——and its charac-

ter—must be taken account of; though it is reasonable to think that most scholars would agree that stylistic, lexicographical and ideological

factors must be included in any assessment. And while all would recognize that the second verse of the first chapter is part of the book, it need not follow that the truth of its claim must be accepted. If the only evidence is the verse itself and our chronology, then presumably our hypothesis accepts the content of the verse, but if we broaden the evidential base we are at liberty to offer an alternative. And if we can explain how it is that the verse is mistaken about the * My impression, based upon general reading of history books, is that historians will usually prefer this simplest hypothesis, and where other evidence is wanting will assume that their documents are rather exhaustive. For an example of this see Belkin ({3]: 5£.), who seems to think that because “we know of only one visit of Philo to Palestine” we must assume that he never made others, though so little is known of his private life that his having gone every other spring is entirely consistent with our evidence. For a cogent discussion of this problem see Johnson ([15]: 47ff).

EVIDENCE AND EVENTS IN HISTORY 7 date, presumably our hypothesis that it is mistaken 1s even stronger. It may prove instructive to pursue this illustration further by presenting, however briefly and inadequately, an alternative that was proposed by the late Charles Cutler Torrey.” ‘That this proposal is generally regarded as extreme does not affect its usefulness here. In fact, Torrey’s general tendency to be extreme is not a quirk which

finds expression in a variety of ways, but actually represents one extreme interpretation of the historical status of the bulk of the later Old Testament books. From that point of view, his work is an excellent exemplification of what I am trying to say about history in this paper. For Torrey’s hypotheses are an attempt to offer an explanation for a large body of historical evidence. The evidence from the biblical book consists not only of its statements, but also of any number of elements that its writer or editor

never dreamed they would be contributing to the work of future historians. There is, for example, the character of the Hebrew in which it is written: how does it compare with the character of the Hebrew of other books the dates of which may be less problematic? Torrey is of the opinion that its Hebrew reflects the strong influence of Aramaic and considers that this, among other reasons, supports a

later dating than would be warranted were we required to decide merely on the basis of the statement in the first chapter. In addition, there is what, for want of a better term, may be called the ideological element. Torrey thinks the book is rather late—a view supported by several ancient traditions—not only for lexicographical reasons, but because he thinks its ideological content requires that it be treated as part of a group of late works, such as the books of Ezra, Nehemiah and Chronicles. But he thinks an attempt was made to make it appear earlier. The books of Ezra and Nehemiah reveal a conflict between

the returnees from the Babylonian exile and the Samaritans; and Torrey believes that the purpose of making Ezekiel seem earlier was to establish a continuity of tradition for the returnees to oppose to the manifest continuity of the Samaritan tradition which, of course, had not been uprooted. Whether or not ‘Torrey is right, his argument provides a useful illustration of our point. One does not simply look

at the document in order to know what happened. The author or > [33]. It will be noted that in using this work as an example I am not to be taken as endorsing its conclusions. Interested readers may consult Pfeiffer ([24]: 525ff) for other theories as well as references to Torrey’s critics.

8 EVIDENCE AND EVENTS IN HISTORY editor of the document which provides the point of departure for Torrey’s book may well have wanted his contemporary readers to believe that something was the case which in fact was not. That it was or was not is part of the hypothesis put forward which explains the document and other parts of the evidence.® Scholars who disagree with Torrey—which means practically every-

one else in the Old ‘Testament field—would offer different reconstructions. One might say that they offer entirely different historical events as better explanations of the evidence. Indeed, these different events might be radically different from one another, and the various temporal locations which have been proposed for the book of Ezekiel

range over hundreds of years. As we come closer in time to the present, divergences are likely to be less startling, for evidence is apt to be fuller and the possible events which would make it intelligible

are fewer. Too, in fields of history where such general agreement exists, the actual starting point of the individual historian is likely to be the professional tradition concerning its character, which is easily

mistaken for the real past. But it is clear that what we have said about ancient historical events obtains for later ones as well. When our concern is with late events about which general agreement exists and we seek to fill in details here and there, it is often the practice to look at documents and assume that the details are filled by being drawn from the documents. But one could just as easily say that what is drawn from the documents enlarges our evidence, and the event as enlarged in detail is a slightly improved hypothesis which explains all the old evidence plus the newly added details. This view has the advantage of giving the same status to all events. Not, to be sure, in the sense that all events are equally well established, but rather that from the point of view of an inquiry into the methodol-

6 Cf. Oesterley: “In regard to much that has been said we are prepared for the objection that the evidence of the Old Testament does not offer sufficient justification

for the assumptions made. We agree that this is so if we are to rely upon the Old Testament alone. But the object of the whole of our investigation will be to show that the beliefs and practices of any one race of people must, to do them full justice, be studied in the light of analogous beliefs and practices of other peoples. Only so can one fill up the dacunae which inevitably exist in the records of the races of antiquity” ({23]: 42). ‘Thus the kind of historical event or reconstruction offered will

depend upon what one takes the relevant evidence to be. One may suspect that it is precisely this which is at the heart of the disagreement between philological-literary and folkloristic-comparative religionist approaches to ancient literatures; (cf. [10]: IXff).

EVIDENCE AND EVENTS IN HISTORY 9g ogy of history all events are hypothetical constructs introduced for the purpose of making our evidence intelligible (cf. [2]). Il. Aistorical Evidence

In this section of the paper I just want to discuss precisely what 1s meant when I talk about historical evidence, and what is meant by insisting that concern with it, rather than, say, with the explanation of historical events, is that activity which distinguishes historical research from any other intellectual activity.

Relevant to any theory is a body of evidence which is said to confirm or disconfirm it. From another standpoint, it may be said that

the theory in question explains the evidence which confirms it, or fails to explain it if the evidence disconfirms it. Thus, it appears that the term “evidence” is a correlative term; something is not evidence simpliciter, but rather evidence relative to some theory or hypothesis. Since we may believe that everything is subject to being explained,

there is nothing that cannot be evidence. Yet in itself it is not evidence, or, if you like, “evidence” is not a monadic predicate word. In using the term “historical evidence,” I intend to depart from this usage. It is not that this departure is required for what follows, yet it seems to me somewhat convenient to have the term “historical evidence” refer to a determinate body of data—that very same body of data to which historians make reference when using the term— even when no hypothesis exists for its explanation. Thus it would have been perfectly correct to refer to the very first cuneiform tablets

discovered in the orient as historical evidence even though no one had the foggiest idea as to what they were evidence for. One could perhaps see that, as the scratchings on the tiles were not produced by nature but were rather man-produced, they could have some possible bearing upon our understanding of man’s career on earth. One walks along and discovers rocks of all kinds, sizes and shapes. Presumably men have always found them, been fascinated by them, and made use of them. Being rocks, they need not be evidence. They do not become evidence—say, for the age of the earth, the geologi-

cal character of the area in which they are found, or the material cultural attainments of prehistoric communities which may have made

use of them—until geological or archeological theories are formulated. So it is with historical evidence; it does not become evidence

10 EVIDENCE AND EVENTS IN HISTORY until historical hypotheses are formulated. Stl, a certain class of phenomena may be recognized as being rocks, and another class of phenomena may be recognized as being historical evidence. ‘Thus the word “evidence” in the phrase “historical evidence” does not mean, at least in this paper, what “evidence” ordinarily means. Rather,

the logic of “historical evidence” is like, if not actually the same as, that of “rock,” for each is the name of a class of objects. It is with historical evidence that historical research begins, and if history as a discipline may be thought to have made progress over the course of years it is hkely owing to growing sophistication in the treatment of historical evidence. It is no longer possible to write Roman

history by reading the great historians of antiquity—Livy, Tacitus and their confreres—and recasting the events they describe in a new narrative form. If the narrative of the historian is, as many of them say, a creative and original act, it 1s, at the same time restricted by its need to conform to historical evidence. No one can properly appreciate what is involved in historical research who has not paid attention to those disciplines—so-called ancillary disciplines—which are devoted to the assessment and classification of such evidence. ‘The following paragraph, taken from a work

by a historian, will make clear what these are. Protracted intensive study of these manuscript sources has given birth to a numerous progeny of highly specialized arts, such as paleography, epigraphy, diplomatics, sphragistics, and heraldry. ‘To the paleographer, a specialist in handwriting, the historian owes the use of many manuscripts which, unnamed, undated, and sometimes undeciphered, would have been closed codices. The study of epigraphy has made inscriptions available for historical purposes. And because the chanceries of Europe had different methods of drafting, signing, sealing, and attesting decrees, ordinances, charters, and other records, a fund of special knowledge had to be accumulated before scholars could identify and interpret the documents. The term diplomatics covers this highly important art; but diplomatists must often summon two other specialists to their aid—the student of seals and the expert in heraldry ({15]: 16f:;

cf. [16]: chs. 1 & 2; [4]: ch. 3).

To Johnson’s list one might add numismatics as well as the treatment of artifacts by archeologists. What the practitioners of these various disciplines do is classify and correlate historical evidence. Once

the scholar has become proficient in his work he has but to look at

a document, say, and know where and when it was written. The form of the letters, the character of the dialect, the use of technical

EVIDENCE AND EVENTS IN HISTORY 1] or legal terms all give away secrets which would be forever lost if we

had to depend only upon explicit dating and locating within texts themselves. That modern scholarship has been able to discover that so many ancient texts are pseudepigrapha is owing entirely to the successful pursuit by scholars of these ancillary disciplines. ‘The simple-

minded counterfeiting of coins, as in the Renaissance period, to be passed off as products of classical antiquity is simply no longer feasible,

and only a scholar as learned as he is perverse could hope to do it successfully today.’

The widely held view that historians are not concerned with generalizations or that history is not a generalizing discipline is entirely untenable in light of the fact that these ancillary disciplines, which are the peculiarly historical disciplines, are clearly generalizing. It is,

of course, true that the generalizations they use are often unformulated,

and certainly never formulated with the precision that characterizes generalization in physics. Yet they are generalizations for all that.

Johnson. , And they have had much successful application to their credit. I suspect

that there are some branches of the study of documents or coins for

which the generalizations require very little qualification, and the result of their application may be very striking indeed. Again, we may quote

The results achieved by these laborious studies are often amazing. The German historian Giesebrecht, for example, pointed out that certain chronicles of the eleventh century drew their information from a com-

mon source which was no longer available. So sure was he of their dependence upon this early chronicle, which he named Annales Altahanses,

that he constructed the missing manuscript from these later derivatives. Some twenty-six years later the missing chronicle was found, and it confirmed these shrewd conjectures in every important particular ({15]: 17).

Such results show clearly how well established is knowledge of the

writing of the period in question. On the basis of what he knew about how works influence subsequent ones, our historian was able to propose that a bit of historical evidence of determinate character must be assumed in order to make sense of certain pieces of historical evidence which were known to exist. In its own way, it is no less ’ With all the progress that has been made in the application of physics and chemistry to the dating of artifacts and other finds, even this may no longer be possible; cf. New York Times, January |, 1960, pp. 1, 8.

12 EVIDENCE AND EVENTS IN HISTORY impressive than the anticipation of the discovery of the planet Neptune,

particularly inasmuch as it reflects a degree of systematization in historical research of which many of the philosophers who write about history seem unaware.

In the course of working with the results of the various ancillary disciplines, historians discover a kind of convergence. By this I mean

that certain documents seem to belong with certain artifacts, coins, and, if the period is sufficiently ancient, ruins. There emerges in the course of research a constellation of kinds of historical evidence. To account for the particular constellation becomes a question of some interest, and it is at this point that the historical event enters into consideration.® It is the function of the event to explain the evidence, that is, 1t must make intelligible the grouping together of some particular constellation of historical evidence which is believed to belong together. We have seen that differences between historians often stem

from failure to agree on what the proper constellation is. And even were such agreement to be had, some disagreement could result from the incompleteness of the evidence. Different reconstructions may begin

from and be consistent with the same evidence,’ but would have different implications so far as evidence not known or not yet known is concerned. Should new evidence turn up, it would be possible to choose between differing reconstructions or propose still a new one (cf. [11]: 475). But, even then, we would have no reason for taking

| the historian’s event to be a description of a real event. Il. Explaining Events

Though we have seen that historical events are constructed for the purpose of explaining historical evidence, it is possible to treat such events as descriptions of states of affairs. And one may wish to have an explanation of the state of affairs quite apart from our recogni° In prehistory, the constellation of correlated finds is taken to represent a stage of cultural development, but we sometimes discover a realistic tendency to think of the stage as independent of the constellation of evidence and to treat it as real even when newer evidence leads to the breaking up of the original constellation. If one thinks of the stage as an hypothesis to explain a particular constellation this would be less likely.

’ Though not if the view that reconstructions are deduced from the evidence were correct, for in that case agreement about evidence should entail agreement about events.

EVIDENCE AND EVENTS IN HISTORY 13 tion of the function of the historical event in explaining the evidence.

It does not seem that this problem is in any way different from attempts to explain states of affairs in general; that is, that they are states of affairs described by historians, hence not capable of being witnessed by anyone now, does not seem to raise any special problems.

They may be classed with states of affairs described by present-day sociologists or anthropologists, journalists, or Just perceptive eye-witnesses. That an historian finds it best—given his historical evidence—

to believe that a state of affairs of determinate character took place at some past time in no way affects the character of the event. Presumably, what will explain it would have explained it no matter when it

happened, and if events of like character happen in our own time, like explanations will be required. If one distinguishes between the phenomenal panorama passing before the eye-witness and the report he makes of what he observes, one could argue that the latter is rather like our historical event in that it is possible to treat it as explaining the phenomenal panorama. Likewise, a sociological or ethnographic report may be treated as introduced for the purpose of making sense of what the sociologist or ethnographer saw or experienced. In all of these cases there seems to be a certain parallelism, and one is at liberty to conclude that all events as characterized in reports are hypothetical. ‘The difference, between historical events and other kinds of events seems, then, to be even less, than we commonly think. But it would take us too far afield to explore this here. It will suffice simply to note that when we

are concerned to explain an event as a state of affairs we are not required to make distinctions of a logical character which depend upon the event’s temporal location.'° . As has been noted earlier, the problem of explaining historical events is precisely one of the problems of history which has received

most attention from philosophers in recent years. The question to which most inquiry is directed is whether or not such explanation requires that reference be made to general laws. I do not think it appropriate to burden this rather long paper with a review of what

'0 ‘While one frequently hears it said that concern with time is especially characteristic of history—-(cf. [30]: 24)—this can only mean that historical events are ordinarily understood as not being instantaneous. ‘Temporal location, however, 1s simply a specific or more determinate form of existence, and if existence 1s not a predicate neither is temporal location.

14 EVIDENCE AND EVENTS IN HISTORY has been written on the subject, but will, rather, turn directly to the task of presenting some views of my own. Let us assume that we are watching two little boys at play, and that suddenly one punches the other in the nose. The victim of this outrageous assault proceeds to burst into tears. Further, assume that soon thereafter someone else comes up, sees the crying child, and asks us why he is crying. We say that the other punched him in the nose, and intend by this to answer the question. Is it, however, an explanation? Are we not, perhaps, required to connect crying and being-punched-in-the-nose before we may claim actually to have explained the crying? Michael Scriven would say that the answer given is a complete explanation of the phenomenon to be explained, and that we are required to separate the explanation itself from any justification of the explanation, which is what connecting the two would be. He believes that the former may be had without the latter

([29]: 445). But rather than having someone else come up and witness the crying of the child, let us assume that instead it is a creature of some sort who has never before witnessed the phenomenon of someone crying, and let us assume that it asks the question. (For economy of expression, I shall use the made-up word “hume” to mean “a creature

of some sort who has never before witnessed the phenomenon of someone crying’). Surely the answer does not explain the phenomenon

to hume. To hume, the answer may seem very strange indeed; he asks about crying and we tell him that someone was punched. For hume the connection is by no means obvious, for either one may be thought about without involving or entailing the other. Far from being

a truism, as Scriven thinks, the answer seems beside the point. It will not do to dismiss this as a silly and improbable example. The difficulty of hume could easily be extended to every so-called truism or common sense explanation of an event, historical or other. If the answer to the question is actually an explanation it is only because implicit in it is the generalizing of our experience with being

punched and crying. Too, if anyone thinks that reference to his cupidity explains Cortez’ decision to make a third expedition into Lower California—to use one of Scriven’s examples—it is because we know something about cupidity and its effect upon the actions of men. Thus, it emerges that we accept an explanation as explaining only if we can justify it. To the extent that we cannot justify it, it remains problematic or doubtful or even unintelligible.

EVIDENCE AND EVENTS IN HISTORY Id An alternative to the kind of justification offered above would require that the denial of the necessary connection between empirical

concepts be itself denied. The mind might be said to grasp intuitively that between the “truism” and the event explained there is a direct connection, and this intuition in no way depends upon previous experience in order to be discovered. I do not know how many of the recent critics of the view that explanation in history presupposes general law would be willing to subscribe to a view such as

this,'' and it may well be that, in his firm present grasp of the meanings and uses of the truisms he discusses, Scriven has simply forgotten that he did not always know them. The claim, then, of some philosophers that in history, as in any discipline, explanation requires reference, at least implicitly, to general laws, does not arise out of a study of historical writings, but is rather

a consequence of a certain position in epistemology, a position traditionally associated with the name of David Hume. Thus, the fact that historians do not as a rule’ seek to formulate generalizations cannot be taken as an argument against the view, as Dray seems to think it can ([8]: 11f., 139, and elsewhere). One might have wondered before raising such a point how it is that such methodologists

as Hempel and Popper, who surely know that historians do not generally formulate laws, nevertheless persist in their view. It seems clear that neither one of them considers this fact about historians to be pertinent, and that is because ours is a problem in epistemology, not historiography.'’ On the epistemology of empiricists no intuitive connection between an event and any other event is descernible. If

one wants to attack this view, one ought to do it outright and not obfuscate a philosophical issue by appeal to historiography. Such an appeal does not actually lead to a resolution of the question. It is, of course, the case that historians generally do not specify laws, but rather seek out antecedent causes of a highly specific character. ‘' It may be that Mandelbaum does, or at least did at one time; (cf. [18]: 206ff.) and see my [11]: 467ff. '? See [31] for an interesting exception to the rule. ‘> Cf. Donagan: “It would not follow that the Hempelian theory is mistaken even

if it should turn out to conflict with historian’s or common reader’s estimate of recent historical work.” That he adds “Berkeley’s criticism of fluxions has worn better than the replies of contemporary mathematicians,” may mean that he is only suggesting that the Hempelian theory may be a sounder account of historiography than that of most historians, which is not of course, the point that I am making in the text ([6]: 145f,).

16 EVIDENCE AND EVENTS IN HISTORY From the empiricists’ standpoint, however, the claim that this specific event led to or caused the event we seek to explain implicitly makes reference to general laws because on their epistemology there can be

no other way of justifying the claim that the two events are connected. It may be that the reason why the epistemological basis for the explanation-by-general-laws position is never taken up is that such

of its critics as Dray and Scriven are themselves in the empiricist tradition.

The outstanding accomplishment of physical science has resulted

in that philosophy of science has tended to the view that it must analyze the logic and procedure of science, but must be careful not to dabble with matters of a substantive character. It is certainly the case that only one of unusual audacity could presume to challenge the material knowledge claims and methods of procedure of physics without first becoming thoroughly grounded in that subject himself: It is habits of work which come from the philosophy of physical science that writers such as Scriven bring to their treatment of explanation in history. Just as one starts the philosophy of physics by careful study of what physics has accomplished, so one begins philosophy of history—in the critical rather than speculative sense—by becoming familiar with what historians have done. And historians have not been formulating general laws. ‘To say they do would be erroneous; to say they should presumably presumptuous. Yet neither history nor social science has so impressive a record as physics, and it may be doubted that there is much that is worthwhile to be derived today from a philosophy of social science which adopts the scrupulous hands off policy of philosophy of physics. We must be prepared to say what we think history and social science lack no matter how much we may hesitate to speak similarly about physics. With this much said, let us now consider the character of historical or sociological laws. Things being what they are, some of the next section will be speculative. IV. Social Laws

Those who have cniticized the possibility of the application of general laws to the study of history have often buttressed their views by pointing to the simple-minded or trivial character of the laws that would be involved. Popper, whose views on explanation require the

EVIDENCE AND EVENTS IN HISTORY 17 utilization of general laws, has insisted that they must be trivial, presumably because of his belief that any non-trivial law in history would be historicistic.'* Thus, he suggests that one way to explain

the partition of Poland in 1722 would be by application of a law such as “If of two armies which are about equally well armed and led, one has a tremendous superiority in men, then the other never wins” ({26]: 448). He allows that “never” may be interpreted as “hardly

ever,” no doubt to allow for such happenings as the bigger army marching into quicksands. Not significantly different is Michael Scriven’s conception of the character of laws that might be used in history, if any were at all. In

considering ways in which we might explain his aforementioned example of Cortez’ decision to make a third expedition to Lower California, he thinks that advocates of the view that general laws are relevant to explanations in history would have to consider the following: “All confident wealth-seeking people undertake any venture which offers very great wealth” or “All confident people seeking very great wealth undertake any venture which offers very great wealth,” or, finally, “All confident people with Cortez’ background of experience, seeking very great wealth, undertake any venture involving the hazards of this one, which offers very great wealth” ([29]: 454). In what follows, he argues that the most general of these—and presumably of any set of such statements no matter what historical event or set of events they were introduced to explain—is most likely false, whereas the one least likely to be false, because of its increased specificity, is hardly worth formulating. But for all their differences, the three are alike in showing that Scriven considers that general laws in

history could only be generalized from the particular event to be explained. It is no wonder, then, that he can think only of trivialities. Of little interest, likewise, is Gardiner’s conception of a law adequate to explain the unpopularity of Louis XIV at the time of his death," and Gardiner is one of the defenders of the general laws-in-history

'* In his sense of “historicism” as belief in laws of development rather than in Mannheim’s sense as the social determination of thought. That Popper has failed to distinguish these will be apparent to any who compare Mannheim’s [19] with Popper’s caricature of it in [27]. 1S ({9]: 82-90), another example of this kind of thing is to be found in ([22]: 303), and the notion of “Hempelian” general laws implicit in Donagan’s criticism of them ({6]: 161) is of like character.” For a more serious consideration of the character of the laws that would explain Gardiner’s example see [25]: 583ff.

18 EVIDENCE AND EVENTS IN HISTORY point of view. In sum, the kind of example that philosophers have made up are too specific to be interesting, and this seems to be because of their view that laws in history would have to be fairly direct generalizations of what they explain. This, of course, is never explicitly stated,

but it is clearly implicit in the examples they make up. Since each historical event is in many, if not most, ways qualitatively different

from any other, it is no wonder that the program implied in the view that general laws are required or presupposed in history—as everywhere else—seems not worth taking time about. For a more serious consideration of the character of laws in history, we may turn to the work of the late Frederick John Teggart. An intensive study of records bearing on certain wars and uprisings in antiquity leads ‘Teggart to the conclusion. ... that between 58 B.C. and A.D. 107 barbarian uprisings in Europe were preceded invariably by the outbreak of war either on the eastern

frontiers of the Roman empire or in the “Western Regions” of the Chinese. Also it has been found that the invasions which followed disturbances in the Roman East occurred both on the lower Danube and on the Rhine, whereas the uprisings which followed disturbances in the T’ien Shan affected only the upper Danube. Further, there were

no uprisings in Europe which were not preceded by the respective disturbances in the Near or Far East, and there were no wars in the Roman East or the ‘T’ien Shan which were not followed by the respective

outbreaks in Europe... . ([31]: 236).

Teggart claims that there are no known exceptions to these conclusions and “in spite of the unsystematic character of the Roman sources and their imperfect preservation,” evidence is lacking for the correspondence of events in only a small number of cases ([31]: 237). The application of Teggart’s discovery is very limited, indeed, and we shall discuss this problem shortly. But two things may be remarked about it at once. It is general without being merely the generalization

of a particular event. And it is not trivial. It is only in light of it— or something rather like it—that it makes sense to say that evidence of a determinate sort is missing, in this case evidence that would establish that the uprisings we conjecture because of Teggart’s hypothesis, but know nothing more about, might be spoken of more confidently.'® Too, should evidence turn up of a war or uprising similar '© T.e., evidence an explanation of which would be an historical event of the kind that would confirm Teggart’s hypothesis. Of course, we could conceivably find evidence of an opposite kind, the explanation of which would be an event which would disconfirm the hypothesis.

EVIDENCE AND EVENTS IN HISTORY 19 to those characterized by Teggart but somewhat earlier or later than

the span of time covered by his investigation, it would surely be reasonable to try to find a war or uprising of the kind that Teggart correlates with it. There are obvious limitations to ‘Teggart’s hypothesis. It cannot, for example, be used in explaining wars or uprisings much before or

after the span of time Teggart mentions, nor in explaining wars or uprisings which took place during that period but in other parts of the globe. It is, in other words, restricted in its scope both spatially and temporally. In this, however, it resembles Kepler’s account of plan-

etary motion, the scope of which 1s restricted spatially to our solar system and temporally to the period of time between the beginning and ultimate end of that system. Related to its limitation of scope is the fact that its formulation is likely to evoke a feeling that the hypothesis cannot stand by itself, but rather that something ought to be done to make it intelligible. If theories of more general character could be introduced into this field, not only would our perhaps reasonable hesitation about accepting Teggart’s hypothesis be overcome, its limited scope would cease to be an issue at all.'’ Here, too, we see a resemblance or parallelism of sorts between Teggart’s account and Kepler’s, the more general laws of Newton doing for the latter what more general laws in the social-historical field might do for the former. Teggart was himself fully aware of the problem, as the following paragraph from his study makes clear. The establishment of correlations in historical events does not, indeed, solve the problem of the recurrent invasions in central Europe, but it does define this problem in new and explicit terms. How, then, are we to account for the circumstance that Roman wars in Armenia occasioned barbarian uprisings on the lower Danube and the Rhine, and that Chinese wars in the ‘Tien Shan occasioned similar outbreaks in

Hungary? Now, stated in this form, the problem assumed an unexpected aspect, and in an embarrassing manner called for inquiries of a sort not represented in the chronological data previously compiled. In other words, the focus of inquiry shifted from the critical examination of dates to an explanation of the possible linkages, considered geographically, between peoples situated in Armenia and in Rumania, separated

as they were by the Black Sea, and between peoples situated in the Tien Shan and in Hungary, separated as they were by Dzungaria, '’ Failure to discover such more general laws would not, however, deprive Teggart’s

account of such independent worth as it may have.

20 EVIDENCE AND EVENTS IN HISTORY Western Siberia, Russia, and Poland... Suffice it to say that the outcome of this phase of the undertaking was the conclusion that the correspondence of wars in the East and invasions in the West was due to interruptions of trade ({31]: IX, his italics; cf. pp. 239f).

Teggart’s suggestion here is that the various barbarian peoples involved depended upon contact with their civilized Roman and Chinese neighbors in order to procure needed supplies and that these sources were cut off during times of war. This forced them to make

uprisings against or incursions into the settled territories of the Roman Empire. This explanation is still formulated in terms about as restricted in application as the original hypothesis discussed above, yet if it has serious explanatory force it would have to be—as, surely, it can be—formulated so as to apply more generally. Analogous economic relationships between large and civilized peoples and barbarlans may be found elsewhere, and the newer explanation could con-

celvably be tested against a body of data wider than that provided by Teggart’s own study. And, clearly, if the more general proposition tends to be confirmed, the intelligibility of the narrower hypothesis Is given some degree of additional support.

It is impossible to anticipate the degree of generality that may ultimately be attained in the formulation of historico-social theories, but we may speculate about certain possibilities even if these remain,

in the end, unrealized. Even in the more general of the theories mentioned above, the restriction in its range of application 1s appar-

ent. It deals, or would if anyone troubled to formulate it, with a highly determinate kind of event, and if it may be applied to other times and places than those with which Teggart is concerned, it does not begin to deal with the wide variety of socio-historical phenomena

in which we are interested. May we not conceive of at least the possibility of formulating social theories which are less specific in character, so general that in fact no specific kind of event is mentioned in their formulation? We may begin by considering another social theory, also rather specific in its application yet, like ‘Teggart’s, in no way trivial. G. P. Murdock’s book, Soczal Structure, is concerned to explain how systems

of kinship nomenclature develop and change. It would take us too far from our present problem to discuss here the precise character of this problem in cultural anthropology, and the way in which Murdock deals with it. It will be sufficient to observe that he thinks that there are basically four aspects of the social system which are relevant to

EVIDENCE AND EVENTS IN HISTORY 21 the development and change of kinship systems: rules of residence, of descent and of marriage, and the form of the family; whether or

not this is adequate, need not concern us. The problem is to see how the various forms that these four can have affect kinship systems. Before presenting his evidence for this, but after a detailed discussion of the phenomena involved, Murdock offers what he considers to be a very general formulation of his theory: The relatives of any two kin-types tend to be called by the same kinship terms, rather than by different terms, in inverse proportion to the number and relative efficacy of (a) the inherent distinctions between them and (b) the social differentials affecting them, and in direct proportion to the number and relative efficacy of the social equalizers affecting them ({21]: 138; original in italics)."*

In the context of Murdock’s investigation, “inherent distinctions” and “social differentials” refer to social factors which lead to the distinguishing between kin-types; “social equalizers” refers to factors which result in distinct kin-types acquiring similar roles.'? But surely these notions have applicability to a much wider range of social phenomena. One may suspect that social scientists could provide long lists of situ-

ations which might be characterized in terms of such ideas. ‘Then why, one may wonder, might we not expect that all of this might be unified by some system of theory in which these notions would be used and within the formulation of which such more restricting terms as “kin-types” and “kinship terms” do not appear? Presumably such a more general theory would explain the more limited theory constructed by Murdock—or if not his something like it—and much else besides. I do not say that this will ever actually be done, nor have I any idea at all concerning what such a theory would be like. But I can think of no argument that leads me to think it impossible, nor

must it of necessity be so obviously false as the more general of Scriven’s three examples. '8 An instance of two kin-types called by the same kinship term in our society is mother’s sister and father’s sister. Whether or not Murdock’s use of “inverse proportion” and “direct proportion” is to be taken seriously, need not concern us here. '9 In a matrilocal society, say, the sister of a person’s mother may often be able to do many things for or concerning him that his own mother might do, and this social equalization may result in one term being used both for “mother” and “mother’s

sister.” In such a society, one’s father’s sister lives elsewhere and does not do the things in question. Thus, between mother’s sister and father’s sister there is social differentiation, they are referred to by different kinship terms, and our category aunt does not apply.

22 EVIDENCE AND EVENTS IN HISTORY As long as we are speculating about unrealized possibilities, one may wonder about the possibilities of general theories using the concept of anomie. Since Durkheim introduced this notion into sociology (cf. [7])

social scientists have characterized a wide variety of social and psychological phenomena as anomic, that is to say, reflecting a breakdown of norms, the realization that expectations cannot be fulfilled or that one’s situation cannot be structured in terms of the norms and values to which one has become habituated. Sebastian De Grazia’s The Political Community will provide the interested reader with many

examples of this sort of thing, entirely apart from whether or not one subscribes to his point of view. If it should prove possible to unify theoretically the wide vanety of anomic phenomena, presumably we would have a theory of great generality within the formulation of which no mention would be made of particular kinds of anomie, just as we suggested for the ideas discussed in the previous paragraph. We are now in a position to consider the circumstance under which an historical event may meet the requirement that it offer an adequate

explanation of historical evidence and yet have to be rejected on other grounds. To be sure, an event may be rejected because its rejector thinks that the evidence it explains is properly part of a wider configuration of evidence, a view perhaps not shared by its original formulator and others who accept it. But considerations such as this

have been dealt with above, and I want to end the present section with consideration of another point. Should the possibilities concerning which we have just been speculating ever approximate realization, it

may well be the case that certain logically possible historical reconstructions, which seem admirably to account for the evidence, must be rejected on the ground that their acceptance would be in conflict with acceptance of fairly well-established theories. Thus, we might have theoretical reasons to choose between two or more conflicting reconstructions, reasons which we do not have today. One can imagine

that such reasons would be of great value for research in ancient history where evidence is sparse and conclusions somewhat tenuous.

It may be worth noting that even in the present state of social theory there is at least one known instance of this kind of possibility.

Murdock notes that while it is logically possible for an amitolocal rule of residence—one according to which upon marrying the couple settles down with the husband’s father’s sister—to exist, in point of fact none has ever been discovered ({21]: 71). Subsequently, Ward Goodenough argued that there are good reasons for believing that

EVIDENCE AND EVENTS IN HISTORY 23 this logical possibility can never actually be realized, that the realization of the rule is nomologically impossible [13]. ‘Thus, if there were scraps of data—perhaps owing to incompetent ethnography—which might be explained by supposing either that the society in question had an amuitolocal rule of residence or some other determinate rule,

if we could reject the former on nomological grounds, the claim of the latter would be strengthened.

V. Subjectwe Intentions and the Explanation of Events

In this final part of the paper, I should like to consider the view that historical explanation, at least sometimes, makes reference to the intentions of an agent, an historical person of importance ([8]: ch. V). It would be far, indeed, from my purpose to suggest that the decisions of such persons are not important and even intrinsically interesting. But I do intend to say that knowledge of this kind cannot be offered in explanations of historical events. It would explain only how

it was that the agent in question initiated that chain of happenings

which is the event. At the very least, the notion that an event is explained by reference to the subjective intentions of the agent presupposes that the best laid schemes 0’ mice an’ men do not gang aft agley. If one cannot explain the defeat of the Spanish Armada by reference to the subjective intentions of its commanding officer, I do not see how one can expect to explain the victory of the English in terms of the subjective intentions of its commander. Such subjec-

tivity may well enter into an account of the strategy which each officer sought to carry out, and a student of naval tactics might be very much interested in determining what led each of them to make the decisions he made. But having arrived at the end of this inquiry, our student of naval tactics has still not offered an explanation of the course the battle took. The notion that an “historian’s explanation... will involve a detailed examination, maznly in rational terms, of the activities and motives of countless individuals and groups” ([8]: 142, italics added), is clearly a confusion of psychology and history, the kind of confusion one associates with a theory called “methodological individualism.””° *© See [12]. While methodological individualism purports to explain social phenomena in terms of individuals, it is hard to tell these days just what its most

24 EVIDENCE AND EVENTS IN HISTORY Such a view, with its inability to distinguish between the individual and the non-individual sociocultural content and context of his life and experience, is entirely incapable of appreciating the possibility for developments of the kind speculated about in the previous section

and can never do more than describe the behavior of individual behavers. In our present context, such a view confounds the initiation by an individual of a course of action or event with the event itself. Actually, this last is but one of two possible interpretations of the view in question, but inasmuch as the other is so improbable it seems reasonable to suppose that none of the explanation-by-reference-tosubjective-intentions theorists holds it. It seems that according to some theological cosmogonies, the creation is never a fully completed event after which the universe continues on its way, each subsequent state of which being accounted for by reference to the laws established by

the creator and a state description of the universe at its previous state. Rather, on this view, during each successive stage it is sustained by the care of the creator, who is in some way taken to be the causal agency of the changes within it. It seems to me that the only alternative to what I have said about the view that historical events may be explained by reference to subjective intentions in the previous paragraph, is an historical analogy to this cosmogonical active recent proponent, Mr. J. W. N. Watkins, is trying to claim. From a recent note on the subject ([35]) one may suspect that his individualism is non-methodological and his methodology non-individualistic. He insists that only individuals exist, which is an ontological and not a methodological point, and that social explanation can be in terms of anonymous individuals displaying anonymous dispositions. But these individuals and dispositions are in no way individualistic, being entirely

social in character; cf. my [12]: part 2). Only a perverse logic could take my remark, near the end of the aforementioned paper, that “there is no science of the anonymous,” as denying that in the formation of theories we do not make reference to actual particulars, a denial which would seem to make me a methodological individualist. If on Watkin’s methodology we can explain only in terms of individuals, and if he thinks we can explain by reference to anonymous individuals and anonymous dispositions, then he treats the latter as being not conceptual or theoretical but as real entities. The point of my remark, as is obvious in its context, is precisely to insist upon the conceptual rather than individual character of the anonymous. Watkins likes to say that he has consistently defended the same position over the years and that I keep missing its point, yet any comparison of this recent note with his characterization and defense of “individualistic ideal types” in [34] will show that this is mistaken. If attempts to treat his anonymous individuals and dispositions as individualistic, in the way his preferred ideal types are taken by him to be, are not properly representative of his views, then perhaps he is no longer a methodological individualist even if he insists upon the title. One may venture to suggest that his words are like the handiwork of Socrates’ ancestor Daedalus.

EVIDENCE AND EVENTS IN HISTORY 29 theory. One must imagine that the initiator of the action “keeps control” of it throughout the entire course of its duration. This theory is even more complicated than the theological one, however, because

presumably some way must be found to harmonize the “control” exercised by the initiators of overlapping events—such as the winning by the English and the losing by the Spanish of the battle of the Armada. Since no one wishes to hold such a position, we need consider it no longer, and will simply assume that it is the failure to distinguish between the initiation and continuation of an event that is in question. That which follows from the initiation of the action follows not merely because the action was initiated, but because it was initiated under the particular condition which obtained at the time. Initiating actions of such character, quite apart from the subjective intentions of the

initiator, in contexts of such character have certain results, and in any given case the results constitute the event the historian describes or constructs. But to explain it, we are required to know considerably more than the motives, intentions or dispositions of an agent. A full explanation must presumably await the satisfactory formulation of socio-historical theories.

It may be noted that the view of this paper does not make the initiation of the course of action by an individual agent irrelevant. Nor does it require that the individual initiation be explained away so that it cannot be taken to be a genuinely new departure in the socio-historical scene. Our view is only that the new departure becomes

an element in a context, and its effects within that context are amenable to explanation within a possible system of socio-cultural theory. Such theory would be non-individualistic, but it would not be historicist

in Popper’s sense. The theory could explain what happens in history, but it would not require that what happens happens of necessity. Our explanations are hypothetical, and there is no reason why our “if” clause cannot, in part, assert that such-and-such was done by so-and-so. Our explanation of so-and-so’s doing such-and-such would have to take account of motives and intentions. Our explanation of the historical event which follows from the initiation would not.*! The construction of the event is often, perhaps mostly, quite ?! Though made in a context different from ours, some remarks of Hempel seem apposite at this point: “This kind of information about purposes and beliefs might even serve as a starting point in explaining a self-regulating feature in a human

26 EVIDENCE AND EVENTS IN HISTORY independent of the account of its initiation, and could be explained even if on the basis of available information the intention of the initiator could not be explained or even determined.”

REFERENCES

[1] Banron, M., White and Coloured, the Behaviour of British People Toward Coloured Immigrants, London, 1959.

[2] Becker, GC. L., “What are Historical Facts?” in H. Meyerhoff, ed., The Philosophy of History in Our Time (Doubleday Anchor Book), New York, 1959, pp. 120-37. [3] BeLxin, S., Philo and the Oral Law, Gambridge, Mass., 1940. [4] Biocu, M., The Historian’s Craft, New York, 1953. [5] De Grazia, S., The Political Community, Chicago, 1948.

[6] Donacan, A., “Explanation in History,” Mind, vol. 66, 1957, pp. 145-64. [7] Durkuerm, E., Suicide, Glencoe, IIl., 1951. [8] Dray, W., Laws and Explanation in History, Oxford, 1957. [9] Garpiner, The Nature of Historical Explanation, London, 1952. [10] Gasrer, T. H., Thespis: Ritual, Myth and Drama in the Ancient Near East, New York, 1950. [11] Goxpsrein, L. J., “A Note on the Status of Historical Reconstructions,” Journal of Philosophy, vol. 55, 1958, pp. 473-79. [12] ——-, “The Two Theses of Methodological Individualism,” British Journal for the Philosophy of Science, vol. 9, 1958, pp. I-11. [13] GoopENnoucH, W. H., “Amitalocal Residence,” American Anthropologist, vol. 53,

1951, pp. 427-29. [14] Hempezt, C. G., “The Logic of Functional Analysis,” in L. Gross, ed., Symposwum on Soctological Theory, Evanston, Ill., 1959, pp. 271-307. [15] Jounson, A., The Historian and Historical Evidence, New York, 1926. [16] Lanctiois, C. V., and Seicnosos, C., Introduction to the Study of History, London, 1898.

[17] MacIver, A. M., “Historical Explanation,” in A. Flew, ed., Logic and Language (second series), Oxford, 1953, pp. 187-203. [18] ManpetBaum, M., The Problem of Histoncal Knowledge, New York, 1938. [19] Mannuei, K., Man and Society in an Age of Reconstruction, New York, London, 1950.

[20] Meap, G. H., The Philosophy of the Present, Chicago, London, 1932. {21} Murpock, G. P., Social Structure, New York, 1949.

artifact. For example, in an attempt to account for the presence of the governor in

a steam engine, it may be quite reasonable to refer to the purpose its inventor intended it to serve, to his beliefs concerning matters of physics, and to the technological facilities available to him. Such an account, it should be noted, might conceivably give a probabilistic explanation for the presence of the governor, but it would not explain why it functioned as a speed-regulating safety device: to explain this we would have to refer to the construction [noun, not verb—L. J. G.] of the machine and to the laws of physics, not to the intentions and beliefs of the designer” ({14]: 299).

2 As Banton has said in another connection, “We can never with certainty infer the actor’s intention from observing his behaviour” ([1]]: 17); cf. [32]: 23, 78 and ch. vi.

EVIDENCE AND EVENTS IN HISTORY 2/7 [22] Nace, E., “Determinism in History,” Philosophy and Phenomenological Research,

vol. 20, 1960, pp. 291-317. [23] OrsterLtey, W. O. E., The Sacred Dance, New York, 1923. [24] Premrer, R. H., Introduction to the Old Testament, New York, 1948. [25] Prrr, J., “Generalizations in Historical Explanations,” Journal of Philosophy, vol. 56, 1959, pp. 578-86. [26] Popper, K. R., The Open Society and its Enemies, Princeton, 1950. [27] ———, The Poverty of Aistoricsm, Boston, 1957. [28] Renier, G. J., History: Its Purpose and Method, Boston, 1950.

[29] Scriven, M., “Truisms as the Grounds for Historical Explanations,” in P. Gardiner, ed., Theortes of History, Glencoe, Ill., 1959, pp. 443-75. [30] Social Science Research Council, The Social Sctences in Historical Research (Bulle-

tin 64), New York, 1954. [31] Teccarrt, F. J., Rome and China, A Study of Correlations in Historical Events, Berkeley, 1939.

[32] ——-, Theory and Processes of History, Berkeley, 1941, paperbound, ed., 1960. [33] Torrey, C. C., Pseudo-Ezekial and the Onginal Prophecy, New Haven, 1930. [34] Watkins, J. W. N., “Ideal Types and Historical Explanation,” British Journal for the Philosophy of Science, vol. 3, 1952, pp. 22-43. [35] ——, “Third Reply to Mr Goldstein,” British Journal for the Philosophy of Science,

vol. 10, 1959, pp. 242-44.

DISPOSITION CONCEPTS AND HISTORY Alan Donagan’s “Explanation in History” (Mind, 1957) attempts to oppose to the theory that in history explanations presuppose general laws—the so-called “Hempelian theory”—the proposition that what historians in fact do is explain by use of law-like statements formulated in terms of what Donagan, following Ryle, calls disposition concepts. I am far from being convinced that Donagan’s critique of the Hempelian theory is adequate to its demolition since he nowhere

faces the fact that the theory 1s an application to history of the Humean—taken broadly—epistemology and not a reconstruction of what historians do. As such, it should be attacked on philosophical grounds and not on the irrelevant ground of historiographic practice. Too, 1t seems to me that the Ryle-Donagan rejection of the claim that dispositional statements rest logically on general law statements of the usual kind on the ground that we learn to apply dispositional statements to individuals before we learn to express general correlations

(p. 158) is entirely beside the point, since we are concerned with a problem in the logic of explanation and not in the psychology of learning. But I do not intend to discuss these important matters here. Rather, I wish to devote this short note to a consideration of certain difficulties about the application of dispositional statements to historical

explanation entirely apart from whether or not the Hempelian theory is judged to be tenable.

In the course of his discussion, Donagan recurs to a statement from J. A. Wilhamson’s The Evolution of England (Oxford, 1931): “The

Norsemen and Danes who sailed south to the Insh Sea and to the shores of the English Channel were plunderers first and settlers by an afterthought. Like the early Anglo-Saxons, they came to sack a civilised land, and only when they had stolen all they could get did they think of occupying its soil” (Donagan, p. 151). This is said to explain the Scandinavian invasions of Britain in the first half of the ninth century. Donagan thinks he shows that this explanation is not capable of being made compatible with the Hempelian theory by generalizing the particular event to be explained—“all men, or all Norsemen and Danes, and perhaps Anglo-Saxons too, are plunderers first and settlers by afterthought”—and observing that it would

DISPOSITION CONCEPTS AND HISTORY 29 be “fantastic” to think that Williamson would consider his explanation to be weakened by the discovery of exceptions to it. Though it does not seem to me at all necessary that the general laws implicit in historical explanations be immediate generalizations of the historical events they explain, it will take us from our present task to pur-

sue the point here. More pertinent will be an attempt to explore what is involved in Donagan’s view that Williamson’s statement “1mmediately implies” a law-like or dispositional statement to the effect

that if the Danes in question had sufficient opportunities to plunder a given territory they would not settle in it (p. 156). There are, I suppose, at least two ways to discover what Williamson’s statement implies—in whatever sense is suitable to the present context—one by means of logical analysis and the other by ascertaining just what Williamson himself intends that it imply. We might usually wish to distinguish between what a writer intends to say and what in fact his words may actually mean, but we have seen that

for Donagan one reason for not treating the events described in Williamson’s book as instances of a general law is that it would be “fantastic” to think that the disconfirmation of what in Donagan’s view that general law would be could have any affect on Wilhamson’s acceptance of his own explanation. And yet, I suppose that Donagan and other critics of the Hempelian theory have little option here, for

once they allow that the intentions of historians need not be given a position of primacy in our determination of what their explanations mean, what seems to be one of their most potent arguments is very

much undercut, namely, that historians do not explain the way Hempelians say they should. The question, then, emerges as to whether or not Williamson can possibly be understood as intending to say what Donagan says his statement means. And one may wonder if it is really credible that an historian discussing events in ninthcentury Britain wants to talk about the dispositions—mental or psychological, I should suppose—of particular, individual Danes concerning whom he knows virtually nothing. It is not about Danes and their dispositions that his evidence leads him to speak, but rather about certain things that those ancient Danes did, namely, plunder rather than settle. If Williamson had some independent way of knowing about the relevant dispositions of those Danes, he might have wished

to refer to them in order to explain the events in question. Since he cannot be expected to have such knowledge, one must disagree with Donagan’s interpretation of his statement. An alternative would be

30 DISPOSITION CONCEPTS AND HISTORY to trivialize it entirely and say that since such-and-such was done by the Danes they must have been disposed to do it. But this can hardly pass muster as an explanation of their action. Thus, it appears that Williamson’s statement is not at all an explanation of Danish plun-

dering and not settling, but is, rather, a summary or general statement to the effect that Danes did in fact plunder when they could and seemed to settle only as a last resort.Since Donagan makes so much of the lesson he thinks is implicit in his quotation from Williamson, perhaps it is worth digressing a moment from the main purpose of this note—which is to impugn the conception of historical explanation defended by Donagan—and say something additional about Williamson. To begin with, it may

be worth noting that the kind of general summary statement—not explanation—that I take the quoted statement to be appears elsewhere in his book, and one may suspect that it is simply a stylistic device, one which is not unhelpful to the reader. ‘Thus, to add one example, a few pages after the quoted statement, Williamson writes, “,.. the Anglo-Saxon was no town-dweller. His fields were his home. War swept over them, but if there were any survivors they returned to the plough. ‘There was nothing else to do” (p. 50). It would certainly not do to see here any reference to a disposition of Anglo-Saxons to

return to the plough in the sense that this would explain the fact that they did so return. ‘That they did as a rule return is all that the quoted statement reports. What is oddest of all is that Williamson should even be cited by a writer who wishes to say both that the Hempelian theory is inapphcable to historical explanations and that historians share this opinion. Whatever one wants to say about Williamson’s announcement in the preface of his book that he has a penchant for geographical

explanations (p. v), he is the author of the following paragraph (p. 930):

Why were these men, of similar blood and speech to the Saxons of England, so utterly different in instinct and character? The answer 1s that though heredity counts for much, environment counts for more. The way of life shapes the people who live it. We may see this in nations as in individuals. The English stock in North America and Australia are of different character from the English stock at home. If three brothers are born on a farm, and one goes to sea, another to work in a city, whilst the third stays on the land, by the time they grow old they will have less in common with one another than with strangers in their own several callings. This is a vital fact in the evo-

DISPOSITION CONCEPTS AND HISTORY 31 lution of a people, and it has its bearing upon imperial federations and

leagues of nations. There is a limit to the unity attainable by such means. It must at best be a unity without uniformity, based upon a sentiment needing deep root.

Writing as an historian, not a logician, Williamson makes no effort to separate his general law and the conditions to which he apples it, but both of these elements of what Hempel calls the explanans of an explanation are present. And Williamson goes so far as to use his general law not only for the purpose of historical explanation, but also, at the end of the paragraph, to make some observations as to the future solution of a present problem. Williamson is clearly an historian whose approach to historical explanation does not begin to accord with the views of Donagan. What these views are, we have already seen. Donagan holds that historians explain by reference to the dispositions of individuals. Against

this view, I wish to offer two considerations. The first has already been mentioned in our discussion of Donagan’s quotation from Williamson, namely, that the historian almost never has access to those dispositions. His records permit him at most to offer reconstructions of events but give him no special information as to the dispositions of those who were involved in the action. It may be that upon rare occasion personal papers or memoirs do give information of this kind about certain leading figures, but this is not at all the usual situation. And even if we had such information about leading Danes—-as we do not—that would hardly permit us to interpret statements about the Danes is terms of thezr dispositions. ‘That Danes were

disposed in the way Donagan—surely not Williamson—says is an inference from their actions and one which serves no explanatory purpose of Williamson. If we were concerned to deal seriously with dispositions——not in the trivializing way I noted above—we would want to be able to specify dispositions in psychological terms rather than in terms of the outcome of their applications. And this could be

done only if we had the kind of information about the particular people who were, say, those ancient Danes such as we never shall. This leads to the second of my two points. It is one thing to say that dispositions to plunder rather than to settle result in Danes plundering and settling only as an afterthought. But what would be entailed

by a characterization of the Danes in terms of psychological dispositions formulated without reference to the particular historical consequences that their having such dispositions is alleged to have effected?

32 DISPOSITION CONCEPTS AND HISTORY Knowledge of such dispositions would be relevant to our understanding

of the individual’s particular behaviour. If the individual is an historically important person, we would say that sometimes his behaviour was such as to initiate an historical action or event. That his particu-

lar action was initiatory in the way it was, is, it may be presumed, owing to the context in which it was performed. But such an action in such a context would have had that outcome entirely apart from the particular disposition of the given actor or historical figure. Thus, it emerges that knowledge of psychological dispositions, however relevant to our understanding of the individual’s action and, indeed, of the initiating performance of the historically important person, is entirely irrelevant to our understanding of historical events.

THEORY IN HISTORY*

I. Introduction

It is obvious enough to those who read what philosophers writing in English have to say about problems of history that the main focus of interest and point of contention is the nature of historical explanation. And even within that general theme, interest is confined more narrowly to the question of whether or not explanations in history rest upon general laws or theories, “covering laws” in the increasingly-used term of William Dray. On the one hand, we have the view, often associated with the names of Hempel and Popper, yet including among its supporters such writers as Brodbeck and Gardiner, and actually rooted

in writings earlier than those of any of these, according to which explanation must always presuppose general laws for only these can warrant an inference from one particular state of affairs to another particular state of affairs. And on the other, we have writers such as Donagan, Dray, Gallie, and Scriven offering a variety of arguments to the opposite effect, and in particular, making much of the claim

that historians do indeed explain or offer explanations which are complete and reasonable, yet make no use of general laws. Some may think that the debate has grown stale and tedious, and that essentially the same arguments—though occasionally with greater subtlety—are repeated time after ttme. And some may deem it a matter of regret that neither side seems able to reach the other, or if you like, that neither side seems amenable to persuasion. But surely,

most who follow the debate have no doubt that both sides are attempting to cast light upon the same problem, that of how historical events are to be explained. But this view of the matter is mistaken, and far from being engaged in a discussion of a common problem, each of the sides to the “debate” is concerned with something quite different from that which

engages the attention of the other. One side, those who argue in defense of the claim that explanation in history must rest upon general * This is a slightly revised version of a paper which was read to the Boston Colloquium for the Philosophy of Science (November 1, 1965).

34 THEORY IN HISTORY laws, are concerned with methodology of inquiry and the logical problems of validating claims to knowledge. For the most part, these interests do not arise out of preoccupation with problems of historical

knowledge, and I think it can be shown that any number of the contributors to the development of the general-laws point of view are not particularly sensitive to the epistemological idiosyncrasies of history. Rather, proponents of the general-laws point of view tend to ascribe to history a conception of explanation in general, and while all of them would acknowledge that the particular features they look for in a logically sound or complete explanation are not always, or even usually, to be found in the explanations that historians actually offer, they nevertheless argue that in principle these might be sup-

plied. If not, the defenders of the general-laws position would not allow that a proffered explanation could pass muster as an explanation. In order to show that their position is neither arbitrary nor unreasonable, some of these writers construct examples to illustrate what they take laws in history to be, but these examples are almost always immediate generalizations of the given particular explanation.

Thus, if some historian—or philosopher in quest of an example— suggests that some particular a is explained by some particular }, the

general law allegedly called upon is “a-like events are caused by b-like events” or “a-like events are usually caused by d-like events” or some such thing. And this, unfortunately, leads historians to wonder

about the relevance of such discussions to their own work. In addition, these examples give a somewhat distorted idea of the purpose of historical inquiry, the relation of historical research to other kinds of social inquiry, and the function of theory in the realization of the goals of all such research. I shall talk about all this in what follows, but before I do so, I want to make some passing comments on the other side of the “debate” over the nature of historical explanation.

Strange as it may seem, the writers on the other side are not interested in history either. They are, rather, very much taken with the elucidation of modes of discourse and, for whatever reasons, have been attracted to the discourse of historians. ‘The constant preoccupation of the writers who oppose the conception that explanation in

history presumes general laws is with what they find in historical narratives. It seems assumed that any historical narrative is complete

in itself in every possible respect, as if the given narrative has and could have no relationship to other matters. And, more particularly, in dealing with the logic of explanation in history, these writers are

THEORY IN HISTORY 39 of the view that in any given case we are to be limited to the idiosyncratic intentions of the given historian—presumably because he is

the author of the narrative to be elucidated. Since the well-known book of William Dray [2] is likely the best known contribution to this point of view, I shall draw several examples from it, but there 1s evidence of this attitude in the writing of others as well (see the papers of Alan Donagan, L. B. Gallie and Michael Scriven in [4)]).

On p. 28, in rejecting the possible claim that “a person cannot mean anything by the explanation if he denies the law,” Dray says that “The historian has denied that he arrived at his explanation by means of a law,” as if to suggest that if the given historian did not himself derive his explanation by deduction using general laws such laws are irrelevant to the explanation. On p. 36, he talks about an argument which “may lead to misunderstandings of the structure of typical explanations in history;” on the following page, he contends “that the historian, when he sets out to explain the French Revolution is just not interested in explaining it as a revolution ...; he 1s almost invariably concerned with it as dofferent from other members of its class,” and on p. 57, in criticizing “The Humean assumption that nothing but ‘regularity’ can justify a ‘because’” and the consequence this has for the arguments of Gardiner, he writes: If the historian does not use a precise ‘rule’, then a vague one must be found; if no universal law is available, then a qualified one must have been assumed. The alternative which is too much to accept is that, in any ordinary sense of the word, the historian may use no law at all.

It seems clear that in each of these passages, the general-law point of view is rejected as not being an account of what is actually written in works of narrative history, and that “the historian” of whom Dray

writes is one who makes use of a certain mode of discourse—not one whose work raises certain epistemological problems or whose explanations lead to further questions of fact and theory. And thus, when, on p. 58, he says of possible claims that the covering law is a sufficient condition of an explanation that it, too, “is artificial and misleading,” it seems reasonable to take it to mean that it fails as an elucidation of the historian’s discourse. On p. 76, Dray says “Hempel’s formulation begins by laying down the logical structure of explanation as he believes he finds it in physics,” which would suggest that he takes Hempel’s task to be the elucidation of the language of physicists, rather than attending to the philosophical questions that physics raises.

36 THEORY IN HISTORY It is hard to believe that Hempel would himself subscribe to this conception of what he does, but one need not be surprised that Dray should think it. (All the italics in this paragraph are Dray’s) In sum it should now be clear why I say that writers like Dray are not actually interested in history—the discipline—and its problems, but only in the language in which histonans formulate their finished product.’

Il. Some general laws

If the elucidators of modes of discourse seem to suggest that there are no problems of history or historical knowledge and by implication treat history as a literary genre only, it does not follow that the view of that discipline implicit in some of the writings of their opponents is more useful. I do not think it profitable to seek to ascribe to history

any particular set of goals, and it is clearly the case that historians are engaged in a variety of activities. Yet zmplicit in what is found in

the statements of any number of those who subscribe to the view that historical explanation must always presuppose general laws is a conception of the historian’s task which is both restrictive and untrue to the actual practice of that discipline. I do not suppose the writers I have in mind would actually wish to maintain such a view, but I think that it does emerge from what they write, particularly from the kinds of example they make up when they seek to defend their point of view. I think it best to present some examples of what philosophers have supposed might serve as general laws applicable to explanation in history and then draw from them what I think they imply.

, I shall first present examples used by Hempel, Gardiner and Nagel, all of whom subscribe to the general-laws thesis; then one from Popper,

who shares their view on the matter, but, for reasons which are not relevant here, thinks that the laws in question must and ought to be trivial and so might not be bothered by some of the consequences I

' Cf. May Brodbeck’s distinction between “language as communication” and “language as description” in Minnesota Studies in the Philosophy of Science, vol. 3, 1962, pp.

236ff. For a rather extreme consequence of the preoccupation with modes of discourse, see Kai Nielsen’s assertion that the issue is about, not how historians explain, but how they use the word “explanation” (in Sidney Hook, ed., Philosophy and History, New York 1963, p. 319).

THEORY IN HISTORY 37 shall be drawing; and, finally, one from Michael Scriven, who does not subscribe to the general-laws point of view at all but who nevertheless seems to agree with those who do about the character of laws in history if they were to be had. In the course of what is most likely the best-known individual expression of the general-laws thesis, Hempel says the following: Consider, for example, the statement that the Dust Bowl farmers migrate to California ‘because’ continual drought and sandstorms render their existence increasingly precarious, and because California seems to them to offer so much better living conditions. ‘This explanation rests on some such universal hypothesis as that populations will tend to migrate to regions which offer better living conditions [7].

Hempel goes on to write that it would “be difficult accurately to state this hypothesis in the form of a general law which is reasonably well confirmed by all the relevant evidence available,” and it is clear enough from his discussion that he puts the hypothesis forward as an

example of the sort of thing he has in mind and is not to be held responsible for 74. Yet one may wonder just what evidence would be

“all the relevant evidence available” to test a “general law” which seems to be little more than a direct generalization of the historical event it is being called upon—or might be called upon—to explain. As we proceed with our catalogue of examples, it will become clear that all of the proffered laws have this same feature: they simply generalize what they are supposed to be explaining. What this implies about the discipline of history and the goals of historical explanation is precisely what we have to deal with in this present section of the paper. At the beginning of a list of explanations culled from historical writings, Gardiner quotes the following from a work of Charles Seignobos:

Louis XIV died unpopular... having caused France to lose... the incomparable position she had gained by the policy of the cardinals ([5], p. 65).

Later, in another context, he asks: But are we able to interpret... the historian’s explanation of the unpopularity of Louis XIV in terms of the generalization that rulers of nations become unpopular when they pursue a policy resulting in the decline of the country over which they rule? ([5], p. 87).

38 THEORY IN HISTORY This makes it seem that in Gardiner’s view one who held that Seignobos’ explanation rests upon a general law would be committed to a law which was hardly more than a generalization of what

was to be explained. In this case, that a king became unpopular because he pursued policies detrimental to his country is explained by appeal to a generalization to the effect that rulers who practice policies detrimental to their countries become unpopular. That this, indeed, is Gardiner’s view becomes unmistakable a bit further on when he says: Let us return again to the explanation given as to why Louis XIV died unpopular. ‘The positivist interpretation would be that Louis XIV represented a case of the law ‘Rulers are unpopular whenever their policies prove detrimental to the fortunes of their countries’ and that the explanation in question was deduced from this law taken together with circumstances of the case ([5], p. 89, Gardiner’s italics).

Our next example is the sort of general law Ernest Nagel thinks one might call upon in order to justify F. W. Maitland’s explanation of the first Queen Elizabeth’s use of “et caetera” in her title, rather than stating her title in full. According to Maitland, the reason for this seemingly odd way of styling herself was that she hoped thereby to avoid taking a public stand on the issue of the separation of the Church of England from Rome by not proclaiming herself head of the church.* About this, Nagel says the following: Now an examination of Maitland’s discussion shows beyond doubt that

the event he sought to explain does indeed logically follow from the explanatory premises, provided these include an assumption essentially as follows: Whenever anyone acquires a position of great political power,

is faced with an issue fraught with peril, but is required to announce immediately a policy, then such a person will make a statement that is momentarily noncommittal [8].

The assumption seems little more than Maitland’s original explanation with the proper name and specific circumstances removed and replaced by “variables.” The example from Popper takes but one long sentence in his Open Society:

* Fredenc William Maitland, Historian, selections from his writings edited, with an introduction, by Robert Livingston Schuyler, Berkeley and Los Angeles, 1960, ch. 12. The editor, in his introduction to the essay, informs us that Maitland was mistaken in thinking Elizabeth was the first to use the “et caetera” in this way, but this has no bearing on our point.

THEORY IN HISTORY 39 If we explain, for example, the first division of Poland in 1772 by pointing out that it could not possibly resist the combined power of Russia, Prussia, and Austria, then we are tacitly using some tnvial general

law such as: ‘If of two armies which are about equally well armed and

led, one has a tremendous superiority in men, then the other never wins’ [10].

I must admit, that in a sense it is not really fair to include Popper in this list. Popper holds that in the socio-historical field we cannot have comprehensive laws and must, in consequence, make do with trivial ones. The writers mentioned above might hope we could have serious laws, but Popper apparently has no such expectation. ‘Thus, unlike their cases, that his is but another example of a trivial law is

not just a bit of bad luck. But if what I shall soon be saying about the implication of these immediate generalizations for our conception of the discipline of history is sound, then we shall have to take issue with this example as well, and possibly, for the sake of fair play, we might agree to include Popper’s generalization in our list of

bad examples but omit Popper himself from any list of authors of bad examples. Our last example comes from Michael Scriven and is taken from his account of how one committed to the general-laws thesis might set about justifying an explanation of Cortez’ third expedition to Lower California by reference to his cupidity. He offers three generalizations from which one might choose: “All confident wealth-seeking people undertake any venture which offers very great wealth” or “All confident people seeking very great wealth undertake any venture which offers very great wealth” or “All confident people with Cortez’ background of experience, seeking very great wealth, undertake any venture

involving the hazards of this one, which offers very great wealth” [11]. Scriven, to be sure, is opposed to the general-laws thesis, but that is not immediately to the point. What is of interest is that he, too, seems to think that the only kind of generalization we could have in history, if we could have them at all, would turn out to be directly generalized from the explanation it was called upon to justify.

Clearly, if the alternatives offered by Scriven to those who would seek nomological justification for the explanation of Cortez’ expedition

were, in fact, all we could choose from, the formation of theories relevant to explanation in history would be a trivial enterprise indeed. In the course of a paper largely favorable to the claim that theory is relevant to explanations in history, Samuel H. Beer takes a number

40 THEORY IN HISTORY of opportunities to oppose what he calls the “dogma of universality”

[1]. The name of the dogma he seems to derive from what he has read in the writings of some of the authors I have just been mentioning. Thus, he cites Popper to the following effect: “To give a causal explanation of a certain specific event means deducing a statement

describing this event from two kinds of premises: from some unwersal laws, and from some singular or specific statements which we may call the specific initial conditions.” From Hempel, he quotes a character-

ization of a general law as being “a statement of universal conditional form which is capable of being confirmed or disconfirmed by suitable empirical findings.”? The word “universal” figures prominently

in both these statements, and one may suspect that it is this which led Beer to the name he gave the noxious doctrine he wrote to attack. But it is not these cited passages alone which figure in Beer’s con-

ception of the dogma. It is the examples that are used to illustrate universal laws, and only a few lines after he quotes the passages given

above, he produces the examples of Hempel and Popper which are included in our sample. The dogma of universality against which Beer quite properly protests is the view that explanations in history are to be justified by appeal to general laws which are arrived at by the simple expedient of generalizing—or universalizing—the particular explanation to be justified. It is, of course, not unreasonable to suggest

that Beer has not fully caught the drift of the writers he deals with and does not realize that the theories of physics which they have in mind when they offer their accounts of what general laws are, are not at all the simple-minded generalizations of the particular causal explanations to be justified. Rather, what they have in mind are highly

complex theories which may be applied to a wide variety of initial and boundary conditions in order to explain a wide variety of kinds of occurrence. Thus, it would seem that the very writers cited as the source of the “dogma of universality” may not themselves subscribe to it. But Beer thinks they do because he makes the mistake of taking seriously the examples of historical generalization they make up. But before we go on to the next section and its presentation of examples which point to the possibility of theory in history without the dogma of universality, it may be worth our while to take our bad examples seriously enough to wonder what they would imply > ({1], p. 7.) The quotation of Popper is from The Poverty of Historicism, Boston

1957, p. 122; that of Hempel from [4], p. 345.

THEORY IN HISTORY 4} about history if they were not as bad as Beer and | take them to be. I think that one consequence of taking them seriously would be the isolation of the discipline of history. That is, history would be seen to have no particular connection or contact with the work of other disciplines that seek to make sense of human social existence—the social or behavioral sciences—and, indeed, it would even seem that the particular problem of any given historian at any given time of his career would have little if anything at all to do with other particular problems of historical research. I suppose that this seems a most extraordinary thing to say, and for two reasons. First, it surely seems odd to say that proposed examples of general laws have any sort of isolating effect when it is clearly the case that the subsumption of an historical explanation under a general law would seem to have just the opposite effect, namely, to show that it is one of a class of similar explanations and is in consequence not unique or conceptu-

ally isolated. And, second, it is one of the main theses of a wellknown paper of Morton White that there are no essentially historical

laws or theories, and that explanation in history must make use of theories which have been formulated by the special social sciences [15]. This, too, suggests just the opposite of the isolation I see lurking in our bad examples. Yet, it seems to me that the “dogma of universality” that Beer has extracted from these examples has precisely the effect of isolating historical work, because of the way in. which it seems to make each historical explanation something complete in itself. We are concerned,

say, to make sense of certain internal American migrations in the 1930’s which saw farmers from the Dust Bowl settle in California. We discover that the migrants were people whose existence had been rendered precarious by certain natural occurrences and who were in a position to hear about the more promising possibilities in California.

And so, it appears as if we have an explanation for the migration: finding life in the Dust Bowl no longer tenable, the migrants moved west in hopes of improving their fortunes or because they believed that life would be easier for themselves or some such thing. But for

the usual Humean reasons, which we not need review here, it 1s concluded that that which is explained and that which explains it are not essentially bound to each other, since each may be conceived by itself, and so we are required to justify the explanation by appeal to a general law. According to the “dogma of universality” which is suggested by the examples reviewed above, this is done simply by

42 THEORY IN HISTORY generalizing the given case. This satisfies the logical need revealed

by our Humean analysis. But it also isolates the case in that only Cases pretty much like itself—few if any more of which we are likely to turn up——has any bearing upon it or its justification. That seems to be one of the lessons drawn by critics of the general-laws point of

view. hey not infrequently point out that any threat to the general law offered in justification of some explanation would be met by spelling out in increasing detail just what sort of event the general law is applicable to until it becomes entirely clear that only the situa-

tion confronting us to begin with can be explained by its means. The point of the critics is that such “general laws” are not worth bothering about. But what makes them not worth bothering about is

precisely the isolating character that they have. They isolate the explanation and the kind of event it is about from every other possible interest of students of history and society.*

But it is certainly not the case that the only option open to those who would seek nomological justification for historical explanations is to apply the dogma of universality. In the next section of the paper,

I shall present a number of examples of the sort of thing I have in mind taken from the historical literature. Here, I shall briefly sketch what this involves. Clearly, if the dogma of universality is to be avoided, what we are required to do is not quickly to conjure up a general statement which will enable us to deduce the explanation we wish to justify.” Rather, we focus upon the particular explanation, which is, in a way, the starting point of our theoretical enterprise— that this particular situation, state of affairs or events came to be because

of these other particular situations, states of affairs or events—and seek to discover in it what we might think of as its general features. There are no rules for this, though I believe that what this involves * It is interesting to note that those who would elucidate the historian’s mode of discourse would also have an isolating effect upon historical work. Each narrative is treated as a totality in itself, the elucidator being concerned with what the narrator says, and why, in the particular story. This is quite explicitly affirmed by W. B. Gallie, who says: “The fact is that the kind of explanation that I claim to be characteristic of histories cannot be confirmed, or even be preferred against other possible explanations, except via the acceptability of the narrative which it enables the his-

torian to reconstruct or resume. If the narrative has now been made consistent, plausible, and in accordance with all the evidence, if it is the best narrative that we can get, then the explanation that helped us to get it is the best explanation as yet available.” (Philosophy and the Historical Understanding, London, 1964, p. 124.)

> For the distinction between explaining events and justifying explanations, see the first part of the “Postscript” to this paper.

THEORY IN HISTORY 43 will become clearer in the course of the discussion of the next section. But this kind of procedure seems not peculiar to the historical or social studies. Consider the following statement of N. R. Hanson: In the thinking which leads to general hypotheses, there are character-

istics constant through the history of physics from Democritus and Heraclitus to Dirac and Heisenberg. Kepler did not begin with the hypothesis that Mars’ orbit was elliptical and then deduce statements confirmed by Brahe’s observations. ‘These latter observations were given,

and they set the problem—they were Johannes Kepler’s starting point. He struggled back from these, first to one hypothesis, then to another, then to another, and ultimately to the hypothesis of the elliptical orbit ([6], p. 72; his italics).

Hanson then takes us through the stages of Kepler’s investigation during which he proposed and tested a number of hypotheses. He concludes this part of his discussion as follows: Kepler typifies all reasoning in physical science. Would it have required so much time, and genius, to ‘observe’ the elliptical orbit in Tycho’s data? De Motibus Stellae Martis is more than a compendius expression of Brahe’s observations. Nor is it concerned with deducing geometrical consequences from the elliptical orbit hypothesis, ‘thought of? Kekule-

fashion. Kepler’s task was: given Tycho’s data, what is the simplest curve which includes them all? When he at last found the ellipse his work as a creative thinker was virtually finished. Any mathematician could then deduce further consequences not included in Tycho’s lists. It required no genius to take Kepler’s idea and try it for other planets ({6], p. 84).

What Hanson seems to say is that what Beer calls the “dogma of universality” plays no role in the formulation of theories in physics, but he does not wish to suggest that it is not a goal of that science to formulate theories. I think that this way of looking at the problem has merit in history, too, not only as a possibility, but as something that is actually to be found. We are now about to see that this is the

case, but I think that one point is worth emphasis before we do. Some of those who take offense at the general-laws point of view think it outrageous that writers who are rarely historians should presume to suggest that this or that well-regarded explanation is, in fact, defective since its proponent has failed to provide general laws for it. I would lke to suggest that this is not actually a consequence of subscribing to the general-laws point of view. Hanson makes it clear that Kepler’s problem was to explain Tycho’s observations which latter were not to be rejected or in any way disesteemed. I think that

44 THEORY IN HISTORY the same consideration obtains here. A sound historical explanation, say, one which excites the historical community and wins its acclaim, is not to be deemed a bad one or one not worthy of serious respect simply because it is not attached to some general law or other. Rather, it may very well serve as the starting point for a subsequent theoretical inquiry the initial point of which is to explain 7, or to justify it. From the standpoint of the present paper, then, one function of the search for theory is precisely to preserve all good typical historical explanations.°

Ill. Some promising examples

According to Hanson, one difference between what he was trying to do in the book to which we have referred and what some other philosophers of science have written about scientific explanation is that they have regarded as paradigms of physical inquiry not unsettled, dynamic research sciences like microphysics, but finished systems, planetary mechanics, optics, electromagnetism and classical thermodynamics ([6], p. 1).

Thus, they tend to see these sciences from the viewpoint of the present,

when they are complete, and not from that of Newton, Kepler and Galileo, for whom they were sciences in the making. If advocacy of the general-laws conception of explanation in history is in the hands of those others, who write about it with these paradigms in mind, it is quite easy to see that it might very well seem to take the form of defending the dogma of universality. There is nothing in history which even remotely resembles what mechanics has become, and it is this, I suppose, as much as anything which makes the “dogma” seem dogmatic. Yet, there is a great deal of historical work which points to possibilities for development of theory if only scholars wish to pursue it. For some of it, the pursuit ® Though I do not wish to deal with it here, another important function of theory is to aid in choosing between seemingly plausible, but not always mutually consistent, explanations of the same historical happening. Harry Eckstein—in “On the Etiology of Internal Wars,” History and Theory, vol. 4, 1965, pp. 133-63—gives a striking list of explanations of the outbreak of the French Revolution (p. 137f.), and argues that only theory can enable us to decide among them (pp. 139 and 163). Cf. my “Evidence and Events in History,” Philosophy of Science, vol. 29, 1962, pp. 17594; p. 190f. (In this volume pp. 3-27; p. 22f.)

THEORY IN HISTORY 45 has, in fact, been attempted; for others possibly not. In what follows, I shall discuss a number of examples which illustrate what I am trying

to say. In no case can we say that the required theories have been reasonably established. Yet, in each of them we could hope to have them—not by the immediate generalization of what we must justify, but through the trial-and-error testing of plausible hypotheses. We will begin with Max Weber’s famous study of the spirit of capitalism. The Capitalist Ethos. Though Max Weber is widely regarded as one of the major contributors to the development of sociology in the early decades of the present century, and one whose influence is still said to be felt, it is nevertheless the case that his famous essay, The Protestant Ethic and the Spint of Capitalism [14] is an historical and not a sociological work. It is concerned with capitalist enterprise in Europe and America, and is, more particularly, concerned with the answer to one question about it: How did Western capitalism come to have

the character it had? ‘Thus, it is an instance of one fairly typical sort | of historical writing, that concerned with making sense of some particular something or other. I do not say that this is the only sort of

thing historical books do, but it is one fairly frequent purpose of authors doing historical research. Some, of course, want to describe

or recreate a given course of events, but others are interested in more limited happenings. In one famous example, Henn Pirenne sought to delineate certain factors which affected the development of cities in the Middle Ages [9]. In another, about which I shall have

some things to say later on, Mrs. Woodham-Smith tries to understand the factors which led to the destruction of the Light Brigade in the Crimean War [16]. In just the same way, Weber seeks to answer a particular question about Western capitalism, and, as is known, he traces what he takes to be distinctive of it to certain Calvinist doctrines which he believed tended to encourage certain modes of behavior favorable to the development of “capitalist”-personality types. There are any number of passages in Weber’s essay which sup-

port the view that the problem it deals with is historical and not sociological (cf. pp. 27, 36, 46 and 55), and even though The Protestant Ethic and the Spint of Capitalism is the opening essay of a series called Gesammelte Aufsatze zur Religionssoziologie, Weber makes it clear

that his comparative sociological studies of non-Western economic and religious ideas and institutions were undertaken because of his desire to clarify the relationship between such ideas and institutions in the west:

46 THEORY IN HISTORY for only in this way, [he writes], is it possible to attempt a causal evaluation of those elements of the economic ethics of the Western religions which differentiate them from others, with a hope of attaining even a tolerable degree of approximation.

And he adds that the comparative studies are “definitely oriented to the problems which seem important for the understanding of Western culture” ([14], p. 27; italics added). If one looks at the character of the voluminous literature which has come out of the scholarly reconsideration of the Weber hypothesis one sees that it is both historical and sociological. The formulation of general laws in terms of which to justify Weber’s original historical explanation does not in itself settle the historical question. And,

certainly, a generalization formulated in the spirit of the dogma of universality would only be distrusted by historians and confirm them further in the view that general laws have nothing to do with explanation in history. Scholars concerned with the historical questions discussed in Weber’s essay, continue to seek answers to the question

“How did Western capitalism acquire the character it had?” And scholars whose interests are sociological or nomothetic continue to seek general propositions which would, among other things, make intelligible the answers of the historians. The theoretical problems emerge from scrutiny of the historical explanation and by focusing upon the particular relationships which the historical explanation empha-

sizes. If there is a sense in which the formulation of theories leads to

a more complete statement of the historical explanation, it in no way involves that the particular historical explanation is in any way something inferior. For just as the explanation 1s justified by the theory,

the theory is supported by the explanation. Reflection on the explanation raises questions for theory, but no theory can escape the need to accord with the claims of particular explanations which are accepted as sound by the community of historians. Slavery in North Amernca. One of the more interesting contributions

to recent historical writing is Stanley Elkins’ collection of studies of American slavery [3]. The first of the four studies deals with slavery as a problem of historiography, and the last with the ways in which intellectuals have responded to slavery. From the point of view of our interest here, it is the middle two studies which attract our attention. The first of these two studies takes it for granted that historians

have some clear conceptions of the institutional character of slavery

THEORY IN HISTORY 47 in the southern states and is concerned to make that character intelligible. ‘he other study is possibly based upon a less firmly established historical conclusion. It seeks to account for the personality of the slave, but Elkins’ conception of that personality emerges rather indirectly, and it is not unlikely that other historians do not share it. I do not say this with the intention of being adversely critical, for what we know about such matters must be constructed out of sources written when sensitivity to and techniques for determining personality types were far less developed than psychologists have made them

in recent decades. Yet for our purpose here, the study in question illustrates the same sort of thing as the others mentioned: given an histonian’s conception of the character of the personality of the slave,

how are we to make sense of it? For reasons which had best be assessed by other historical and psychological scholars and cannot detain us here, Elkins concludes that the so-called “Sambo” stereotype was characteristic of American Negro slaves. But he gives no comfort thereby to those who would see this as evidence of racial peculiarities or inferiorities of Negroes,

for, as he says, “One searches in vain through the literature of the Latin-American slave systems for the ‘Sambo’ of our tradition—the perpetual child incapable of maturity” ([3], p. 84). And this sets the problem which is to be explored in his third study, how is this differ-

ence between the personality types of North and South American Negro slaves to be explained. Some scholars have sought to account for the characteristics in question in terms of African culture, but against this it is argued that the Sambo type was not to be found in Africa and, too, that more elements of their African cultures were preserved by slaves in Latin America than by those in the United States ((3], pp. 89ff.). The explanation Elkins finally offers has to do with the psychodynamics of the reaction to capture, enslavement and the passage from Africa to North America ([3], pp. 98ff.). A series of shocks is said to have had the effect of detaching them from their previous cultural moorings and to have made them amenable to the thought and personality re-formations out of which the Sambo type emerged.

To put the matter simply, Elkins claims that there is a particular historical phenomenon which requires to be explained, namely, the existence of a certain personality type among some people which was not to be found among other people related to them. And he

48 THEORY IN HISTORY explains this by appeal to certain antecedent happenings. In this, his explanation resembles a great many others recognized as typically historical. But in éypical instances, the explaining happenings are gener-

ally felt to be of the sort that could explain the explained phenomenon, and I should think that even historians who might disagree with a given explanation could usually understand how someone might believe

that the proposed cause was the cause of the phenomenon to be explained. Why that should be is not germane to our present interest, and I shall not bother to deal with it. But I do think it clear that the explanation offered by Elkins is not of the typical political or economic sort. And it is this which requires that he interpose between his phenomenon-to-be-explained and his explaining causes a psychodynamic mechanism which is thought to connect them. But except for the atypical addition of the particular character of the connection which he takes to bind what he explains with what he thinks explains it, Elkins’ explanation seems to be rather like a good many others that historians offer. But to say that the connection is psychodynamic is certainly not to end the matter. The relevant psychological theories have to be spelled out and their relevance to history clarified. None of this is done in Elkins’ book, nor was it his intention to do so. But we do get some indication from Elkins of what might be involved in this, and here, too, the procedure seems to be to seek out general features of the subject of the inquiry which might be discovered in other contexts. And this leads him, possibly to the surprise of many of his readers, to a consideration of the experiences of those people who were imprisoned in Nazi concentration camps and the effects that experience had upon them. He is particularly interested in pointing to the possible similarities between the shock-and-detachment phenomenon he hypothesizes for the captured, enslaved and transported

Africans and the adaptations of the camp prisoners to their new environments. The one procedural difference that may distinguish Elkins from Weber is that it seems that Elkins simply spotted the similarities in the adaptation of the two kinds of people he discusses— perhaps he was led to the shock-and-detachment hypothesis through reading memoiristic and psychological accounts of the camps—whereas

Weber set about systematically to discover whether or not certain relationships he discovered in the course of his investigations of the capitalist ethos may be found in other times and places. But it is the similarity between the two that concerns us here. Both seek to justify

THEORY IN HISTORY 49 their particular historical explanations by discovering in the situations with which they deal general features which may be discovered at other times and in other contexts. Both point to the need to justify

their explanations theoretically, yet neither is led to the dogma of universality. Counter-revolution in the Vendée. Following his critical remarks on the

dogma of universality, Beer proceeds to offer two examples of historical work which he takes to be entirely acceptable yet they do not satisfy the requirements of the dogma. Actually, what he does is to discuss the two papers immediately following his own in the journal in which it appears. One of these deals with the counter-revolution in the Southern Anjou section of France during the 1790’s and was written by Charles Tilly [13]. This is the only one of interest to us

here, and our interest will be less in Tilly’s paper than in Beer’s account of it ([1], pp. 1Off.).

Like the others we have dealt with, Tilly’s, too, is a straightforward historical problem of the usual kind, though his procedure reflects more explicitly the influence of sociology. His problem is to explain how it was that the counter-revolution broke out where it did, rather than in other parts of France, and one would expect that a historian interested in it would seek out conditions peculiar to the counter-revolutionary areas and less in evidence in other parts of the country. According to Beer, it leads Tilly to systematic comparison of localities in the Vendée where counter-revolution took place with those where it did not. He asks what other differences characterized the two classes of localities and initially finds that

revolution and counter-revolution correlate with two types of social organization, the major contrast being between localized, subsistence

agriculture in the areas of counter-revolution and market-oriented nationalized agriculture in the areas where revolution was accepted. Thus,... Tilly... arrives at the generalization... that in the Vendée at that time in history, if (and only if) traditional agriculture prevailed in an area, counter-revolution occurred ([1], p. 11).

This is only one stage of Tilly’s account, and in the course of his investigation he seeks to narrow the focus of his attention to factors relevant to the emergence of counter-revolution still further through additional comparison. I’o quote Beer again, “he observes that the correlation is not simply with the traditional farming areas, but with localities

where there were important nuclei of trading and manufacturing set in the midst of traditional farming” ([1], p. 11). Tilly himself says that

90 THEORY IN HISTORY such a social situation is more favorable to violent local conflict between ‘old’ and ‘new’, ‘backward’ and ‘progressive’ than 1s a uniformly

advanced, or a uniformly backward social setting. In fact, it 1s not far off the mark to say that throughout the West, the peaks of counterrevolutionary activity were not in the backward section so much as the junctions of rural and urban ways of hfe ({13], p. 995).

In his account of Tilly’s work, Beer calls the sort of thing we have been talking about “the method of comparison and generalization,” clearly pointing to the fact that ‘Tilly does both. He compares the different sections of France with respect to certain characteristics he suspects may be relevant to his inquiry and arrives at conclusions not about specific localities in the Vendée or elsewhere, but more generally about all the localities in revolutionary France having the characteristics specified. Beer distinguishes this method of ‘Tilly’s from

another one he finds in the same writer’s essay, “the method of hypothesis and verification.” ‘The point of this method 1s to explain the conclusions arrived at by means of the first, by the formulation of suitable theories or hypothesis, though Beer seems to restrict it to the choosing of available hypotheses from the writings of others ([1]], p. 12). Beer’s own example will make this clear. We have already noted that Tilly, through his method of comparison, had concluded that revolutionary and counter-revolutionary inclinations were associated with the extent to which urbanization had infected a given area. This leads him to suspect that theoretical writings of sociologists concerned with urbanization would be relevant to understanding the question of his particular historical interest ([1], p. 12f.). In effect, what Beer claims is that influenced by sociologists, ‘Tilly came to see the material he was working on as rather like some others and attempted to see if the theoretical solutions proposed to questions raised about those others would work in his case as well. In this respect, his procedure reminds us of Elkins’ attempt to find an analogy between

the experience of concentration-camp prisoners and of captured Africans in the course of their journey to slavery in America. Actually, there is nothing about the logic of this procedure which requires that its historian practitioners limit themselves to selecting among theories already formulated by other social scientists. Any historian who had

some suspicions as to how he might explain the finding with which he had emerged might offer his own formulation which he and others might investigate. It is hard to believe that Beer would not agree to

this, and one may suspect that his account of the “method of hy-

THEORY IN HISTORY o1 pothesis and verification” was intended to describe its use in ‘Tilly’s study—which is most likely the way it will usually be used. There are a number of steps in the course of development of history and theory to be found in Tilly’s study—and its elucidation by Beer— and it may be well to close my discussion of it by summarizing what they are. To begin with, Tilly is clearly interested in the answer to a typical historical question, for he is concerned to account for some-

thing that happened in a particular place or country and at some given time. The philosophical writers whose interest is in the eluci-

dation of modes of discourse and who have taken the historical narrative alone to be the typical product of typical historical research

might not be pleased with this, but the concern with the particular has long been taken to be characteristic—some would make it the defining characteristic—of historical writing, and I cannot see how one could deny that Tilly’s essay is concerned with a particular notwithstanding the fact that it is not a narrative, tells no story. In any event, Tilly answers the questions not by presenting an account of the plans and actions of leading figures, but, rather, by way of a systematic delineation of the ways in which counter-revolutionary sections of France differed from other sections of the country. The procedure, adapted from the work of sociologists, leads to the formulation of general propositions albeit very much restricted in their spatial and temporal range of application, as Beer points out. Actually, whether or not statements of the sort “In France during the 1790’s, sections having characteristics x, y and z will also have characteristic 7,” are general would depend upon what they are about.

My own inclination is to say that they are about France during the 1790’s, hence rather more particular than general. Some might wish to argue that they are about sections having or not having certain characteristics hence having or not having certain others, and, thus, being about a good many sections, are rather more general than particular. They might correctly point out that such statements are exactly like many formulated by social scientists on the basis of the systematic comparison of social units or particular sections of a given country having and not having characteristics being studied. Since

such statements are not about particular social units or particular sections of the country, it is usually assumed that they are general, and the sociologists who formulate them almost always, perhaps even

always, take them to be general. Yet 1t seems to me that a proper understanding of the ceteris-paribus clause, which must always, if only

52 THEORY IN HISTORY implicitly, be attached to such statements, would show that as they are jormulated they are applicable to the particular society of which these

units or sections are part, and in consequence, must be taken as more particular than general. But whether particular or general, this method which Tilly has taken over from the sociologists enables him to discover that in the France of the period of his investigation certain empirically determinate characteristics of sections of the country were or were not associated with a tendency to counter-revolution. The attempt to explain this, to answer why it should be so, leads him directly to the theoretical part of his study, the part Beer calls “the method of hypothe-

sis and verification,” in which he seeks to adapt the theoretical conclusions of sociological investigators to his task. At this stage of his work it becomes very clear indeed that answers to particular historical questions implicate general social theory. One seeks to explain the specific causal relationships uncovered by Tilly’s research by appeal to general laws of one sort or another, and, at the same time, the kinds of data which emerge from the earlier stage of the investigation are clearly evidence of the sort with which the laws in question are to be tested. The Charge of the Light Brigade. Cecil Woodham-Smith’s The Reason

Why [16] is precisely the sort of historical work that the elucidators

of modes of discourse and others who tend to take a dim view of any effort to “make history a science” take to be typical of the field. And it certainly is typical of a very common sort of historical writing, though it 1s by no means a common one of that common sort. On the contrary, it is written in a most striking English prose—narrative history at its best. Nor is it the coldly “objective” sort of work which advocates of scientific history may be thought to demand. It is, rather, a passionate book, moved by injustice, angry at social evil, and, if | may judge from myself, capable of arousing the same feelings in its readers. Surely, it may be thought, no one who subscribes to the general-laws point of view could think much of the book as history, however much he may appreciate it as art. The purpose of Mrs. Woodham-Smith’s book is to explain the suicidal Charge of the Light Brigade in the Cnmean War. In the course of her account, she tells us a good deal about how the British army was managed during the period preceeding the war—the war itself and its patent mismanagement were major factors in the subsequent

reform of the army—and the way this was related to the power,

THEORY IN HISTORY 93 attitudes and conceits of the ruling classes. In the tradition of narrative history, the book focuses particular attention on the two men who figure most prominently in the events with which it is concerned, the earls of Cardigan and Lucan. Through her treatment of these two, the author is able to make clear to us certain typical features of the military careers of noble lords who could rise to the highest ranks of the army, where they would make decisions affecting the lives and welfare of fellow human beings without ever being more than rank

amateurs in matters military or having field experience in lesser commands. But the two earls are not only typical representatives of recurrent types. They are individual human beings, and Mrs. Woodham-Smith’s narrative does not fail to account for the effects of their “idiosyncracies” upon what happens. ‘They were brothers-inlaw, and they hated each other, in itself an evil omen for an enterprise in which the paranoic Lord Cardigan was subordinate to Lord Lucan. But without adding details, let me simply say that the book is a well-woven account in which the typicalities of ordinary expectation are meshed with the particularities of individual behavior in order to give a convincing explanation of the tragic debacle of the Light Brigade.’

If Mrs. Woodham-Smith is nght about what factors led to the climactic events at Balaclava, it is clear enough that there must be social and psychological regularities in terms of which it is these factors

rather than others that are relevant. These are matters to be investigated further, not in a spirit of claiming that the historical account presented in inadequate, but rather with a view to making theoretically

intelligible what seems to be a very sound historical explanation.® This, to be sure, must proceed by way of detailed investigation of determinate relationships, not in the manner of the dogma of universality. And if Mrs. Woodham-Smith herself is not to be expected to pursue any of the special social and psychological disciplines which

are implicated by her book, there is a sense in which her book may be taken as contributing to all of them. To the extent that her account ’ It is, of course, possible that the author was mainly interested in describing some aspect of 19th century English social history, a subject she has been concerned with in other writings, and that the tragedy of the Light Bngade was only a device toward that end. But even if Mrs. Woodham-Smith was not primarily motivated to

explain the tragedy, her account does in fact do so, and that is all that concerns us ere. ° This, of course, is only an assumption of my using it as an example.

o4 THEORY IN HISTORY Is correct, it contributes relevant empirical evidence with which they must come to grips.

Postscript

Comments I have had from readers of the foregoing lead me to think that I ought to add some concluding words of clarification about

some of the matters I have dealt with. 1. I hoped it would be clear that my discussion depends upon distinguishing between the explanations historians offer of whatever it is that they wish to explain and the theories we would require to have if those explanations are to be justified. Thus, in the example taken from Nagel, we can clearly distinguish three elements: |. the event to be explained, namely, Elizabeth’s signature; 2. Maitland’s

explanation of it; and, 3. Nagel’s generalization. Some would say that the explanation offered by Maitland is incomplete and Nagel’s generalization is needed to complete it. But this obscures the fact that Maitland’s explanation may be presumed to be compatible with an indefinitely large number of other generalizations as well. It seems more correct to think of him as having offered one specific explanation, rather than something which turns out to be the common component of an indefinitely large disjunction of explanations. But while my way of putting it focuses on the justification of an historian’s explanation, most of those who defend the general-laws point of view talk about the requirements for the explanation of an

event. Let 7 refer to the statement of theories or general laws relevant to some case, and e¢ to the event to be explained. And let c refer to what most covering-law theorists think of as the initial or boundary conditions, but what most historians might think of as the set of factors which explain the event. My way of thinking about the

problem might then be symbolized as T > (c > e), while the usual view would be symbolized as (T ¢ c) > e. On the usual interpretation of the “D” symbol, these two are logically equivalent, but I think

that the former is better nevertheless. For it makes clear that the starting point for theoretical inquiry is the historian’s explanation, which it takes seriously and actually represents in the symbolism within

the parentheses. But the latter seems to suggest that history without theory is absolutely nothing, that the usual explanations of historians are not explanations at all. It is hard to doubt that the attitude that

THEORY IN HISTORY 99 historians read into the sort of discussions the latter symbolism represents contributes much to the unwillingness of so many of them to take it all seriously. The latter symbolism seems to tell the historian where he has gone wrong; the former how his results may lead to further developments. While he may well resent the first of these, he may equally welcome the second. This leads me to just what it is that I classified above as good and bad examples. Someone wondered what I objected to in the Maitland

explanation. And someone else did not see how I could deny that various examples in Dray’s writings were perfectly good and typical—

I shall return to “typical” in a moment—or how [I could claim that they were inferior to the examples I seemed to favor. But in my lst of bad examples, it is not the historical event that I object to, nor the historian’s explanation of the event. My objection is only to the generalization offered to account for the explanation. Each of them is simply an immediate generalization of the explanation it 1s called upon to justify. While logically it serves to provide what a Humean analysis finds missing in the usual accounts of historians, it 1s too simplistic and trivial to be interesting or even true. What distinguishes,

then, the two samples is not the character of the historical events in them or the ways in which historians might explain those events, but rather the ways in which the justifications of the explanations are offered. In the first set, each justification is an instance of the dogma of universality. In the second, even when no generalization is actually offered we can see how painstaking attention to what seem to be the crucial features of the event and its explanation may result in subsequent development of theory. 2. The word “typical” and its kin are found with some frequency in this paper, and one reader found my conception of a typical historical problem unclear and did not see how I could use it to distinguish between work I find promising and work I am dubious about. But I have no special conception of typical historians’ activity, and I did not intend to use such a conception to make the distinction in question. What historians are wont to do is typical of their craft and certain kinds of questions are typical historians’ questions. Historians typically ask why something or other happened—why Elizabeth used “et caetera” or why the Light Brigade charged or why Dust Bowl farmers moved to California. Or they typically ask why something took the form it took or had the character it had—western capitalism or the personality of the North American Negro slave. These are

96 THEORY IN HISTORY typical questions for historians as other kinds of activity are typical

activity for tailors or shoemakers or barbers. My point is that the question, “Why did the counter-revolution begin in the Vendée and not in some other part of France?” is a typical historian’s question, notwithstanding the fact that Tilly uses procedures not typically used by historians to answer it. And the question, “How did the Sambo personality type come to be as it was?” is a typical historian’s question even though it led Elkins to an atypical concern with psychodynamic factors and away from the historian’s typical preoccupation with social, political and economic matters. It is not a typical historian’s concern

to formulate general laws or to seek out the theoretical justifications for the explanation he offers. But that in no way entails that what historians typically do does not implicate general theory. 3. ‘This leads to the final point of this postscript. It is clear enough from my remarks early in the paper that I am out of sympathy with the elucidators of modes of discourse. Yet, someone challenges, the

real issue confronting the parties to the present-day discussion of explanation in history is precisely whether there is a distinct historical mode of discourse. Perhaps “mode of discourse” is not a sufficiently

clear term, and my challenger thinks the question “Is there a logic of explanation peculiar to history?” is essentially the same as “Is there a mode of discourse peculiar to history?” I think that these may well

be separate questions, though it is clear enough that some writers, W. B. Galhe for one,’ do not. F. J. Teggart once argued that the historians’ tradition of casting their conclusions in the form of narratives

is a stumbling block to the achievement of their goals [12], but if the narrative mode of discourse is precisely what distinguishes history from other intellectual disciplines, poor Teggart was only babbling. (He wasn’t.) Similarly, we would have to admit that ‘Tilly’s problem was not historical after all, notwithstanding its concern to make sense of a particular happening. In my view, Austory is not a mode of discourse but a way of knowing, a way of warranting assertions about past events.

It is not the case that such assertions can only be made in narrative form if they are to be treated as history, just as it is not the case that every assertion about past events is historical.'° It would be out of

place at the end of this paper to pursue what follows from these views. I have only stated them here in order to make clear that my > See note 4. '0 In the sense of “historical” as that which pertains to the discipline of history,

THEORY IN HISTORY 37 earlier remarks about the elucidation of modes of discourse were made

not because I have any objection to that sort of activity as such, but because it does not seem to me that it casts any light at all on the problems of fustorical knowing.

REFERENCES

[1] BEER, SamMuEL H., “Causal Explanation and Imaginative Re-enactment,” FHestory and Theology, vol. 3, 1963, pp. 6-29.

[2] Dray, Witiiam, Laws and Explanations in History, Oxford, 1957. [3] Exvkins, STANLEY M., Slavery: A Problem in American Institutional and Intellectual Life,

Chicago, 1959. [4] GaRpDINER, Patrick, ed., Theortes of History, Glencoe, Ill., 1959. [5] Garpiner, The Nature of Historical Explanation, London, 1952. [6] Hanson, N. R., Pattens of Discovery, Gambridge, 1958.

[7] Hempe., C. G., “The Function of General Laws in History,” in [4], p. 349f. [8] NaceL, Ernest, “Determination in History,” “Philosophy and Phenomenological Research,” vol. 20, 1960, pp. 291-317; p. 303. [9] PrrenneE, Henri, Mohammed and Charlemagne, London, 1939 and 1954. [10] Popper, Karu R., The Open Society and Its Enemies, Princeton, 1950, p. 448. [11] Scrrven, Micuaer, “Truisms as the Grounds for Historical Explanations,” in [4], p. 454. [12] Toner FREDERICK J., Theory and Process of History, Berkeley and Los Angeles, 1949, Paperback ed. 1961; pp. 18-29, 51-66. [13] Titty, CHarzes, “The Analysis of a Counter-Revolution,” History and Theory,

vol. 3, 1963, pp. 30-58. [14] Weser, Max, The Protestant Ethic and the Spint of Capitalism, tr. Talcott Parsons,

New York and London, 1930: references will be to the third British impression, 1950.

[15] Wurre, Morton, “Historical Explanation,” in [4] pp. 356-73. [16] WoopHam-SmiTH, CreciL, The Reason Why, New York, 1953; (Time Reading Program edition, 1962.)

not the course of human events. Only the former sense of “history” is of concern in this paper. Even if there is every reason to believe that a certain proposition about a past event is true because its character as divine revelation is beyond doubt, that proposition would not be historical in the former sense. Similarly, statements I

make about my own past because I remember this or that—say, what I had for breakfast this morning—are not historical in the given sense either. Neither of these kinds of statement is established in the historical way.

IDEALS OF ORDER: HISTORY AND SOCIOLOGY* I In his brief but sumulating Foresight and Understanding, Stephen Toulmin

presents an account of the role of what he calls ‘ideals of natural order’ in the pursuit of research in natural science.' These are the fundamental conceptions of the nature of reality which are said to inform the character of scientific investigation, determine the problems

which are called to the attention of investigators, and even decide what are and are not empirical facts. Toulmin’s account is rooted in a recently emerged conception of the history of science,’ according to which that history is not simply a progressive accumulation of facts and theories, each generation in the course of which pursuing what 1s essentially the same task with increasing sophistication. Rather,

it is now claimed, there is a sense in which there is a certain discontinuity between the physics, say, of one period and that of the next. To see Aristotle as physicist doing badly what Galileo would one day do much better rests, says Toulmin, on an anachronism,’ and it involves a reading of Aristotle in a way no one was capable of prior to Galilleo’s own work. What ‘Toulmin tried to do in his discussion of this is to discover what Aristotle’s model of physical action was and how it was related to what he took to be problematic. This turns out to be rather different from the analogous views of Galileo. As long as Aristotle could confine his attention to cases which resembled his standard type—say, motion against heavy resistance—Toulmin claims his physical theory could work in a fairly serviceable way. But in other cases—‘bodies moving against a sufficiently light resistance’-— Aristotle’s physics ceases to apply.* One requires then to have a new physics based upon other models of physical action, and presumably * Written originally for presentation at a conference on philosophy of history and social science, York University, April 1969. ' Stephen Toulmin, Foresight and Understanding, New York 1963, chs. 3 and 4.

2 One of the best known general accounts of which is Thomas S. Kuhn, The Structure of Scientific Revolutions, Chicago 1962.

> 'Toulmin, op. cit., p. 47f. * Ibid., p. 52.

IDEALS OF ORDER: HISTORY AND SOCIOLOGY 39

the success of the new physics that had its beginnings in the Renaissance is owing to its being based on a model of nature, an ideal of natural order, capable of rendering intelligible a wider range of physical phenomena.

The main impetus for the introduction of the conception of ideals of natural order comes from the history of science. As I intimated above, it functions as a conceptual tool in terms of which to reconstruct

our idea of the history of science in that it replaces the idea of the progressive accumulation of scientific truth with that of a succession of modes of science each one with some accomplishment to its credit until its potentialities are realized, its limits are reached and the need

to surpass it becomes increasingly apparent. The point here is that at any one time there is but one ideal of natural order which pervades

the atmosphere of any given science, and once that ideal has been replaced it never returns. ‘Thus it would seem that the conception of ideals of order would have limited applicability, namely, only to those scientific disciplines which have achieved that degree of maturity so as to be able to have so dominant and guiding an ideal. Consequently, it would seem that it would have no application to the social sciences.

At most one might want to say with respect to them that sometimes certain kinds of techniques may seem to dominate—witness all the talk of behavioural science and behavioural methods—but surely it has never been the case that one fundamental ideal of social order has dominated theory in any social science in a way which would make it the basis of what T. S. Kuhn calls ‘normal science’.® But if one cannot write the history of social science, or of some particular social science, say sociology, in the way suggested by Toulmin’s book, it by no means follows that his conception of ideals of order may not prove useful in dealing with it in some other way. If one attempts to discover what it is that characterizes sociology, what set of assumptions govern its procedures, one would likely throw

up one’s hands in dismay. In terms of what, it may be asked, can it be said that both Max Weber and George Lundberg were practitioners

of the same discipline?’ In every major respect their conceptions of

’ Obviously, not intended in the prescriptive sense. ® Kuhn, op. cit., ch. 3. ’ Their generational difference is not pertinent here: I doubt that a survey of sociologists would reveal that generation is a factor in classifying them according to the character of their points of view.

60 IDEALS OF ORDER: HISTORY AND SOCIOLOGY what they are doing—what sociology is—differs. A few remarks about

each will make this clear. We may begin with Weber.® Central to Weber’s characterization of sociology is that it is concerned with subjectively meaningful action. When he speaks of the subjectively meaningful, Weber is trying to delineate the subject matter

of his discipline and wants to say that the data sociologists study have standpoints of their own. This, of course, is in contrast to the ‘objective’—as he uses the term—phenomena of physical science which are simply there, having no pre-interpretations of their situations which

a behaviouristic physicist would overlook. It would follow from this that the study of the social world is autonomous in that it cannot be understood as a special case of other sciences whose basic concepts presumably reflect the fact that they are concerned with phenomena which do not have a point of view of their own. The crucial methodological problems for Weberian sociology must surely be rooted in this conception of its subject matter, and it is clearly not possible to make sense of Weber’s intention in the matter without paying due attention to his idea of a subjectively meaningful phenomenon.’ The views of Lundberg are fundamentally different.'° In his esti-

mation there is, in the end, only one science, physics, or, rather, ‘physics’, for it need not be the case that it is our present discipline of physics with all of its present-day features which will provide the idiom for the expression of all scientific truth on that day when at last all of the lesser, still groping disciphnes reveal their properly physical nature. It is, in fact, Lundberg’s explicit claim that progress in the history of science has always been the gradual replacement by

physics of disciplines once presumed to be something other than ® My account will be restricted to ‘“Objectivity” in Social Science and Social Policy’ in M. Natanson, ed., Philosophy of the Social Sciences: A Reader, New York 1963,

although interested readers might find it useful to consult “he Fundamental Concepts of Sociology’, the opening chapter of that part of Weber’s Wirtschaft und Gesellschaft translated as The Theory of Social and Economic Organization, New York 1947. Notice

might be taken of the final chapter of Reinhard Bendix, Max Weber: An Intellectual Portrait, Garden City 1962. An attempt to produce philosophical foundations for sociology by elaborating Webenan principles in light of Husserl’s phenomenology may be found in Alfred Schutz, The Phenomenology of the Social World, Evanston (IIl.), 1967. > For two clearly different approaches to Weber’s methodology see Carl G. Hempel,

“Typological Methods in the Social Sciences’, in Natanson, ed., op. cit., and Schutz, Collected Papers, vol. I, The Hague 1962, Part 1. '° See his “The Postulates of Science and Their Implications for Sociology’, in Natanson, ed., op. cit.; this essay is really the opening chapter of his Foundations of Sociology, New York 1939.

IDEALS OF ORDER: HISTORY AND SOCIOLOGY 6]

physics.'! Thus, for him, the goal of all empirical research is to contribute toward the emergence of one unified science, and views, like Weber’s to the effect that the social realm 1s suz generis in that its data may not be reduced—in whatever sense is pertinent here—to something else is simply the ideal of obscurantists possessed of a mystical methodology. Clearly, then, for him the notion that sociology deals with phenomena which are subjectively meaningful 1s not acceptable, and his strongest inclinations here are physicalistic-behaviouristic. Not meaning, but language is, for him, the point of departure. Indeed,

his way of dealing with this leads directly into his conception of the unity of science, for he holds that all science begins with responses to stimuli as these are recorded in language. In this respect there 1s already no difference between natural and social science. Where the two do depart—hopefully not for long—is that the former deal with verbal responses with respect to which there is a greater degree of inter-subjective agreement than do the latter. So anxious is Lundberg to excise meaning, thinking and purpose from the universe that he is not able to understand scientific theory construction as the attempts of human beings to organize the data of experience and experiment, and in his view the eighteenth-century controversy about the nature of combustion was not over the theoretical advantages of one or another

theory, but simply that ‘the older chemists of the day thoroughly habituated to thinking about fire in terms of phlogiston continued to “feel” that the new explanation “left something out’.’'’* What was

left out was a word to which the older chemists were so used to responding that it ‘was therefore as “real” to them as the word “wood”

or whatever other words are used to symbolize the factors assumed to be present in a given fire’.'” Withal the inadequacy of those two sketches of the general stand-

points of Weber and Lundberg, one point does emerge from our discussion: even if it is not possible to write a history of social science

in which one grand paradigmatic determinant of normal science succeeds another or in which successive ideals of social order dominate

the practice of a discipline, such notions are not without use in an effort to comprehend the state of any social science. For what is the conflict between Weberian and Lundbergian approaches to the study "| Ibid, p. 396. 2 Thid., p. 43f. '5 Tbid., p. 44.

62 IDEALS OF ORDER: HISTORY AND SOCIOLOGY

of social phenomena if not the rivalry of ideals of order? Each of these offers, at least implicitly, an account of its own with respect to all manner of things that sociologists may be concerned with, and most particularly about the very nature of sociology and its subject matter. It is, of course, possible to characterize verbally what might be done in order to choose between them: we need only formulate

testable theories in terms of the prescriptions of each of them and discover empirically which of them survives the test. But even without introducing difficulties from without the limited framework of our present discussion, this is far more easily said than realized. Each of the ideals of order in question has a rather clearcut conception of what is a social fact; these conceptions seem quite irreconcilable and it is not to be expected that the advocates of one point of view will accept as evidence what cannot be assimilated to its own conception. Too, it is quite likely the case that testable hypotheses and evidence

relevant to them which are produced in the spirit of one of these competing ideals of order, may well be subject to reinterpretation in terms of the other.'* In such cases it might not be too obvious just

what bearing the outcome of the investigation actually has on the controversy. In any event, while some may think that in principle the controversy ought easily to be solved, I suspect that this is not in fact the case. What is clear, however, is that rather unlike the situation in natural science, in the social sciences we have not only the usual problems of discovery and explanation, but the more fundamental issues of what sorts of thing the subjects of our discovering and explaining are.

I If we have confined our remarks to the field of sociology, it 1s not, of course, the case that our considerations are limited to that field. I dare say that no small part of the controversy and ferment within the social sciences could be reconstructed in terms of the notion of ideals of order. Indeed, one may suspect that such a reconstruction could prove to be of immeasurable value, but only if the effort were '* Cf., for example, the way Erwin Straus interprets the experimental results of Pavlov’s work so as to use them against Pavlov’s own point of view; The Primary World of Senses, New York 1963, Part I.

IDEALS OF ORDER: HISTORY AND SOCIOLOGY 63

made to get far beyond the inadequate sketches I offered above in order to produce rather detailed characterizations of the ideals in conflict. But without seeking to work out in detail the character of any particular ideal of order, I should like to consider whether or not the application of this notion may be of use in dealing with still another issue in the philosophy of social science, namely in the degree, If at all, to which the social sciences are relevant to history. The range of topics which may be discussed under this heading is far wider than what we are likely to discover by a survey of the literature,'’? but in what follows I expect to remain well within the limits of what methodologists will recognize as acceptable discourse.

But I should like to mention in passing that there are at least some issues of this field that might well be of interest to philosophers were they willing to stray somewhat from the usual assumptions made in these discussions. ‘(hus some who are committed to the well-known covering-law conception of historical explanation strongly affirm that

the social sciences are indeed most relevant to that task, for inasmuch as there are no specifically historical laws it is precisely to the social sciences that history must turn for the laws its explanations require.'° There is a general taking for granted that the sorts of event which are presented in historical accounts are simply instances of the sorts of event with which theoretical social science is concerned, but

a moment’s reflection makes it clear that there are at least some questions which have to be dealt with before we can be secure in this belief. One increasingly recognizes that explanations are not of concrete events but of events with respect to certain characteristics they may have, under this rather than that description, as it were. But the descriptions we find in history books are never offered in the languages of the social sciences, hence they do not make use of the

concepts in terms of which explanation in this or that branch of theoretical social science are offered. Presumably, all that is then required is that the historical event be redescribed in the appropriate language.'’ But what are we assuming when we say this? One assumption is that the kinds of descriptions we find in history books For an interesting and wide-ranging exception see Nathan Rotenstreich, Between Past and Present, New Haven 1958, ch. 4. ‘© For one of the best known statements of this view see Morton White, ‘Historical Explanation’, in P. Gardiner, ed., Theones of History, Glencoe (Ill.), 1959.

'’ For an example of this, see John T. Flint, ‘The Secularization of Norwegian Society’, Comparatiwe Studies in Society and History, 6, 1963-4, 325-44.

64 IDEALS OF ORDER: HISTORY AND SOCIOLOGY

are not essentially related to the epistemic status of the historical event as reconstructed, and so any number of other modes of description might do as well. Let me clarify briefly what I mean by epistemic status. Much present-day philosophical interest in history is

concerned with explanation of historical events, and so we find a tendency to take for granted that the event is available to be explained.

This taking for granted is reflected in the fact that there is almost nothing in the recent literature on how the historical past—the presumed locus of events to be explained—is constituted in historical research. Taking it for granted, then, we simply assume that just as any present situation with which a sociologist or political scientist may wish to deal might be characterized or described in terms of the different conceptual systems such scholars use, so too may the historical event. But it may well be that, owing to the ways in which our knowing the historical past differs from the ways in which we know events taking place in the present, this may be mistaken. This

is no doubt obscured because of the fact that descriptions of past and present events may be made in identical language. But if we take seriously ether that what a description means or its status as true or false rests upon observables, then we must immediately take seri-

ously that the observables upon which past descriptions are based are entirely unlike those upon which present descriptions are based. The latter are things happening or events taking place; the former are documents, coins, ruins, and other sundry sorts of historical evidence. With this in mind, it is no longer quite so evident that past and present descriptions may be used interchangeably and indiscriminately as evidence for the same theories. Some of the assumptions of the discussion of these matters may well require being made explicit and subject to examination.

What I want to do in what follows is rather different from what I have been musing about. I want to look at two ways of treating— two attitudes toward—what 1s essentially the same problem in order to see whether or not we can say that the differences between them implicate differences in ideals of social order. Obviously, we cannot

say that these would be the ideals of order implicit in present-day history and sociology, for we have already seen with respect to the latter that it has no one such dominant ideal and it may well be

suspected that the same is true of the former. But I think it not unreasonable to believe that the ideal of order implicit in the sort of sociology I shall be discussing 1s not by any means uncommon among

IDEALS OF ORDER: HISTORY AND SOCIOLOGY 65

the practitioners of that discipline. In any event, it may well be interesting to discover whether or not focus upon ideals of order clarifies anything about the attempt to deal sociologically with historical material and with reactions to the attempt. Our problem may be presented in the form of the following question: Why does John Flint,'® an historical sociologist, regard so highly Guy Swanson’s recently published Religion and Regime,'? and William

Bousma,” though making the usual obeisance to theory and its uses one expects from an historian who is wary lest he be deemed obscurantist, hold what is clearly an opposite view? Flint by no means denigrates the ideal of factual adequacy, the desire that an historical account be true, yet the possibility that Swanson is sometimes short of realizing that ideal does not really trouble him: ‘I think it is highly probable that Swanson will be found historically wanting and simply wrong on a number of points. If so, he will be wrong for at least a

few of the right reasons.’ Those right reasons turn out to be that he is pre-eminently nght about research design and his work exemplies what Flint takes to be the most desirable purpose of historical sociology, the testing of theoretical propositions by means of comparative research. Bousma, on the other hand cannot really get beyond the realization that a very large part of what Swanson claims to be the case is simply false. That he cannot find true answers to historian’s questions seems to him to be quite adequate ground for rejecting the work, without a single word of praise for the research design. Since the book cannot be both as good as Flint thinks and as bad as Bousma does in the same respects, one must presume that it is good and bad in different respects. What are these respects? It is here I think, that one may find it of some interest to sketch the ideals of order which underlie the two appraisals. But we must first turn to Swanson’s book in order to have some conception of what it is that is being appraised. One standard text on the Reformation opens with the following words:”? '8 John T. Flint, ‘A Handbook for Historical Sociologists’, Comparative Studies in Society and History, 10, 1967-8, 492-509. '? Guy E. Swanson, Religion and Regime: A Sociological Account of the Reformation, Ann

Arbor 1967. © William J. Bousma, ‘Swanson’s Reformation’, Comparative Studies in Society and Fiistory, 10, 1967-8, 486-91. *! Flint, ‘A Handbook for Historical Sociologists’, p. 505. *? Harold J. Grim, The Reformation Era 1500-1650, New York 1966, p. 1.

66 IDEALS OF ORDER: HISTORY AND SOCIOLOGY The Reformation had its inception in the search for the answer to a typically medieval question, How can I be saved? The answer which Luther found after years of inner conflict, prayer, and study involved basic theological assumptions which threatened to undermine the well-integrated doctrinal system and the highly developed ecclesiastical hierarchy of the church. The rapidity with which the doctrines of Luther

and the other reformers spread throughout Europe is evidence of a general concern over the question of salvation, and also of a strong dissatisfaction with the secularized church for not adequately serving the religious needs of the people.

The same author observes a few pages further on that ‘Although the emergence of Protestantism was prompted solely by religious motives, the hopes and aspirations of the reformers were almost immediately translated into economic, social, political, and cultural terms’.”’ That the problem of the Reformation is that of salvation is, apparently,

not an uncommon view. It is, in any event, stated explicitly by Bousma,”* and in the early pages of his Reformation Europe, the emi-

nent historian G. R. Elton, seems to focus upon the same thing. Speaking of Luther, Elton says:” ... Lhe study of his lectures in the years before he became a public figure has shown the development of his theology and illuminated its close links with late-medieval teaching and mysticism, with views, that

is, on the relationship of God and man which laid the stress not on institutional and sacramental expedients but in the search of the individual soul. Nevertheless, Luther’s views proved revolutionary because they concentrated with such incisive singlemindedness on the total inability of man to help himself to his own salvation. He took seriously

the accepted doctrine of God’s omnipotence and... sole monopoly of free will. In consequence, as the event was to show, Luther had rendered superfluous the whole apparatus of the Church, designed to

mediate between man and God....

Thus, it would appear to be a fairly established view among scholars of the Reformation that its essential motivation was religious in char-

acter, and more particularly, that it reflects the concerns of Christians with the ultimate fate of their souls. The passage quoted from Elton’s book makes it clear that the Protestant turning against the Catholic Church was rooted in Luther’s demonstration that the Church 3 Tbid., p. 5. ** Bousma, op. cit., p. 491. *° G. R. Elton, Reformation Europe 1517-1559, Cleveland and New York 1964, p. 16f.

IDEALS OF ORDER: HISTORY AND SOCIOLOGY 67

was not capable of being what it always claimed it was, the divinely-established human instrument of salvation.”

This conception of the Reformation is in no way reflected in Swanson’s book. We shall soon see that in matters of procedure Swanson and the historians are far apart, indeed, and we shall want at least to speculate on what assumptions they make respectively that may account for the differences. But here we confront a difference of another sort entirely, a difference in the very conception of what it 1s that students of the Reformation study. In Swanson’s view, the various and conflicting theological standpoints of the Reformation

era are articulations of different forms of political integration. On the basis of the secondary sources available to him and his associates, he collected material on forty-one different political units—national states, principalities, free cities, and the lke—and concluded that notwithstanding a certain amount of variation, the forty-one could be reduced to more manageable proportions by a five-part classifica-

tion. The criteria in terms of which the classification is made are ‘the extent to which a regime’s exercise of gubernaculum was controlled by a political community’,”’ and the nature of the regime’s gubernaculum, we are told, ‘pertains to the nghtful, the legitimate, powers of a regime and a regime is the political machinery by which decisions are made to advance the interests of a social system’.”® In Swanson’s view, the period under consideration was one of significant political change, and people require to have techniques whereby they may render seemingly unstable situations more manageable, hence

making their condition more viable. It is the function of the theological articulation of the political system to do just that. The various churches of the Reformation era, not excluding the Roman Church

itself, are said to differ with respect to the extent of their belief in God’s immanence,”’ and in Swanson’s view the members of a politi-

cal community tended to subscribe to that church which gave expression to its own experience of its political order. Since it is not my purpose to deal with the adequacy of Swanson’s understanding of either religion or regime in the Reformation, I do not wish to © So, too, Preserved Smith, The Age of the Reformation, New York 1920, p. 65f. “7 Swanson, op. cit., p. 42.

8 Ibid., p. 33. *? An interpretation which is challenged by Robert M. Kingdon in a review of Swanson’s book published in the Amencan Sociological Review, 33, 1968, 843. I am grateful to John Flint for calling this review to my attention.

68 IDEALS OF ORDER: HISTORY AND SOCIOLOGY

pursue these analyses further. I want simply to underscore the difference between Swanson’s view of the subject of study and those of the historians referred to above. For them, the point of departure is

religion, and the development has in large measure to be seen in terms of the religious aspect of human concerns. For Swanson, the Reformation is a consequence of changes in the political characteristics of European politics, and its religious aspect is owing largely, perhaps entirely, to the fact that theological systems are reflections of social and political reality introduced because people feel the need to make some sense of and offer some justification for the situations in which they find themselves. Our question, then, is what are we to make of the difference in understanding of the subject matter. I shall offer some speculations which I think are not unreasonable. We have, I think, to do here with two elements of the sociological tradition, one of which, the influence of Durkheim in the sociological study of religion, 1s explicitly acknowledged by Swanson,” and the other, the conception of sociology as a positive science,*' is simply taken for granted with no discussion or acknowledgement needed. The influential view of Durkheim may be summed up in his own words as follows: ‘Religious representations are collective representations which express collective realities’.°? A people’s conception of

divinity is taken to be a reflection of its social arrangements, and religion is taken to be an instrument by means of which a social collectivity is enabled to continue its existence. If this view is taken seriously then, implicitly, the study of religion on its own terms is

not encouraged. In an obvious sense, religion is taken to be epiphenomenal and, as such, something without a real history of its own. At each stage of what seems to be the history of some given religion what in fact you have is the consequence of the character of the socio-political order. The two historian critics of Swanson’s book make no explicit reference to this, but we may well wonder if part

of their inability to accept it may be traced to the view that the Reformation is part of an on-going historical development having determinate links to what happened before, on the one hand, and 0 Swanson, The Birth of the Gods, Ann Arbor 1960, pp. 14ff. °' This is not to be confused with the positivism of Lundberg whose views were sketched above. 2 Emile Durkheim, The Elementary Forms of the Religious Life, tr. J. W. Swain, Glencoe

(Ill.), 1947, p. 10; cf. the treatment of religious phenomena in A. R. Radcliffe-Brown, The Andaman Islanders, Gambridge 1922.

IDEALS OF ORDER: HISTORY AND SOCIOLOGY 69

what was still to happen, on the other. We will return to this point

later on in our discussion and will simply note here that the Durkheim-Swanson conception of religion clearly implies the view that religion has no history, and that one understands it only as the consequence of the determinate forms of other aspects of the social world, which, presumably, do. Historians who study religion may well baulk at this.*° I cannot say how successful this view of religion and its systematic study has been, though it may well be that assessments with respect

to this are themselves affected by what one looks for—again the question of ideals of order. But it is worth attending to the fact that this view is clearly related to the other element of the sociological tradition noted above, the conception of sociology as a positive science.

This has typically involved the detachment of the subject of study from any sort of dependence upon individuality or subjectivity. The emphasis, frequently encountered in historians’ writings, on the role of particular persons is clearly avoided in the writings of those who subscribe to the conception we are considering. The historians re-

ferred to above all give a central place in their accounts of the Reformation to Luther’s role. ‘There is, to be sure, a continuity in the history of Christianity, but this is not an impersonal continuity for it is mediated through the experience and thought of significant persons. In Luther, we are told, the concerns of medieval Christianity are continued, but he, the particular theologian, thought through the consequences of certain doctrines, made his conclusions public, and falling upon a Christendom vitally interested in its salvation had the consequences it had. To positivist sociology this is unacceptable for

two reasons. One is that so much is made of the role of Luther as if to suggest that his absence would have made an important difference:

one of the striking characteristics of Swanson’s book on the Reformation is that so little attention is given to Luther. The other is the view widely held among social scientists that regularity—which is what social science seeks—-may only be discovered by attending to classes

of recurrent phenomena, and Luther is patently not a recurrent °° They need not, of course. A historian who takes his task to be the reconstruction of the religious life of a given period may well be able to keep his mind entirely away from antecedents and successors. Yet, historians do acquire a sense of continuity, which may well become an element in the ideal of order implicated by their work, and thus—explicitly or not—react against a conception, such as Swanson’s, in which their sense that Christianity has a history is implicitly denied.

70 IDEALS OF ORDER: HISTORY AND SOCIOLOGY

phenomenon. As we shall see later on, this view is mistaken, but it is most likely a feature of the conception of science to which Swanson subscribes, and since one expects a degree of congruence between a

scholar’s conception of science and his views of the nature of the reality to which science is to be applied, the sort of regularity he requires is presumably a feature of his ideal of social order. Hence, he is led to a view of his subject that enables him to find regular causes—recurrent forms of political organization—for regular effects—

recurrent systems of theology and of ecclesiastical polity. These last remarks lead us to another feature of works like Swan-

son’s, namely, the tendency to see the social world as made up of discrete parts. This is nothing that need be adversely criticized, and any number of valuable contributions to social-science theory would be impossible unless this could be done.** In Swanson’s book, it takes

the form of separating or detaching from the context of the total social world the political and the religious structures which are then subject to statistical-correlational tests. I shall shortly offer a very brief sketch of just what Swanson has to say about the way in which these two sorts of phenomenon are related, but I want first to talk briefly about this sort of thing as still another way in which sociology—at least as practiced by Swanson and Flint—may differ from the practice of historiography. In typical products of the latter, we find what might metaphorically be called a seamless account in which one thing leads to another in an effort to present a reasonably complete picture of a total situation. I do not wish to suggest that historical works must of necessity be of such a character, and I have elsewhere expressed the opinion that the view of history which takes the narrative form to be definatory of it is mistaken.” But it is nevertheless the case that an holistic attitude is implicit in no small amount of his-

torical writing and, thus, may be thought to be an element in the ideal of social order implicit in its practice. This sharply contrasts with the methodological ‘atomism’, if we may call it that, of works lke Swanson’s. Given the procedure he follows that atomism is inevitable; without it there could be no way to effect the correlations which are the hallmark of such work. But though I call this atomism methodological, it by no means follows that it is held in an als 0b or ** George P. Murdock’s Social Structure, New York 1949, is an excellent example

of what I have in mind.

* See my “Theory in History’, Philosophy of Science, 34, 1967, 23-40; p. 29f., n. 5,

37 and 40. (In this volume pp. 33-57; p. 42 n. 4, 52 and 56f.)

IDEALS OF ORDER: HISTORY AND SOCIOLOGY 7]

a conventionalist spirit; in any event, our concern here is with the ideals of order implicit in a given mode of sociology, not in the opinions of sociologists. Presumably, in back of it—at least in any number of instances—lies the assumption of the real separability of elements. And this, in turn, may well seem like a distortion, or violation

of the integrity, of an historical whole in the minds of those whose commitment is to the reconstruction of past events they take to have some total or holistic character. J want now to say something about the actual theory that Swanson offers—though, obviously, I can hardly do justice to a substantial work in the short compass I can allow myself here. My purpose is then to attend to what such a theory is expected to accomplish in order to contrast such expectation with what historian students of the Reformation seek to accomplish. We have seen that in Swanson’s opinion theological changes may be related to changes in the political order. Uhus he says: ‘I have assumed that most members of a society’s body politic will experience changes in the structure, power, and moral force of such collective processes and that they will together seek in

a theology to define the source of their new experiences in order more adequately to cope with the basic changes in their world’.*° In his estimation the various forms of Christianity which are in conflict

differ with respect to the degree to which God is immanent in his creation,’ a point concerning which both Bousma® and Kingdon® are in disagreement. These, he claims, are correlated with forms of polity and he says “We are looking for the characteristics of a central government that may be associated with the belief in the immanence of God’.*® After these words he writes the following: I have argued that immanence refers to the presence of one actor’s powers, of attributes of his personality, in the personality of another. I propose that the presence of a social system’s powers in its regime is what we seek. I conclude that the regime’s nghtful powers, gubernaculum, are those it has as the political agent of the social system and that the social system will be considered immanent in its regime when the regime alone exercises gubernaculum.

This means that the greater the degree to which the gubernaculum is concentrated in the hands of a regime which is not required to °° Swanson, Religion and Regime, p. 231.

7 Ibid., p. ix. 8 Bousma, op. cit., p. 488. *° Kingdon, op. cit. * Swanson, Religion and Regime, p. 36.

72 IDEALS OF ORDER: HISTORY AND SOCIOLOGY

share it with, or answer for what it does with it to, other elements of the body politic, the more likely the people of that political community will subscribe to the tenets of a religion which emphasize the

immanence of God. Swanson thinks that in order of descending degrees of belief in immanence the religions in question are Roman Catholicism, Lutheranism or Anglicanism, and Calvinism. Now let me characterize the five classes of regime Swanson claims to have discovered:*' centralist regimes in which a governor alone, or through agents responsible to himself, exercises power and which Swanson expects to remain Catholic; limited centralist regimes which are rather like the previous kind but in which some power in the application, though not the formation, of policy is given to certain subdivisions of the state and which are expected to become Lutheran or Anglican; balanced regimes in which subdivisions of the body politic had some role in the shaping of the policies of the central administration and of which the anticipated outcome of the Reformation struggle was for them to become Calvinist; commensal regimes which, unlike the previous three have no governor but were closest to the centralist form ‘in that persons who exercised gubernaculum did so only because

they were members of the entire body politic or because, among such members, they were judged to have certain skills needed by the regime’,” and since they did not serve the interests of special groups the degree of the immanence of the body politic in the regime was

maximum, which Swanson takes to mean that a Roman Catholic outcome was to be expected; and, finally, heterarchic regimes in which

‘ultimate authority in gubernaculum was vested in persons only as representatives of constituent bodies in the political community’ and since this is a situation in which the immanence of the regime in its community is minimal Swanson expects a Calvinist outcome. Swanson

devotes about one-half of his book to an examination of these five kinds of regime and to each of the forty-one regimes in his sample. He concludes that on the whole the expected correlation holds with very few exceptions and very few doubtful cases. In referring to certain

doubtful instances, Flint remarks” that

*' Taking advantage of Flint’s convenient summary, ‘A Handbook for Historical Sociologists’, p. 500f.

* Tbid., p. 500; his italics. *8 Swanson, Religion and Regime, p. 212.

* Flint, ‘A Handbook for Historical Sociologists’, p. 502.

IDEALS OF ORDER: HISTORY AND SOCIOLOGY 73

... There are several... cases... which proved problematic as to type of regime but in most cases the alternative decision would have been congruent with his general thesis. ‘hat is, it made no difference to the outcome of his general historical prediction whether a regime was classified commensal or centralist on the one hand or balanced or heterarchic on the other since the anticipated religious outcome would have been the same.

This is, surely, a striking thing to say. We are offered a five-part classification of types of regimes and now we are told that in some instances it really does not matter whether a particular regime 1s classed in one way or another.* The reason, of course, is whichever is correct, commensal or centralist, balanced or heterarchic, the prediction concerning the outcome of the Reformation struggle that Swanson would make would be the same, and it is one obtrusive

feature of sociological practice to emphasize the central place of prediction among the goals of social-science research. But prediction

apart, surely it must matter very much whether a regime is one or another of these types, and not simply because accuracy of classifica-

tion would, in fact, facilitate prediction. This, in any event, would seem to be what leads Bousma to make the following remarks:* For the historian any satisfactory explanation for the confessional choices

of the sixteenth and seventeenth centuries must be based, not on a general theory, but on the specific circumstances of the settlement in each separate case. he historian asks: Who or what groups influenced the decision, for what motives, and under what conditions? And again, what special circumstances made the decision stick? ... From this standpoint, for example, Spain remained Catholic not because the government was ‘experienced’ in a peculiar way by Spanish society but because of a tradition of militant Catholicism associated with the medieval * Somewhat earlier (Ibid., pp. 498ff.), and in private conversation, Flint suggests that the independent variable of Swanson’s theory is really the degree of immanental penetration, that his classification of regimes is constructed along what Flint calls ‘a “degree of penetration” dimension from least penetrated (commensalist, centralist) to most penetrated (balanced, heterarchic) with limited centralist regimes somewhere between these two extremes’, and that, in effect, the whole of the classification is an attempt to provide a criterion or ‘measure’ for degree of penetration. The point is that this particular measure could fail and yet degree of immanental penetration be

indeed the independent variable. If this is the case, then Swanson has misstated what he was after when he says, ‘We are looking for the characteristics of a central government that may be associated with the belief in the immanence of God’ (Religion and Regime, p. 36; italics added). And Flint admits that if Swanson continues to use the classification in the way he does in the book then he 1s responsible to show that

the co-variation of religion and regime is as he says. *© Bousma, op. cit., p. 4809f.

74 IDEALS OF ORDER: HISTORY AND SOCIOLOGY crusades, and because the reforms of Cardinal Ximenes reduced kinds of dissatisfaction that operated elsewhere. ...

Other examples follow, and it begins to be clear enough that some interesting ideas would be made explicit were we to attempt an analysis of the differences between historians’ questions and sociologists’ ques-

tions. Implicit in the quotation from Flint is that it is almost enough

to be able to predict that Spain remained Catholic. In the case of Spain there is no question that in terms of Swanson’s classification the Spanish regime was centralist, but that is not as important as the possibility the theory offers for successful prediction. Bousma, on the other hand, finds that the account offered by Swanson fails entirely

to produce the sort of thing he recognizes as an adequate causal account of the outcome of the Reformation struggle in Spain and elsewhere: it does not make the outcome comprehensible. That in some instances i does not matter if the regime is centralist or commensal only adds to the difficulty of seeing the relevance of such work to an attempt to understand the course of events in question.*’ Let me briefly sum up the general drift of this part of the paper. Though I have not attempted to present formal characterizations of the ideals of order implicit in Flint’s acceptance and Bousma’s rejection

of the kind of work represented by Swanson’s study, and I am by no means certain that such an attempt 1s feasible, 1t has been possible to specify certain conflicting intellectual commitments to which one or the other of them subscribes. For Flint and Swanson the social world is made up of discrete parts which may be correlated with one another in the attempt to discover regularities or co-variations, whereas for Bousma the historical world reflects the sort of continuity which the Swanson study must inevitably destroy. Reflecting the positive-science

tradition, which has deep roots in sociology, Swanson both treats *” IT do not wish to suggest that Swanson is a simple-minded collector and correlator—the complexity of his study certainly would belie that. In addition, he does not want merely to stand on correlations which result in successful predictions, but does consider it part of his problem to seek an explanation for what he discovers. (Cf. Flint, ‘A Handbook for Historical Sociologists’, p. 502f.) After some discussion of this same point, Flint concludes; ‘My point is that Swanson, in making the attempt to formulate a more general theory to account for what he regards as an empirically valid co-variation in forms of regimes and final Reformation settlements has provided an example of a highly desirable mode of procedure for comparative/propositional studies in society and history’ (Ibid., p. 505). Ideally, this would be the explaining of a low-level generalization or correlation by deriving it from a more general theory. But none of this would help Bousma or enable him to see how a general theory casts any light on the separate cases.

IDEALS OF ORDER: HISTORY AND SOCIOLOGY 79

religion as an epiphenomenon and tends to overlook the possible causal-explanatory importance of religion,*® whereas for Bousma Christianity has an integral history of its own and some features of that history make no sense whatever without special attention to the

careers of Luther and Calvin. For the former the central goal of social research is prediction and this is understood in such a way as to make imperative the discovery of classes of recurrent phenomena, while for the latter the focus of attention is upon the understanding of the nature of some unique course of events. It would seem, then, that not only do the two sides represent different ways of dealing with socio-historical data, but their respective conceptions of the nature

of the material to be studied differ as well. And so, far from being surprising that the reactions of Bousma and Flint to Swanson’s book are as opposite as they are, the really striking thing would have been for them to be in substantial agreement.”

Ii It would thus appear that history and sociology are so fundamentally

different that it is not at all easy to see either as relevant to the other. Indeed, it would take no small effort to show that they are two different ways of dealing with the same more fundamental subject matter, for, as we have seen, in some respects these two ideals

of order say some rather remarkably opposite things about their respective subject matters. Presumably, an effort to overcome the * Flint, too, thinks Swanson has gone too far on this point and clearly disassociates himself from the latter’s views: ‘It may be programmatically relevant in these few comments on his dependent variable to note here his apparent view of religious role, ritual and belief systems as nothing more than that—dependent variables in the social historical process. ‘The Roman Catholic Church as an institutional struc-

ture barely exists in this account of the Reformation except as one doctrine attribute in his dependent variable. Neither the Marxists, the Freudians, nor the potential Swansonians have demonstrated (nor, it must be admitted, consistently claimed) that the religious factor is purely epiphenomenal—always an outcome of something else and never at least one of several elements in an historical determinant... .’; ‘A Handbook for Historical Sociologist’, p. 506. * At this point, I should like to acknowledge my gratitude to John Flint for reading this paper and for discussing with me the points it makes, particularly in the present part. He would like it known that while there is a broad range of agreement between Swanson and himself, particularly with respect to the objectives of historical sociology, it is not by any means as thorough as my discussion may make it seem. He certainly takes the autonomy of religion far more seriously than Swanson does.

76 IDEALS OF ORDER: HISTORY AND SOCIOLOGY

opposition would involve the treatment of the two ideals of order as being essentially methodological and in no way entailing ontological commitments with respect to the nature of the subjects to be studied. Whether or not this could actually be done—starting, presum-

ably, with a more rigorous formulation of each of these ideals of order than I have found it convenient or necessary to produce, and working toward a re-analysis into methodological terms of every assertion which seems to be more fundamental than merely method-

ological—I am not prepared to say now. But I should like to end this paper on a note of rapprochement between history and social science in some other way.’ So far as I can tell, in what follows I make only one fundamental assumption, which I shall not argue to defend, and that is, that it is not possible to argue from a particular to a particular. Consequently, any explanation of some occurrence in terms of some other occurrence or set of occurrences needs some-

thing general in addition in order to make it fully satisfactory. Of course, I know that this is by no means the unanimous view and that any number of writers have offered examples to the contrary. My stance with respect to such examples is to say that the everyday language in which they are formulated encapsulates our general experience with regard to the matters they deal with, and even though that general experience is not denoted by a sentence all its own it is

part of the argument nevertheless.” In spite of what I have just said, I do believe that the explanations that historians offer in the course of their work have to be taken seriously notwithstanding the fact that they are almost never accompanied by general theories of any sort. Purists among defenders of the covering-law model of explanation in history might wish to withhold the dignity of ‘explanation’ from what historians typically offer as explanations, but I see no advantage to antagonizing a large group of people. ‘Thus I have no difficulty at all in considering that each of the following sentences contains an historical explanation:* ... Lhe Italian city states remained Catholic through a combination of proximity to Rome and political pressure... England broke away from Rome... because schism seemed in her case more effectively to *° The position to be sketched in what follows has been presented in somewhat greater detail in my “Theory in History’, especially Part 3, pp. 31-8. (In this volume PP: Tor ) example, see my ‘Evidence and Events in History’, Philosophy of Science, 29, 1962, 175-95; p. 183f. (In this volume p. 14.) ** Bousma, op. cit., p. 490.

IDEALS OF ORDER: HISTORY AND SOCIOLOGY 77 protect native ecclesiastical traditions than what presented itself as a novel obedience, and some elements in English society later turned to Protestantism (a very different matter)? for an intricate mixture of motives. Parts of Germany broke away from Catholicism primarily because this served the interest of princes, who could depend upon support generated by a variety of discontents....

And so on for several others. Toward the end of the paragraph from which the quoted passages are taken, Bousma adds the following remark: ‘Comparative study is desirable and needed, but it must begin with such particularities as these’.°** He is, I think justified in this, but

we must attend to what such a programme would involve. Since, on the assumption I stated at the start of this part of the paper, the claim that some particular state of affairs was the result of some other particular state of affairs—defined to be as complex as we require—is not logically adequate, what procedures must be adopted for this to be overcome? If we were to take seriously some of the examples discussed by advocates of the covering-law point of view, we would have only to generalize the case before us. ‘Thus, if we attribute the existence of state of affairs 5 to the antecedent existence of state of affairs a, we would round out the requirements of a full explanation by adding such a general law as ‘d’s always follow a’s’ or ‘b’s usually follow a’s’ or some such thing. But all this is too easy and cannot be expected to produce theories that are interesting, or even true.” Instead of quickly turning to a contrived generalization, one ought rather, to make some effort to turn to the ‘particularities’

Bousma speaks of in order to discover what about the explanation the historian offers makes it seem reasonable. In other words, one might focus upon the general features of the particular explanation, features which in other forms might be expected to turn up elsewhere and provide some test of the original explanation. This, of course, is not as easily done as the simple generalizing of the given explanation,

and there is no reason to think it will be done correctly at first.* Too, what seems to be a reasonable explanation of some event in the opinion of one scholar may not be in the opinion of another, who, in turn, may have an explanation of his own. But one way in which to seek out what the general features of the explanation are 3 Presumably, the parenthetical remark is inserted because Swanson treats Anglicanism as an alternative to Lutheranism in his theory. * Cf. my ‘Theory in History’, pp. 31-8. (in this volume pp. 12-16.) 58 Ibid., pp. 25ff. (In this volume p. 36ff,) °° Cf. Norwood R. Hanson, Patterns of Discovery, Cambridge 1958, pp. 72 and 84.

78 IDEALS OF ORDER: HISTORY AND SOCIOLOGY

might well be to attend to why it is regarded as reasonable. What are the considerations that make it seem reasonable to believe that this was the crucial factor? Why factors such as these, not others such as those? By attending to details we are, if we are lucky, able to move in the direction of comparative studies. ‘This is not merely possible; there

are already actual instances of this sort of thing. I shall here only cite—without discussion—one. Max Weber’s famous study of the nature of capitalism?’ is a typical product of fustorical research. A particular state of affairs, the spirit of capitalism, is explained by reference to another particular state of affairs, the Protestant ethic. Having offered his explanation, Weber then sought to justify it theoretically, and, as he says himself most explicitly, it was this which led to his comparative studies.”* It is his conception of the general features of the particular explanation of the spirit of capitalism which sets the stage for the comparative theoretical investigation. This enables us to see a way in which theoretical social science is relevant to historical research in a way which seemed impossible from

an earlier juxtaposing of Flint and Bousma. But it is a relevance rooted in taking seriously what in particular historians claim to be the case. Historians and social scientists suitably prepared could go seriatim

through each of the explanations Bousma offers for the outcome of the Reformation struggle in a number of European countries and seek to discover why it seems plausible, what its general features are. The stage would then be set for all manner of theoretical inquiries. Our point of departure for the present discussion has been the conflict of ideals of order, and one may suspect that if the sort of rapprochement indicated in the preceding few paragraphs were to

be realized that ought to be reflected in some ideal of order too. Given what we have taken ideals of order to be, this is necessarily the case. In bringing this, discussion to a close, I should like to indicate two modifications which might be expected in the elements we have

discussed: I dare say it is in the spirit of rapprochement that each of the conflicting ideals of order discussed in the previous part of the paper contributes one. Before we proceed with this, however, I must make clear just what >? Max Weber, The Protestant Ethic and the Spint of Capitalism, tr. T. Parsons, New

York and London 1930; third British impression, 1950.

8 Ibid., p. 27.

IDEALS OF ORDER: HISTORY AND SOCIOLOGY 79

it is I am doing. I am not taking the first steps toward the construction

of an ideal of social order. Such an effort would be to reverse the order in which these things are normally done, and it is hard to doubt that both historians and sociologists would wonder about the presumption of someone who is neither one nor the other telling them what to assume in doing their jobs. My point is that 2f there

is to be a rapprochement between history and sociology or the emergence of a discipline which is both historical and sociological, it will involve or presuppose an ideal of order which is somewhat unlike the two I have been talking about. What precisely the ideal of social

order in question will be like must await actual developments, and my remarks in what follows are simply speculations. Obviously, my

speculations have the character they do because I think that the elements of the ideals of order discussed above at which they are directed are dubious enough and sufficiently dispensible to be eliminated with profit.°? In any event, the first element to be discussed is the sociological conception of regularity as requiring classes of recurring events. It is surely this conception that leads a scholar like Swanson to begin not with the explanations offered by historians, but with the sort of classi-

fication he himself introduces of types of regime. It now emerges that we can have regularity even when we do not have the sort of classes of recurrent phenomenon social scientists look for: that is clearly

implicit in the notion of general features of particular explanations. I cannot offer any sort of useful account of this notion, and I suspect that such an account had best be based upon an examination of actual instances of the kind of research we are considering. All I shall say

at the moment is that if the idea is a sound one it need not be an article of sociological faith that a social world amenable to scientific investigation must necessarily consist in classes of recurring events. Second, the conception of historical continuity, which I suggested was implicit in some historical work, though clearly not compatible with Swanson’s study of the Reformation, would presumably have to be modified. It is harder to discuss this than the element just discussed, because it is far less clear and quite likely has a multiplicity of possible * It is interesting to note increasing recent interest in the epistemological problems

of the sort of rapprochement we are talking about; cf. Charles S. Maier’s review of Rolf Dahrendorf, Society and Democracy in Germany, History and Theory, 8, 1969, 132, and Gert Miiller’s review of Talcott Parsons, Soczetzes: Evolutionary and Comparatwe Perspectwes, ibid., pp. 1544f.

80 IDEALS OF ORDER: HISTORY AND SOCIOLOGY

forms. In addition, it is not really all that hard and fast particularly if history is actually pursued on the assumption that particular persons and individual circumstances really make a difference. I would only

suggest that any looking back at the course of the history of such a complex historical existent as Christianity in such a way as to ignore that at each stage it was shaped by the sorts of factors social scientists are wont to deal with—but rather which inclines to see the continuity of the historical existent as some sort of necessary unfolding—is

not likely to fit into the ideal of order necessary to the rapprothement we have been discussing. A continuity which is contingent, each stage of which being explicable in accordance with the sorts of theo-

ries needed to relate what went before to the specific factors now encountered, is what we are likely to have in our ideal of order.

© My colleague, Theodore Mischel, points out that this is rather like the way in which some course of evolution in biology is explained, stage by stage, by reference to the laws which govern the behaviour of the determinant factors involved.

HISTORICAL EXPLANATION AND THE CLOSE OF INQUIRY Those who follow the literature on philosophy of history know that there is little agreement among those who contribute to it; there 1s certainly much disagreement about the nature of explanation in history. There are those who insist that nothing may with propriety be awarded

the accolade “explanation” unless what is explained may be derived by the principles of deductive logic from what 1s said to explain it. Since those principles of logic do not allow that a particular may be

derived from another particular—or from as large a collection of particulars as you may have the stamina to entertain—explanations are required to include in their explanans at least one statement or proposition which is universal or general, and from this requirement not even explanations in history may be exempted. ‘There are others who note that history has as its subject matter the actions of human agents, from which they conclude that an explanation in this field

requires the showing of how the action was decided upon as the outcome of rational calculation. And there are still others who take it that what is peculiar to history—presumably, what distinguishes it from other forms of inquiry—is that the outcome of historical work is the presentation of a narrative, and for an action or event in history to be explained requires that it be shown to be the believable outcome of the account or story the historian is attempting to produce.

These need not be exclusive, and there are writers who have attempted to combine, say, the first and third or the second and third of these. But that the controversies over explanation in history have centered around these positions and the ways in which different writers

have sought to emphasize one or another of them is surely an old story by now—so well known that to add a footnote full of references at this point would be less an appurtenance of scholarship than of pendantry. But with all of the difference of opinion the ongoing debate reveals, an examination of certain aspects of it indicates a rather pervasive

agreement—an implicit one since it 1s never made a theme of discussion in the lterature—concerning one point about the nature of the discipline of history. Those of ironic bent might be pleased at

82 HISTORICAL EXPLANATION AND THE CLOSE OF INQUIRY

this. And those, like the present writer, who tend to dispair of the claims of realism in epistemology, and more particularly in philosophy of science, and so incline to look toward agreement as a sign of truth, might be expected to greet this news with some hope. Unfortunately, this agreement is ground for neither pleasure nor hope: the

point of agreement is simply false. The point seems to be that a successful explanation in history brings inquiry to an end, for there is nothing further to be asked. If this were in fact the case, it would distinguish historical research from every other form of systematic

investigation. It is generally held to the point of platitude that in science we are forever pushing on. There is no stopping point. Each solution implicates at once a new problem—indeed, a range of problems—and it 1s precisely the mark of a good solution that it leads to new opportunities to advance the field. This, of course, is related to our recognition that science is a collective or a collaborative enterprise.

One researcher picks up the problem where another left off; what one accomplishes leads to what another will do. It does not require overly much familiarity with the field of history to recognize that this is the case for it as well. It 1s certainly not the case that every problem

is complete in itself and disconnected from every other. And it is certainly not the case that the work of each historian is divorced from that of his fellows so that when he has solved his problem there 1s nothing more to be asked. Yet the way in which those engaged in the discussion of explanation in history carry out their work implies just such a view. The view is implicated in different ways by the adherents to the different approaches enumerated above, yet it is

the same view that emerges in each case. I shall now attempt to indicate that this is so. We may begin with the view that explanation in history must rest upon covering laws. Inasmuch as those who subscribe to this view are the ones who are most likely intent upon assimilating historical inquiry to scientific inquiry, it may seem a bit odd that their work should suggest that history is unlike science in some particular respect.

But it will be recalled that what we are talking about is a view of history implicit in the way the historical-explanation issue is discussed rather than any view explicitly held and systematically defended. It might be said that in pursuing the explanation issue, the covering-law theorists simply work with an abstract model of the explanation situation without intending to suggest that the practice of history is actually

chopped up into self-contained units, each apparently independent

HISTORICAL EXPLANATION AND THE CLOSE OF INQUIRY 83

of anything else. But this is clearly suggested by the way the position is presented. Actually, a good many of those who defend the covering-law-in-history point of view are mainly interested in the logic

of explanation and not much in history itself, but inasmuch as the discussion purports to be about explanation im history one may expect that the innocent reader will think that he 1s supposed to be learning

something about the nature of that discipline as well. In typical instances of historical writing, we find that some state of affairs is accounted for by appeal to some other state of affairs. Let it be noted that by “state of affairs’ we need not mean something simple. ‘The causes of what some historian may be attempting to explain may be both manifold and complex, yet there is no harm done, given the purpose in hand, by our using the expression “some other state of affairs” to refer to the entire antecedent complex which the historian thinks explains the situation he is attempting to deal

with. One might say, then, that state of affairs b is explained by reference to state of affairs a; if you like, that (a > 6).' That this is an explanation is rejected out of hand by those who subscribe to the covering-law point of view. If @ is, indeed, relevant to the explanation of 6, it can only be because there is some theory or set of theories—call it Y—which jointly with a entails 6. It seems to me that

what we have said so far can be understood as pointing to one or another of two problems. We might be asking why it is that (a > 9).

Or we may be attempting to explain 6. The discussion of the covering-law advocates almost always make it seem that the business in hand is the latter, because, as we have seen, they reject the former out of hand. It may seem—to those who are innocent of the logic of explanation—that the former is an explanation, but to covering-law theorists it is at best an unsuccessful attempt to explain 5. Thus, for them, the only problem is that of explaining 6. ‘They make no objection to using a in the explanation, but a becomes the initial or bound-

ary conditions associated in the explanation with 7 in order to derive b. ‘The acceptable explanation, then, has the general form of

(7 & a) > Bj. From discussions of the sort in question one does not really get a sense of the practice of history, the way in which different problems and solutions affect one another, and so on. What one gets, rather, ' Of course, “a” and “b” are to be understood as descriptions of states of affairs rather than as concrete states of affairs.

84 HISTORICAL EXPLANATION AND THE CLOSE OF INQUIRY

is the somewhat artificial separation of b, creating the impression that in history there are fixed 6’s waiting around to be explained, and that once an explanation—logically and empirically complete— is hit upon, the matter is brought to a close. The way the problems are treated in covering-law literature obscures the fact that there is no 6 the epistemic status of which is independent of anything else. That there is reason to believe 6 is inextricably related to there being reason to believe, with varying degrees of certainty, all manner of other things. In principle, 7 is supposed to enable us to select the a appropriate to the explanation of 6, and if histonans disagree with respect to which a does in fact bear upon the emergence of 3 it is because, at least implicitly, they are in disagreement with respect to f. But not infrequently, the disagreement among historians is not about how to explain some event, but rather over what the event in question actually was. This may not be apparent to those who think in terms of grand occurrences; surely no historian of seventeenthcentury England doubts that there was an English Civil War. Yet there is no little amount of disagreement as to what the facts were which made up that civil war,’ which means all manner of disagreement as to what the 6’s are which require to be explained. What I have been saying, in effect, is that one disquieting feature of covering-law writing 1s that the actual character of the discipline of history is obscured by it. In reality, we do not find historians all bothered about attempting to explain their collections of disparate b’s. Actually, the distinctive /zstorical problems are not about explana-

tion at all,’ but rather concern the attempt to establish in an epistemically responsible way what can never be known by acquain-

tance. There is explaining galore in history, but the most exciting work involves the attempt to determine—in some detail—what happened. The historian does not have a collection of isolated 6’s; he is,

rather, attempting to establish the historical facts, and that this or * A reading of Christopher Hill’s essay, “Recent Interpretations of the Civil War”, in his Protestantism and Revolution (paperback ed., New York 1964), makes it clear that

this is precisely the nature of the disagreements among the historians whose work he discusses. ‘They do not disagree over there having been a civil war, but they do over at least some of what it consisted in, and not all of the historical facts as these are acceptable to historian A are compatible with the historical universe of historian B.

> Virtually all of the issues discussed in the debate on the so-called problem of explanation in history have to do with the fact that what is being explained is in the field of human action or behavior rather than with action or behavior taking place in the past.

HISTORICAL EXPLANATION AND THE CLOSE OF INQUIRY 85

that may be factually believable is only because of a whole context of historical evidence, historical practice, partially established factual bases, and so on. When the historian suggests (a > 6) it is not because some not-explicitly-formulated 7 leads him to choose a among the many disparate facts he knows as the likely causal antecedent of 0. In some of the more exciting, ongoing areas of historical investigation, one may expect that the very /ustorical existence of the one 1s bound up with that of the other. That there is reason to believe— on whatever criteria the discipline of history provides for such matters, the details of which need not concern us now—in the one 1s more than likely the outcome of the same course of investigation as that which provides the reason to believe in the other. ‘They emerge from the same inquiry and they are bound together in epistemically significant ways which the historian seeks to clarify. In the reality of historical practice, the whole tissue of the relationships of the facts, events and actions being established historically is of concern to

historians, not simply the causal explanation of some isolated 3, the very conceptual isolation of which is reflected in the typical covering-law advocate’s view of how the isolated event is explained: [((T & a) > b]. Thus, the historian’s (a > 6) is not simply a mistake

in the logic of explanation, but rather suggests that what concerns

him is the ways in which a and 6 coexist and are related to one another. Even if one thinks, as I do, that the specific explanations— or “explanations”—of historians do, indeed, implicate general theory,

there is something to be said for preferring “[7 > (a > 6)]” as the

ultimate explanation schema. To be sure, this schema is truthfunctionally equivalent to “[(7 & a) > 4]”, yet it more accurately reflects the historian’s problems than does the other. “(a > 5)” designates the area of the historian’s interest; “b” almost never does. The theories implicated by his work must concern the relations of his a’s and 0’s, and not the explanation of his 0’s. If the way in which covering-law advocates present the problem of explanation in history tends to obscure the actual character of the discipline historians practice and to distort completely the way in which historical problems and solutions involve further problems and solutions, the accounts of their opponents do no better. This may seem a bit strange to some, for while everybody knows that what the covering-law theorists advocate is an ideal of explanation virtually no one claims actually to find overly much represented in the works produced by historians, the claims concerning history that are made

86 HISTORICAL EXPLANATION AND THE CLOSE OF INQUIRY

by their opponents are supposed to be rooted in what is actually to be found in historical writing. Yet in spite of this, the real character of history as the discipline practiced by historians does not emerge from what these opponents of the covering-law position have to say. One kind of opponent claims that what is distinctive about history is that it produces narratives, and that what constitutes an explanation in history is the showing that what is being explained makes sense given its place in the narrative. Of course, historians do produce narratives. And since a narrative is not a pointless collection of facts, if an account is a narrative then the facts with which it is concerned must fall into place in some intelligible fashion. But one could easily exaggerate the role of the narrative in history. To begin with, the narrative is not the only form in which an historian may choose to report the outcome of his investigation. But more importantly, to identify history with what is only one way in which the results of its practice may be reported is to overlook the character of that practice itself. History is not a collection of attempts to weave narratives around

a body of established or believed in facts. It is, rather a discipline which attempts to determine what can be known about the human past, to establish what the facts are in which it is reasonable to believe. When historians are in conflict—particularly, in actively pursued fields of research—it is not so much over what story can best encompass all of the known facts as over what the historical facts are, but this conception of a discipline to which matters of truth and

falsity are relevant simply does not emerge from the pages of narrativists’ account of history. For Morton White, “the question is not what reasons the historian has for believing certain statements, but what reasons he has for including in a chronicle some statements that he believes rather than others that he believes”.* In those issues of history which are actually being considered by historians, one must expect that there is controversy and disagreement, but it is difficult

to know how to understand that on the narrativist view. On that view, each historian tells what story he chooses to tell, incorporating what facts he chooses to incorporate; someone else can tell another story.” A story unfolds, reaches its climax and comes to an end; there

need be no connection between it and anything else. Thus, here, too, we find a conception of history which loses entirely its character * Foundations of Histoncal Knowledge, New York and London 1965, p. 227; his italics.

> This view of the matter is quite pervasive in W. B. Gallie’s Philosophy and the Estorical Understanding, London 1964.

HISTORICAL EXPLANATION AND THE CLOSE OF INQUIRY 87

as a collaborative effort by people committed to a determinate methodology for establishing factual truth. Totally oblivious to the ways in which historical facts are established, it simply takes them for granted

and talks about the way in which they are reported. What is overlooked here is precisely the way in which the historical past is known. There is still another alternative to the covering-law position promi-

nently represented in the hterature of our subject, and this too tends to see historical explanation as bringing some distinct question to an end, thus again losing the actual character of the practice of history. This alternative has a number of varieties, but inasmuch as I do not wish to raise critical questions with respect to any specific formulation, I shall not make any effort to distinguish between them. What is characteristic of this alternative in all of its varieties is its focus upon the characteristics of persons or agents in the explanation of historical occurrences, and its insistence that once an explanation 1s offered in such terms there is nothing more to be asked. By “characteristics” of persons or agents I mean such things as dispositions or motives or intentions.° Different writers will emphasize one rather than the others, but all who subscribe to this alternative are agreed that explanations of this kind are final: the inquiry has come to an end. The model for such a conception of explanation in history 1s the sort of explanation with which we are satisfied in ordinary lie. I find something unusual about something my neighbor has done, and I ask about it. Why did he do it? I am answered in terms of his motives or intentions, and, my curiosity being satisfied, [ have nothing

else to ask. But it surely does not follow from this that nothing else may be asked. Even with respect to everyday matters, it is not difficult to move beyond the limits set by the point of view we are considering. Say, my wife does something, for whatever reason, which 1s a sort of

something that irritates me. I ask why she did it, and she states the reason. For the view we are considering, there the matter ends, but certainly it is possible that on a depth-psychological level the action was done in order to irritate me. Even if the first sort of answer is all some given historian requires to have in view of what he is attempting to do in his particular book or monograph, it by no means follows that other scholars might not find in the work he produces—

® Cf. the well-known papers of Alan Donagan and Michael Scriven in Patrick Gardiner, ed., Theones of History, Glencoe Ill., and London 1959; see also, William Dray, Laws'and Explanation in History, Oxtord 1957, especially chs. 5 and 6.

88 HISTORICAL EXPLANATION AND THE CLOSE OF INQUIRY

the questions he raises and the solution he offers—bases for pushing on in other directions. The depth-psychological suggestion above is precisely the model for this kind of thing, but the point is in no way limited to explanations of individual behavior which may go beyond what an individual is likely to recognize as appropriate to the explanation of his own act. We might imagine an historian attempting to account for the fundamental administrative reorganization of some political unit. Presumably, those who initiate the reorganization, perhaps a ruler and his leading advisors, have certain expectations with respect to the governmental apparatus; it is the failure of those expectations to be realized that motwvates the reorganization. Obviously, the historian has to attend to those expectations as relevant to the account he is trying to construct. Seeking to understand why it is that the administrative

apparatus as they have inherited it is not doing the job they expect of it, the initiators of the reorganization might be expected to study it in some depth and come to some conclusions as to how such an apparatus works and how it might be overhauled so as to be more responsive to their wishes. ‘There is no assurance that they are right or entirely nght about this, but if one is to understand what it is that they were attempting to do one must know what it was were thinking when they instituted first this measure and then that one. And with this, the historian has the elements of a perfectly fine historical essay: a problem to be solved, namely, to make sense of the administrative reorganization that his evidence leads him to believe must have taken place, the discovery of a plausible motive for some sort of reorganization, and the further discovery of beliefs which led to the initiation of the sort of acts which resulted in the particular reorganization which emerged. But dare our historian end his essay with the assertion that there is nothing further to be asked? Surely not. There are many further questions which the conclusions he has arrived at open up even if he himself may not be interested in them. If the people who initiated the administrative changes which set our historian’s problem found that the governmental apparatus they had was not responsive to their aims, we may wonder why. We may ask about what it was they sought to accomplish, why they sought to accomplish it, how their aims may have differed from those of their predecessors, and whether or not these differences might

be traced to differences of other determinate kinds, social, psychological, or whatever. Or, we might possibly look at the government

HISTORICAL EXPLANATION AND THE CLOSE OF INQUIRY 89

apparatus itself, attend to its history during periods prior to the rise of the regime which initiated the changes, notice the extent to which it was or was not responsive to the aims of government during those earlier periods, and discover such things as whether or not a once efficient instrument was geared for new stress, whether or not the personnel of the administrative apparatus had interests of its own other than those of either the political unit or the regime, and so on. In a somewhat different vein, someone could raise questions as to how the various facets of the questions dealt with by our historian are illuminated by sociological theories which deal with bureaucratic organization, the transmission of information and the like, or how our theoretical understanding of such matters is affected by the contribution our historian has made to the historical study of one.example.

Thus, there is no point at which inquiry comes to an end so that there are no more questions that may be asked as a matter of principle. At most, all one might say is that this scholar does not wish to ask any more questions so far as this subject of investigation 1s concerned. On the analogy of the everyday sort of situation referred to above, his curiosity is now satisfied. But the satisfaction of his curiosity is fully compatible with the further sumulation of that of someone else. I have been trying to indicate how so much of recent philosophical

interest in history has tended to project a conception of that discipline which is quite inaccurate. That historians are mainly concerned to explain settled facts and that successful explanation brings the inquiry to an end—nothing anyone explicitly asserts, yet clearly suggested by much of the recent literature—are far from being the case. I have already indicated something of what I think the situation is in

those areas in which historical research is being actively pursued,

and I do not think that there is any need to expatiate upon that here. In any event, what is needed is less repetition by me and rather more familiarity with what historians do prior to that point in their

activity when they decide that the time has come to write up the results of their inquiries. What I prefer to do here is to bring this discussion to a close by observing that few of the writers who have been engaged in the explanation-in-history discussion have any need to be committed to the conception of history which emerges from

their work. |

It is obviously possible to offer a covering-law theory of explanation

which attends to the role that general theories are believed to play

90 HISTORICAL EXPLANATION AND THE CLOSE OF INQUIRY

in such matters, without creating the impression that the discipline

of history is what it is not. If explanation is and must be what covering-law theory says it is, then if we are to have explanations in history they must satisfy covering-law criteria, but the covering-law discussions of explanation in history that we have in the philosophical literature have surely not persuaded anyone antecedently dubious about the covering-law view that there might be something to it after

all. On the contrary, people antecedently dubious about it, and antecedently familiar with the practice of history—say, because they happen to be historians—find their dubiety confirmed by their failure to see in the account of history contained by suggestion in covering-law writings anything that remotely resembles the reality of that discipline. All the talk about explanation i fistory tends to deflect attention from the arguments in favor of the conception of explanation the covering-law advocates wish to defend. And the validity of those arguments are in no way affected by the attempt to exemplify explanation by discussions of isolated facts from the human past.’ Similar considerations obtain for most attempts to offer alternatives to the covering-law point of view. What is usually at stake for those who argue for alternatives has nothing to do with pastness. It

has, rather, to do with the special character of the human subject matter, with the view that human actions, as the results of motives and intentions—not impersonal causal factors—cannot be subject to

general laws constructed on mechanistic models. The arguments offered in defense of views such as these, if valid, affect not only the ’ Not only does the introduction of the made-up examples with which the literature of our subject abounds not really contribute to making the ‘systematic point a

writer may wish to make, it often gets in the way as it becomes the subject of discussion in place of the point it was intended to illustrate. Note, for example, Frederick Olafson’s treatment (History and Theory, vol. 9, 1970, pp. 271-3) of a point

that Arthur Danto attempted to establish. He moves away from the point in order to deal in detail with an example Danto introduced to illustrate the point, and the character of his critique shifts to Danto’s having “not provided a satisfactory example of an explanation in which a law as distinct from a principle of action figures as the major premise” (p. 272) and to Danto’s having “not attempted to carry out” (p. 273) an analysis of one sort or another. But who would normally expect that a professor of philosophy—who is neither an historian nor a social scientist—would be producing the sorts of explanations and analyses Olafson calls for on pain of having his philosophical work rejected? So far removed from reality is the image of the discipline of history which is contained in the philosophy of history of recent years, that examples—good, bad or indifferent—invented by philosophical writers for whatever purpose are treated as if our understanding of them affected our understanding of the discipline of history itself.

HISTORICAL EXPLANATION AND THE CLOSE OF INQUIRY 9]

possibility of explaining past human events and actions along nomological-deductive lines, but also the possibility of offering such explanations for the present-day human phenomena studied by sociologists, political scientists and psychologists. ‘The mechanism they find lurking in every suggestion that covering laws have application to the human sphere must be the object of their attention, as of course it is. But this can be done without the distorting image of history as essentially concerned to offer explanations of isolated past actions. Narrativists, of course, are less open to the charge that they isolate what is to be explained since they seek to put what is explained into

a context of fact, event and choice in terms of which it becomes intelligible. The recent interest in narrative history and its character is no doubt motivated in the same way that the emphasis upon motive, disposition and intention was, namely, by a belief that covering-law

explanation is possible only in domains which are mechanistic, and that to believe that covering-law explanations may be had—much less must be had—in history is to be committed to a mechanisticdeterministic view of man which flies in the face of our experiences of ourselves. ‘The recent explications of the nature of narrative are attempts to offer alternatives to the covering-law view by showing how human actions are rendered intelligible—hence explained—by

being seen as the reasonable outcome of a course of events. It is clearly possible to argue this way even if one does not go to the extreme of identifying narrative form with the essence of history. One could simply insist that sometimes—even often—historians produce narratives, and that these contain valid explanations which make no

use of covering laws. ‘This argument, however, would not require that history be essentially narrative. Nor would it require that history be understood as being pursued by writers working independently of one another each trying to tell a story the validity of which is in no way affected by what anyone else is doing.

BLANK PAGE

PART TWO

THE WHAT

BLANK PAGE

A NOTE ON THE STATUS OF HISTORICAL RECONSTRUCTIONS

I

According to Teggart, “Historical criticism yields only isolated ‘facts.’

The academic historian pursues his inquiries ‘in the faith that a complete assemblage of the smallest facts of human history will tell in the end; the labour is performed for posterity.’ ‘The facts having been arrived at, it follows that something must be done with them; and, in point of fact, the academic historian, without further questioning, casts the results of his investigations in the mold of traditional historiography.”’

In discussing some views of the origins of certain Jewish festival lights, I. H. Gaster comments, “Whether any of these theories 1s right cannot now be known; all of them are based on deductions trom fragmentary and inconclusive evidence.”* Here we have expressed a rather common view concerning the nature of historical reconstructions, viz., that they are the results of logical deductions from the evidence. The evidence is said to point, in some sense, to the particular

reconstruction advocated by the historian. But this seems precisely the point that Teggart finds difficult. The narrative produced by the academic historian, on the basis of the facts’ he has established, is a non sequitur.

The point to be made here is that the status of the historical reconstruction is far better understood if we view it as something other than a logical deduction from the facts. This may seem even more reasonable after we have considered the following, perhaps somewhat ' Frederick J. Teggart, Theory and Processes of History, Berkeley and Los Angeles, 1941, p. 32; cf. Gh. V. Langlois and Ch. Seignobos, Introduction to the Study of History,

London, 1898, p. 314. 2 “The True Glory of the Maccabean Revolt,” Commentary, vol. 6 (December, 1952, pp. 529-536, p. 536; my italics. ° Just what these “facts” are need not concern us here. Teggart believes that the application of the various techniques of historical research to documents, coins, and other artifacts does yield some result, and the result is what he calls the “isolated facts.” Whatever they are, they are not the reconstructed historical event, and Teggart’s difficulty concerns the relation between the facts and the reconstruction.

96 NOTE ON THE STATUS OF HISTORICAL RECONSTRUCTIONS

extreme, example. The historical status of the seemingly quite simple Old ‘Testament Book of Esther has more than exercised Biblical schol-

arship. Without presenting their arguments, all of which reflect impressive erudition, I shall state the general conclusions of four repu-

table scholars. ‘'wo of them argue that the work in question reports historical events, one pretty much as they happened in ancient Persia during the reign of Artaxerxes II,* the other as an allegorical account of a religious conflict—between worshippers of Babylonian and Iranian deities—consequent upon some action taken by Nebuchadnezzar, the

Chaldean conqueror.’ The other two see the book as fictional, one as an historical novelette with a nationalistic moral,° the other as containing two items from a possibly larger collection of folk-tales, the difficulties of which are understood as resulting from the canonization of the work before the transformation of Persian motifs into Jewish ones could be completed.’ Inasmuch as all of these scholars begin with more or less the same facts,® it is quite difficult to believe that their different results were all obtained by logical deduction. And this again brings ‘Teggart’s problem into focus: What, indeed, is the relation between facts and reconstruction? Or, if you like, what is an historical reconstruction?

It seems to me more in accord with what gets done in historical writing to treat the reconstruction not as inferred from the facts, but rather as postulated to explain the facts. If the account is adequate

to the evidence, we ought to be able to deduce the facts from the * Jacob Hoschander, The Book of Esther in the Light of History, Philadelphia, 1923. > Julius Lewy, “The Feast of the 14th Day of Adar,” Hebrew Union College Annual,

vol. 14 (1939), pp. 127-151. ° Robert H. Pfeiffer, [Introduction to the Old Testament, New York, 1948, pp. 732747. ’ Theodor H. Gaster, Purim and Hanukkah in Custom and Tradition, New York, 1950,

pp. 6-38. ® I suppose what would be desirable at this point would be a list of the sentences containing the facts used by each scholar in order to see if this is actually so. It may indeed be that it is not quite the case, yet it will surely be admitted that most of the

facts used by any one of the four were used by the others as well. They all had access to the same Biblical text and ancient versions, the writings of the same Greek historians and geographers, the same, more or less, archeological reports, etc. There was, then, sufficient identity of facts to make the diversity of conclusions most amazing if the conclusions were really deduced from the facts.

It sometimes happens that historians not only disagree, but that an element of personal animosity enters into the disagreement. And this is not to be wondered at. For if we take seriously their belief that a reconstruction is deduced from facts, then, if there is no disagreement about the latter but only about the former, it is implicit in the disagreement that each casts doubt upon the other’s powers of reasoning.

NOTE ON THE STATUS OF HISTORICAL RECONSTRUCTIONS 97

reconstruction. The minimum criterion of adequacy we may demand from a proposed reconstruction is that it take account of all the known facts. Sometimes the facts available are far from complete, and so it

happens that more than one reconstruction satisfies this criterion. Readers of the above cited treatments of the Book of Esther will discover how often the different writers seek to put the same fact into the different

contexts required by their respective theories. Nevertheless, there are advantages to this way of considering historical reconstructions that

are not shared by the notion that they are deduced from the facts. If from the reconstruction we may deduce the known facts, we may deduce a good many other things as well. This will enable us to specify objective problems for historical research, and perhaps thus silence the repeated charge that history must always be a subjective discipline. Those historians interested in the particular reconstruction

would be able to learn what sort of facts, as yet not established, would have to be established if the reconstruction—as hypothesis— is to be confirmed. This might be followed by a restudy of the evidence

with particular questions in mind, by decisions to excavate sites as yet unstudied, etc. ‘To be sure, scholars do these anyway, to the extent

allowed by their resources; but the point being made here is that if we view the reconstruction as postulated, not inferred, we may see how to formulate objective questions and to direct our research toward the confirming or disconfirming of hypotheses in quite the same way

that we find in other empirical disciplines. (And, in fact, in history too, for it is sometimes said of an historical work that, while it is wrong, it has stimulated the research that refuted it.) If logical analysis of the reconstruction enables us to determine what must be established in order to confirm it better, it also allows us to determine what will enable us to choose between two or more conflicting reconstructions. To be sure, this will not actually enable us to choose, for some of the required facts may never be, and perhaps no longer can ever be, established. And so we may be forced to such less than satisfactory solutions as just noting the relative confirmation of competing hypothetical reconstructions, without being able to decide

in a more conclusive way.’ But even if we can do no more than note what it is we require to prove without being able to do much about it, at least to know this is to know something important about 9 This involves matters in the theory of probability, which are beyond the scope of this paper.

98 NOTE ON THE STATUS OF HISTORICAL RECONSTRUCTIONS

the status of some realm of inquiry. And future research may yet manage to yield a more satisfactory solution.

I It may be noted that what has been said above about the postulational character of historical reconstructions bears a marked resemblance to Arthur C. Danto’s notion “that stores play an important cognitive role in historical inquiry, that a story is an hypothetical recounting of what happened in a more or less determinate stretch of the past.... In a sense, such a story is a loosely articulated ‘model’ of what happened.... Such a story must satisfy at least two conditions: (1) it must account for all the known relevant traces, or at least be consistent with them. But since an indefinite number of stories may be consistent with the same set of traces but at the same time be inconsistent with one another, there is the further condition (2) that the model must suggest further research, so that the positive or

negative outcome of such research will tend either to confirm or disconfirm the model.”'° It is Danto’s contention that in putting forward the above view he obviates the need to consider the function of general laws in history (p. 21f.), to use the name of a well-known paper by Professor Hempel!! with which Danto explicitly takes issue. Nonetheless, I cannot but feel that the satisfaction of the two criteria specified in Danto’s paper does actually presuppose at least some small number of general laws; whether or not these are formulated by people called “historians” is of no great matter. In the few paragraphs that follow, I shall attempt to show why this seems to be the case. I do not expect to show that it actually is. It seems to me that this last would require a reasonably

full logical analysis of the notion of relevance, and I do not feel prepared to undertake it. The first of the criteria mentioned by Danto is that the historical reconstruction “must account for all known relevant traces, or at least be consistent with them.” In order to know that a given reconstruc-

© Arthur C. Danto, “On Explanations in History,” Philosophy of Science, vol. 23 (1956, pp. 15-30), p. 22; his italics.

'' Carl G. Hempel, “The Function of General Laws in History,” in Feigl and Sellars, eds., Readings in Philosophical Analysis, New York, 1949, pp. 459-471.

NOTE ON THE STATUS OF HISTORICAL RECONSTRUCTIONS 99

tion satisfies this requirement, we must certainly know what relevance

means or what determines that a given trace is relevant. One of the few serious attempts to understand this notion in the context of historical inquiry is to be found in Mandelbaum’s The Problem of Histoncal Knowledge,'* and a few words on his views are in order here. “What

we mean by characterizing such statements’ as irrelevant is that the fact which is asserted by one statement has no bearing on the facts

asserted by the statements which form its immediate context... relevance must be taken as a category of facts, and not as a product

of our apprehension and description of them” (p. 208f.). “For in saying that we cannot understand a statement in a given context we must mean that we cannot see the connection of the fact which the statement asserts with the facts asserted by the other statements” (p. 210). “One fact is relevant to another when they are so connected

that the mind cannot apprehend the nature of the latter without an understanding of the former”'* (p. 211). Finally, “The relationship of relevance cannot be reduced, nor can it be defined in the usual sense, for it is an ‘ultimate’ in our common apprehension of the world” (p. 21 If.) Thus, according to Mandelbaum, the basis of an assertion of relevance is the relevance of one fact to another. ‘The determination that one fact is relevant to another is simplified by the view that relevance is an ultimate and indefinable category in terms of which the common world is apprehended. But does this really make matters simple? ‘The doctrine of necessary connection seems no more obvious

here than in the account of the nature of causality. It is, of course,

quite true that very often we are able to “see” right off that x is relevant to y whereas it is not at all to z, quite in the same way that we may be said to “see” that x is a cause of y but not of z. But as clear as it may seem to be to us on first acquaintance, the self-evident nature of the connection seems to become very weak indeed when we are called upon to provide a rationale for the claim that there is 2 New York, 1938, pp. 206ff. '' The reference is to the preceding sentence: “If we characterize a statement concerning human freedom as irrelevant to a statement concerning the behavior of electrons, or if we characterize a statement concerning Newton’s first law as irrelevant to a statement concerning the course of the French Revolution, it is on the basis of the belief that these entities have nothing to do with each other.” '* The proposition that Billy punched Johnny in the eye may well be relevant to the proposition that Johnny is crying, yet surely the latter may be apprehended by the mind without any reference to the former.

100 NOTE ON THE STATUS OF HISTORICAL RECONSTRUCTIONS

a connection. (In other words, there may very well be a phenomenological sense of relevance or of causality, but these ought not to be confused with the logical sense involved in the various branches of research, as exemplified, for instance, in the quotation from Mandel-

baum in note 13.) And if, in the end, we justify claims that something is the cause of another by appeal to general laws that connect the two, I suppose that we are required to make similar justifications for claims of relevance." Nor does this view require an atomistic view, either of the world or of our experience of it. Many things are experienced together,

and it requires an intellectual effort of a high order to determine what is and what is not relevant to some object of study. When some fact is isolated for inquiry it may well be doubted that it cannot be apprehended without an understanding of other facts allegedly relevant to it. That they are relevant is precisely what is to be established by investigation; since the connection is not inspectable, some other way of justifying the claimed connection must be discovered. And it seems to me that such justification would require the use of general laws.'®

'? I am not really certain that we have to do here with two different problems for I shouldn’t be surprised if it could be shown that causality is simply a special kind of relevance. If claims of causality are supported by showing that the claimed cases are derivable from specifiable general laws together with specific statements of fact (i.e., if we may take the account of explanation in Hempel and Paul Oppenheim, “The Logic of Explanation,” Part I, in Feigl and Brodbeck, eds., Readings in the Philosophy of Science, New York, 1953, pp. 320-331, to be at the same time an account of causation that attempts to avoid the difficulties raised by Hume), and if claimed relevances are to be justified in a similar way, it may even occur to some that the formal explication of one is identical with the formal explication of the other. This would seem a highly paradoxical possibility, presumably to be accepted or rejected in the end depending upon whether one’s commitment is to the analysis of natural language or to the construction of formal languages. '© If by “categories” we mean those fundamental concepts in terms of which we order the world we experience, which is presumably the sense we are supposed to give to Kant’s table of categories, it is difficult to see how we can accept Mandelbaum’s

view that relevance is an ultimate category of facts. Does he mean that it is constitutive of phenomena or of reality—depending upon whether one is a phenomenalist or a realist? It seems to me that any sense of relevance to be considered in connection with history—and this could not be a phenomenological sense—is likely to be a logical relation between sentences or propositions, which relation presupposes some kind of system within which it obtains. It may be noted that the question of relevance has been discussed by John M. Keynes in A Treatise on Probability, London, 1948 (first printed in 1921), pp. 54f. and 147; and in great detail by Rudolf Carnap, Logical Foundations of Probability, Chicago,

1950, Ch. VI.

NOTE ON THE STATUS OF HISTORICAL RECONSTRUCTIONS 101

It seems, then, that historical reconstructions or true stories, rather than dispense with the dependence upon general laws, must presup-

pose them if they are to satisfy Danto’s first criterion and if my suspicions concerning the nature of relevance are correct. [his does not, of course, mean that historians must now all stop what they are doing in order to turn to the job of formulating general laws. It does mean, however, that their results do presuppose such laws, and that the justification of their work requires that, wherever possible, such laws be made explicit. Finally, without laws I am unable to see how we can ever know that Danto’s first condition has been met. (Nothing short of the actual establishment of Mandelbaum’s conception of relevance in history could obviate this.) As for Danto’s second condition, “that the model must suggest further research, so that the positive or negative outcome of such research will tend either to confirm or disconfirm the model,” its satisfaction seems to involve the same sort of problem as that we have been considering. If a narrative, complete in itself, seems to suggest other things, it must be because events of the kind it is, in the context it may be in, tend as a rule to suggest other things of the same kind. Once again, general laws are presupposed.

THE “ALLEGED” FUTURITY OF YESTERDAY The purpose of this note is to comment on the notion of history and the historical past—as well as on the source of that notion—which clearly informs Richard M. Gale’s “Dewey and the Problem of the Alleged Futurity of Yesterday.”' Gale’s paper is, as the title indicates, actually intended as a criticism of what are said to be some views of Dewey, but I shall not concern myself much with that. Whether or not Dewey’s views of history and the past are open to criticism, the very character of Gale’s criticism indicates that his own views, rather commonly held views I think, may well be subject to review. Under the rhetorical rubric “the futurity of yesterday” Gale has subsumed two very different ideas (p. 503), and we may well begin

by separating them. There is first of all the futurity of the past. If Dewey could be shown to hold a position which entails that the past

of common sense and ordinary language is in the future, or that propositions about them mean something about the future, then he would obviously be in a very difficult position. It is because I doubt that he or any other philosopher has held such a view that I deem the phrase “futurity of yesterday” to be merely rhetorical. It will turn out that what interests me in this note is precisely that—and why— Gale should think that the position he attacks can involve such a view. But first let us quickly note that the second idea subsumed under Gale’s rubric is the futurity of the instrumentality of historical knowledge. This is clearly forward looking, and is a reflection of the usual orientation of Dewey’s philosophy. As is well-known, he was inclined to think that knowledge is pursued because men wish therewith to solve problems. If knowledge of the past is indeed knowledge,

then Dewey must insist that it is intended to resolve a problematic situation, of whatever character. Even if Gale is right and “there are problematic situations which are concerned solely with what took place in the past” (p. 503), these problematic situations are not themselves in the past, and from the standpoint of present inquiry their resolution is in the future. The claim that the utility or instrumental-

ity of historical knowledge points to the future is not on its face ' Philosophy and Phenomenological Research, June 1962, 22: 501-11.

THE “ALLEGED” FUTURITY OF YESTERDAY 103

unreasonable. And it is quite likely that Dewey subscribed to it, just as it is highly unlikely that anyone has even held to the futurity of yesterday in the sense that Gale suggests. Gale talks about “the past” or “past events” and hardly uses “history’ or “historical,” but I think it will be useful to suggest that all claims that such-and-such happened in the past be deemed historical. Thus, all assertions to the effect that something or other happened— clams to determinate knowledge about particular past events which

are, of course, susceptible to confirmatory test through historical research—are historical assertions. Assertions about the past, the mean-

ing of “past,” or the nature of past events are not historical. They are clearly metaphysical and are quite likely often based on the notion of past which is implicit in our ordinary language. My point is that Gale has failed to distinguish between determinate claims to histori-

cal knowledge and claims about the nature of the past. Though Gale makes but infrequent explicit reference to ordinary language and common sense (see pp. 504, 505), his notion of the past is clearly rooted in them and not in historical inquiry. There are clues to this in his paper, the two which follow being sufficient, I think, to make this clear. At one point he says (p. 509) that “we must assume that the present traces of past events have retained their present shape or form without alteration during the period separating these past events from our present. Likewise, we must assume that historical records have remained unaltered during the period of time separating their having been written and our present. These considerations are so obvious that they are often overlooked.” I am afraid that far from being “so obvious,” neither of these assumptions is made at all. The study of historical evidence is quite central to historical research, and historians must be constantly alive to the possibility that their documents are not what they purport to be. That they are or are not what they purport to be and that they have or have not undergone changes are important problems which must be resolved. Historians certainly do not make such simple assumptions about historical evidence.’

More important as a clue to Gale’s notion of the past and its source is the following: “Actually, part... of the meaning of a statement about the past involves its means of direct verification, 2¢., a * Cf. Leon J. Goldstein, “Evidence and Events in History,” Philosophy of Science,

April 1962, 29: 175-94; parts 1 and 2. (In this volume pp. 3-27.)

104 THE “ALLEGED” FUTURITY OF YESTERDAY

set of observational statements describing the experiences an ideal observer who was suitably equipped and stationed in space and time would have had” (p. 507; Gale’s italics). To speak about the verification of statements about the past seems to suggest that we are considering the methodology of inquiry, and this would seem to suggest

further that Gale’s remarks are about the historical. But no reflection on the methodology of historical inquiry could ever yield a statement about ideal observers. Historical inquiry begins with historical evidence,

which is, of course, present evidence, and proceeds to explain that evidence by constructing hypothetical historical events which are compatible with it. I have discussed this in detail elsewhere and will not repeat it here.’ My point is that the notion of an ideal observer who might confirm statements about the past is derived not from the historical events of historical inquiry, but from the common sense notion of past events which may well be imbedded in ordinary language. ‘[’his is the idea of a fixed and immutable past, which once having happened is impervious to change. I do not wish to challenge that notion. Nor do I wish to deny that such a past really was. But I do agree with G. H. Mead “that the metaphysical demand for a set of events which is unalterably there in an irrevocable past, to which these histories seek a constantly approaching agreement, comes back to motives other than those at work in the most exacting scientific research.””*

Gale writes: “When we say that a past event is an actuality, a determination, over and done with, unalterable, inviolable, we are referring to those traits and characteristics expressed by the logical proposition about this event, which is the meaning of a past tensed statement about this event in terms of its logical verification” (p. 508;

my italics). ‘he question is: When who says? Surely not the philosophical student of historical methodology. But quite likely most of us in our ordinary, common sense belief that there was a past which

is complete and is no more. This, however, cannot be said about historical events—those events constructed by historians to explain

their evidence and which are always subject to review in light of new evidence or new understandings of the original evidence—for these are not located in the real past of common sense and ordinary language at all. > Thid.

* The Philosophy of the Present, Chicago, London, 1932, p. 28.

THE “ALLEGED” FUTURITY OF YESTERDAY 105

I said near the beginning of this note that I did not believe that any philosopher has ever, directly or by implication, defended a view which could be called “the futurity of yesterday,” but I think we can

now understand how Gale and others may come to think that they have. It simply represents a failure to distinguish between the real past of common sense and the historical past of historical research. To those who have not paid special attention to the methodology of history—-to which the real past is irrelevant—the former is more natural. And so they tend to think of history as concerned with the real past; they intrude the past of ordinary language into a context in which it has no place. But to be concerned with verifying statements about historical events is to be concerned with history, not the past. To suggest, then, that such statements and their verification are directed toward the future, is not to suggest that statements about the past contain reference to the future in their meanings. It is only because

the historical event of historical research and the real past of common sense are intertwined in Gale’s thinking that he can suspect that what Dewey must surely be saying about the former is actually said about the latter.

HISTORICAL REALISM: THE GROUND OF CARL BECKER’S SCEPTICISM That the historian Carl Becker became an historical relativist is too well-known to require documentation here. But it may be of some interest to wonder why. The point of wondering is not that there is any reason to believe that Becker was so profound a thinker that the course of the development of his thought might prove to be instructive.

: Rather, the point of interest is why so highly-esteemed an historian who devoted so many productive years to historical research and writing should ultimately arrive at a position which says in effect that all that work has no epistemic worth. It has recently been suggested that Becker’s relativism has its origins in similar views of Croce,'

but this is implausible on its face. [ do not wish to say that Becker might not have been aided in the formulation of his relativism by what he found in writings of Croce, but that a practising historian should be converted to a sceptical position concerning what he does by something he read in a philosopher’s book is not easily believed. Unless he had already some inclination in that direction, he would rather likely dismiss the philosopher’s view as philosophers’ views are so often dismissed. In my view, one can find that inclination in Becker’s

reflections on what he was doing as histonan, and the ground of his sceptical views will prove to be a growing awareness that his conception

of the goal of the historical enterprise 1s not compatible with what historical inquiry actually is hke. One may discover this by attending to what he has to say in such general, reflective papers as ‘“Detachment and the Writing of History’ and ‘What Are Historical Facts?” What I want to do in what follows is, first, to present an account of what Becker says on the themes of interest to us here and, second, to move from that account to a consideration of why I think that the views which underlie what Becker came to believe are mistaken. I think that the sorts of view that le back of Becker’s relativism are ' Chester McArthur Destler, “The Crocean Origin of Becker’s Historical Relativism’, History and Theory, 9, 1970, 335-42. 2 Both may be found in the collection of Becker’s essays and letters entitled Detachment and the Wnting of History, edited by Phil L. Snyder, Ithaca, New York 1958,

paperback edn. 1967. Page references given in the text will be to this volume.

HISTORICAL REALISM 107 held rather widely—by historians, by philosophers and by others— but they are not for the most part made the subject of even inchoate reflection. That alone makes it seem worth spending some time teasing them out of what he explicitly says; that Becker was himself an

eminent historian adds to our interest in the matter. But before I actually turn to what he said, I should like to call attention to one point. The views actually presented by Becker are relativistic views,

and it is his relativism which is frequently discussed by others. I, however, shall be concerned with his scepticism. He is in fact both a relativist and a sceptic with respect to the prospects of historical knowledge, but the two are to be distinguished. In the paper referred to in my first footnote, Destler tries to trace Becker’s relativism to the influence of Croce. I rather think that his relativism is rooted in his scepticism, in a way I shall make clear in the sequel, for without the scepticism there would have been no relativism. The relationships between relativism and scepticism are intricate and subtle: to analyse them in detail would take us quite beyond the

limits of our present concern. I do not believe that all relativisms need be rooted in scepticism. / should want to say, for example, that

systems of knowledge are relate to the disciplined techniques or methodologies of knowing in terms of which knowledge is established.

Thus, historical knowledge is relative to those methods of inquiry which constitute the discipline of history. Presumably anyone who took historical knowledge and the ways in which it 1s acquired senously would not expect to be called a sceptic with respect to the possibility of such knowledge, and thus it would seem that one could be a relativist without being a sceptic. And this leads to the delineation of the first of the two terms we are to compare. By ‘scepticism’

one refers to the being dubious about the possibility of there being established knowledge of some sort. I suppose that one may be a sceptic about the possibility of knowledge in general, but such scepticism as something avowed is rather rare in the history of thought. As something with which to charge an opponent it is rather less rare, but we need not go into that. Apart from general scepticism, there are to be found scepticisms with respect to the possibility of specific sorts of claim to knowledge, and these are usually grounded in some kind of epistemological commitment. ‘Thus, one who 1s sceptical with respect to the knowledge claims of mystics is so because he cannot assimilate those claims to the conception of knowing to which

he subscribes. It would appear that the temperament underlying the

108 HISTORICAL REALISM rejection of the mystic’s claim is less one of scepticism per se than of affirmative commitment to some view of knowing which imposes limitations upon what can be accepted by those who subscribe to it. We would expect, then, that someone who was sceptical with respect to the possibility of historical knowledge was so because he took the object of historical inquiry to be of such a nature that there could be no licit way of knowing which could make it known. Suppose one believes that the goal of historical inquiry is to describe some part of the human past as it would have appeared to someone for whom it was present. And suppose one believes that matters of fact are only to be known by acquaintance, and that the only theory of empirical

truth 1s the correspondence theory of truth. It would take little reflection upon what these views committed one to believe to realize that one had little option but to be an historical sceptic. We shall see that some such combination of views lies back of Becker’s scepticism. Although it may be quite impossible to separate the kind of relativism I am about to sketch from historical scepticism, it is not the same thing as scepticism. Even if this kind of relativism always presumes historical scepticism, it can be characterized independently of it. The relativism in question sees the outcome of the historian’s work as largely affected by the historian himself: his interests, values, aims

and so forth. These, in turn, may be affected by the situation in which the historian finds himself: the kind of social world in which he lives, the sorts of problem his community confronts and such like. Becker makes frequent reference to the historian’s milieu and, clearly

enough, that is the sort of thing he has in mind. Thus, it would appear that one learns more about the historian and his age from the work he produces than about the portion of the human past it purports to deal with. And, it would seem, the most useful kind of information to be had by one who sought to understand some work of history would be information about the historian himself: what sort of man he was, what sorts of interest were his, what sorts of problem beset his community, and so on. One may wonder why anyone should hold such views. Or why one who was himself an historian should hold such views. Or, more particularly, why Carl Becker should have held such views. I suspect that he came to hold them because he could not think of how else to deal with the historian

and his enterprise once he found himself an historical sceptic. The root of Becker’s difficulty 1s a position I call historical realism. Not being a systematic philosopher, he did not develop that position

HISTORICAL REALISM 109 in a systematic way; rather he seems simply to have held to it in the way that one holds to most of one’s commonsense beliefs. And, at first blush, historical realism does seem to be the commonsense of history. What, one asks, is the historian supposed to be doing? There

are, of course, any number of answers involving any number of conceptions of the value to us of knowing about the human past: the moral and practical lessons such knowledge may bring, the widening of perspectives which comes with knowing how people other than ourselves lived and thought, and other benefits along these lines. But all this depends upon our actually having knowledge of the human past. Thus, prior to all the values that accrue to us from that sort of knowledge, is the actual acquiring of historical knowledge. History— the discipline—is the means whereby that knowledge is acquired, and its job is to acquire that knowledge with as much accuracy as possible. And what, we may ask, is accurate historical knowledge? The obvious answer would be an account of the past which accords with what an eyewitness could have recorded had he been present when the event

in question took place. An historical account is true, it would be argued, to the extent that it corresponds to the real past as that was

experienced by those people who were there when the past was present. I do not wish at this point to subject this to analytical scrutiny.

I do think that what I have characterized is what generally accords with commonsense and what generally accords with what many historians have vaguely in mind when they choose to become historians. There is evidence in Becker’s writing that some such view was

held by him as well. The ground of his scepticism, it seems to me, is precisely his growing realization that éhzs ideal of historical knowledge

cannot be realized—or at least that there is no way to know that it has been realized if it should be the case that from time to time it is—given what the actual procedures of historical research are. Those reflective papers of Becker which give expression to his relativism are obviously products of a time when he had abandoned any commitment he may have had to historical realism, and that point of view is explicitly rejected in the texts before us. I take it that he himself once did subscribe to that view both because of the vehe-

mence with which it 1s rejected and because I find it difficult to believe that anyone who was ziutially committed to the relativism expressed in those papers would have chosen to become an histonan

at all. I dare say that this is not the sort of evidence that would satisfy an intellectual historian. But in the end I am less interested in

110 HISTORICAL REALISM the actual course of Becker’s intellectual development than in the relationship of certain ideas. What is clear, as I shall show from Becker’s own texts, is that the belief that history is not a way of establishing knowledge is rooted in a rejection of historical realism—

or, rather, in the recognition that the ideal of historical knowledge that historical realism involves cannot be realized. Becker seems to be saying that without the prospect of realizing that ideal, there is no point in looking at history as a way of knowing, a point with which I shall briefly take issue later on. In the essay on ‘Detachment and the Writing of History’, the realistic idea is put down with sarcastic gibes at the ‘modern historian’ who wants to record everything just as it happened (p. 5f.) and for whom the historical fact— he uses the term ‘historical reality’—‘is a thing purely objective, that does not change; a thing, therefore, that can be established once for all beyond any peradventure’ (p. 7). The goal of the writers he criticizes 1s summed up as follows: “The truth, which alone changes not, is what must be got at. The objective reality must be caught, as it were, and mounted like a specimen for the instruction of future ages’ (p. 8). This is to be accomplished by historians who are so detached from the age in which they live that they function rather like automatic recorders than intelligent human beings (p. 9), historians for whom ‘it has come to pass where the historical fact seems almost material too, something that can be handed about and pressed with the thumb to test its solidity’ (p. 10). Becker’s reaction to the point of view he is impugning is not simply a mindless rejection. He has attempted to give some thought to the practice of history, and it is that thought which leads him to the conclusion he has arrived at. But this is simply not reflected in the sarcasm of the essay from which I have been quoting; it is rather more apparent in the one he produced somewhat later, ‘What Are Historical Facts?’, and we may turn briefly to some relevant passages from that. It will be of interest to attend to the way in which Becker

treats two instances of what are said to be historical facts, for that will enable us to get some idea of the way in which the actual practice of history intruded itself into his consideration of the issues of interest to us here. The first of the two is the historical fact that Caesar crossed the Rubicon (p. 44f.). The question for Becker is what makes that an historical fact. I do not wish to examine the character of the confusion which leads him to say that an historical fact is made up of lots of other historical facts—the crossing of the

HISTORICAL REALISM 11] Rubicon being analysed into all manner of actions of all manner of people, thus failing to see that short of an infinite regress—which 1s simply not possible here—he ought to be trying to tell us what the simplest sort of historical fact is. To wander in that direction would be to lose the main point of interest here. What is interesting is the claim that the crossing of the Rubicon is an historical fact because of the way in which it is tied to so many other events of interest and importance. ‘Apart from these past events and complicated relations, the crossing of the Rubicon means nothing, it is not an historical fact properly speaking at all’ (p. 45). He then adds that the crossing is a symbol which stands for the long senes of events involved. He does not make clear what ‘an historical event properly speaking’ is,

and there is no need for us to bother about it. Nor is it clear in what sense the crossing is a symbol of anything, though he will later argue that historical facts are located in historians’ minds or they are

no place at all (pp. 48ff), and to think of them as symbols. would tend to emphasize this mental character which is attributed to them. But this may be to find more systematic integration than this essay actually has. The point that matters for us is that ‘the simple historical fact turns out to be not a hard, cold something with a clear outline,

and measurable pressure, like a bnck’ (p. 45). ‘That is, measured against the actuality of the practice of history, that conception of the aim of historical inquiry which takes it to be the discursive reproduc-

tion of the reality of the human past as it would have appeared to eyewitnesses motivated by nothing more than the disinterested desire to describe what they actually saw? is simply false. It is false because it fails to take into account such notions as importance and relevance. Becker seems to be saying that no historian would care at all about

Caesar’s crossing of the Rubicon except for what that means in its determinate political context: countless people have crossed the Rubicon, but only Caesar’s crossing meant that the Roman Senate was being defied and that its authority as the sovereign voice of Rome

would soon be lost. In point of practice historians have no desire at all to construct a realistic reproduction of the human past, but have the more limited goal of presenting responsible accounts of those

> We need not discuss the difficulties involved in the accurate formulation of such an ideal particularly the difficulties introduced by the problems of perspective—-both spatial and temporal.

112 HISTORICAL REALISM events of the human past which, for whatever reason, are judged to be important. The second example, the assassination of Lincoln (pp. 48ff.), is even more important for us in that it clarifies the actual character of Becker’s scepticism. The discussion of the assassination is in a way rather more radical than the discussion of the crossing of the Rubicon. The latter sumply indicates that historians do not as a matter of practice

want to reproduce verbally the entirety of the real past as it was when it was present, but it says nothing as to the theoretical possibility of such an aim. The treatment of the former, however, basically impugns the aim itself, for its outcome is that no part of the real past ever enters into the historian’s work. To be sure, it does not emerge in so bald a way; it 1s never explicitly asserted as the outcome of the analysis. Yet it is there, as we shall see, and it is that outcome that leads Becker to the statement of historical relativism that the essay contains. Becker begins his account with a statement of the assassination, its place and date, and then goes on to say the following (p. 48): ... Lhat was an actual event, occurrence, fact at the moment of happening, but speaking now, in the year 1926, we say that it z an historical fact. We don’t say that it was an historical fact, for that would

imply that it no longer is one. We say that it was an actual event, but 2s now an historical fact. ‘The actual occurrence and the historical fact, however closely connected, are two different things . . . (Becker’s italics).

The distinction between actual event and historical fact seems clear enough, but those who want to use the latter expression to refer to the actual events of the past may stick a bit at it. But without worrying whether or not Becker was justified in his use of this or that term, let us try to understand what he is trying to say. On the one hand there is the realm of past happenings, the sorts of thing which

took place in the real past and which, it 1s supposed, provide the subjects which historians investigate. ‘he assassination of Lincoln is taken to be one of an indefinitely large number of past occurrences of that sort. But that occurrence is gone. It may never be subject to analysis—the historian may never come up against it. I put it this way advisedly; I want to suggest to the reader Becker’s own rejection of the claim—if it ever was actually claimed in quite this way—that the historical fact is ‘a hard, cold something with clear outline, and measurable pressure, like a brick’ (p. 45). What is left for the historian then? Becker says historical facts. He does not work out in detail

HISTORICAL REALISM 113 how these are to be had, but it is not all that difficult to do. Without saying anything for or against the position—our problem for the moment is the grounds of Becker’s scepticism, not its soundness or unsoundness—what Becker is pointing to is the following. The histo-

rian begins his work by confronting not a real past which he has either to describe or interpret, but, rather, a body of remains which are taken to be evidence for that past. By applying the techniques and methods which are the procedures of the discipline of history to those remains, by thinking historically, as it were, the historian comes to a conception of what it makes most sense to believe took place at some past period of time. It is the event as the historian has constructed it in the course of his investigations that Becker takes to be the historical fact. That is the fact with which he deals and attempts to describe for the benefit of those who take history seriously. ‘Uhe actual event, as Becker understands that, never plays a role in the historian’s work. Immediately after his discussion of the assassination of Lincoln and the conclusions he draws with respect to the distinction between actual

events and historical facts, Becker raises the question as to where

and when historical facts are. They turn out to be in the present and located in the historian’s mind. ‘These are expressions of Becker’s

relativism. The actual event is located safely in the past where it is entirely insulated from the historian and his milieu. Not so the Austoncal

fact. It is the historical fact only with which the historian can work. He has no access at all to the actual event. It is this, I would suggest, that makes Becker a sceptic with respect to the possibility of historical

knowledge. What, after all, is the business of history if not to offer adequate descriptions of actual events of the human past and to assess their significance? It is hard to avoid the conclusion that the more Becker considered the distinction he had drawn between actual events and historical facts the more he came to realize that there was no way of using the former as a touchstone in terms of which to determine the adequacy of the latter. There was, in effect, no way to get at the actual event at all. Given what was supposed to be the goal of the historical enterprise, this 1s obviously a sceptical conclusion. Historians are not, it would seem, engaged in an epistemically licit enterprise inasmuch as there exist no criteria in terms of which to

determine that the outcome of any piece of historical research is successful. Yet there are historians. ‘he subject is professed at universities. All manner of written results of what is called historical

114 HISTORICAL REALISM research come tumbling from the presses. How are we to understand all this activity and its product? It is the logical function of Becker’s relativism to account for the discipline of history once he has reached the conclusion that history as a way of knowing cannot

be sustained. The relativist’s history is present oriented. It is not produced by detached scholars determined to record the actual event

as it was when it was really present, but rather by men who are attuned to their milieu and its needs, and who write history in order to clarify the problems of their own times.

If one takes seriously the distinction between actual event and historical fact, Becker’s sceptical conclusion indeed seems compelling.

But this 1s so only if one is committed at the same time to the ideal of historical realism, if one sees no alternative to it. But there is an

alternative, and I wish to sketch briefly the direction it takes. A complete exposition and defence of the alternative would be far too extensive to tack on to the end of this paper.* What I want to say

here can be dealt with under three headings: (1) the autonomy of history; (2) the mistaken ideal of historical realism; (3) the concept of description.

(1) By the autonomy of history I mean substantially what Collingwood meant, not that the subject matter of historical investigation consists of some special and ontologically distinctive entities, but, rather, that the discipline of history is to be taken seriously on its own terms. A critical philosophy of history determined to take seriously the au-

tonomy of history is required to attend to the way that discipline actually carries out its tasks; it ought not to be too much in a hurry to judge history in terms of criteria drawn from elsewhere. The focus of attention ought to be on what sort of knowledge a discipline practised the way this one is may be expected to yield, not on determining the extent, if at all, to which history can conform to an ideal of knowledge hammered out in the course of philosophical reflection independent of history. Against this, one may readily imagine

it being argued that it is precisely philosophy’s major task to subject claims to knowledge—and, more significantly, claims made on behalf of ways of knowing—to critical scrutiny in the light of whatever criteria

have come to appear plausible. Yet that cannot be permitted to undermine the autonomy of any discipline which has some degree of * | have written, but not yet published, a book called Histoncal Knowing, in which the

themes I am about to sketch, and others related to them, are dealt with in detail.

HISTORICAL REALISM L15 success to its credit. We surely may not become sceptics with respect to the possibility of historical knowledge on the ground that it fails to satisfy some criterion we require claims to factual knowledge to satisfy, overlooking whatever may be claimed on behalf of the discipline

apart from that. In spite of the many disagreements among historians, the fact is that the practice of history has resulted in a good deal of generally-agreed-to knowledge about the human past. It 1s that which gives it the right to have its procedures for the acquiring of knowledge taken seriously on its own terms, and not subject to the criteria of truth or objectivity or factuality which come from modes

of inquiry other than itself. There is one point of fundamental 1mportance lurking behind the distinction between actual events and historical facts which Becker did not have the philosophical capacity to draw. Actual events are experienced or known in perceptual ways, whereas historical facts are established by means of intellectual pro-

cedures rather unlike perception. Most criteria with respect to the nature of empirical truth, factuality or objectivity are drawn from reflections upon the human experience of the present world of nature,

that is, from experiences which are essentially perceptual, and it is not to be wondered at that claims to knowledge established in other ways find it difficult to satisfy such criteria. Becker’s scepticism is a consequence of his realization that historical facts cannot be established

in the way that actual events are known, together with his failure to see that owing to the widespread agreement among historians as to so much of the historical past we might well be interested in articulating the criteria of truth, factuality and objectivity for what is known in the historical way.

(2) ‘The way in which the epistemic ideal of historical realism is mistaken ought to be rather clear now in view of what I have just been saying. Although I think that the point of view I call historical realism is rather pervasive in writings about history, I do not know of anyone of its proponents who has presented and defended it systematically.’ It is simply presupposed in what they write. It would take a disproportionate amount of space were I to attempt to offer such a statement here,° and rather than do that I shall simply indicate > It is never far from the surface in Maurice Mandelbaum’s The Problem of Histoncal Knowledge (New York 1938), but it is never explicitly formulated there.

° | have rather more to say about it in a chapter of the work mentioned in note 4 called ‘Historical Realism and Scepticism’.

116 HISTORICAL REALISM something about it, enough to make clear what its epistemic ideal is. Like realistic epistemologies generally, historical realism sees the process

of historical knowing as involving some sort of interaction between the knowing subject and what is to be known. This tends to involve a conception of empirical knowledge as knowledge by acquaintance

and empirical truth as correspondence. All this may be harmless enough in itself, but owing to the common tendency to take Becker’s actual event to be the object of historical knowledge, we soon find ourselves in a situation in which our conception of factual or empirical knowledge cannot be realized with respect to the human past. I should like to say that historical realism is that point of view according to which the historian must approach the past events he studies

as they might have been approached when they were present. The conceptions of truth and factuality in terms of which the work of the historian is thus treated are such that can only be applied to the sorts of claim to knowledge one might make about things that are present. But while I should like to say this, I shall not really do so here. That the human past is to be treated as if it were present is so patently foolish, that readers would not doubt for a moment that I am making it up, setting up a straw man, in order more easily to press some point or other. And their suspicions would be heightened

by the recollection that I have already said that I know of no one who has explicitly formulated and defended the historical-realist position. I do think that one finds all manner of discussions in philosophy of history which make sense only if the noxious doctrine is presupposed, but to present some examples would take us far from the subject I am dealing with here. And so I shall take advantage of the fact that the point of departure for our discussion comes from Becker and observe that historical realism is the view that the subjects of historical investigation are actual events. Actual events are, of course,

the real past, and forever beyond the scope of investigation. Why should that be? Only because matters of fact are understood to be subject to study only under the sort of conditions that actual events can no longer satisfy once they are past. The epistemic ideal of historical realism leads to Becker’s scepticism. And all this leads to the important conclusion that while it is logically possible to be a sceptic sumpliciter—although there are pragmatic inconsistencies in the formulation of such positions—in point of actual philosophical practice such things are virtually never to be found. Rather, one is a sceptic with respect to certain sorts of knowledge only because one’s non-

HISTORICAL REALISM 117 sceptical epistemology-in-general cannot assimilate claims to such knowledge. But if the area concerning which one is sceptical 1s 1mportant enough or successful enough, it becomes necessary to sacrifice

epistemic ideals to the actuality of ways of knowing. If history has some success to its credit, we must attend to how it operates, not become sceptical with respect to its possibility because it fails to accord with some antecedently held conception of what empirical knowledge is like.

(3) Finally, some remarks on the concept of description. It is frequently said that historical narratives are descriptions of some past course of events. There is nothing technical about this way of speak-

ing, and I certainly do not wish to suggest that there is anything improper about it. But unless one is sensitive to the differences in the ways in which different kinds of description are arrived at, the idea of an historical description will lead to difficulties. It 1s, I suspect, because we think that historians attempt to describe the past that we

think of them as concerned with actual events; but a moment’s reflection on the character of their genesis must make it evident that an historian’s description is totally unlike that of an eyewitness. ‘Vhe latter describes what he sees; no purpose of the present discussion would be served by introducing the complications of perspective and

the like. ‘'o be sure, no one thinks that historians see what they describe, but the use of ‘description’ in connection with the outcome of the historian’s work leads to certain expectations with respect to

it. Someone returns from a place and describes what he saw. We expect that if one of us were to go—or had been there—that is what he would see. And all this is entirely correct given the genesis of the witness’s description. Given that genesis, it is not unreasonable for it

to be claimed that the description is true if it corresponds to what was seen. But is it reasonable to say that same thing about historical descriptions? Inasmuch as they originate in totally different ways from the descriptions of witnesses, clearly not. Historians do not witness what they describe. This may be the merest platitude, yet it needs to be said in view of all those who approach the descriptions of historians as if they have to satisfy the same criteria of factuality, objectivity and truth as do those of witnesses. The historian’s description is the result not of perception, but of certain intellectual operations, those operations which are the methods of historical research. The description is historically true not because it corresponds to an actual event as a witness may have observed it, but, rather, because given

118 HISTORICAL REALISM the evidence in hand and the ways in which historians deal with and think about such evidence it is reasonable to believe that some part of the human past had such-and-such characteristics. It is, in other words, reasonable to believe in these-and-those historical facts. What we need is a theory of historical knowing which makes sense of how we know—and come to describe—historical facts. Any theory which

assimilates historical facts to actual events and imposes upon the practice of history a theory of knowledge which at best is only applicable to knowing the present that is given to perception must lead to scepticism with respect to the possibility of historical knowledge.

A NOTE ON HISTORICAL INTERPRETATION What is historical interpretation? In one sense, it seems rather obvious.

It is the discerning of the significance or the special importance of events of the historical past. It would seem, on the face of it, that to judge that an event was significant in a specifiable way must involve some determinate idea of causation in human affairs, for to say of an event that it was significant in that way is to say that it had a specifiable effect on what was yet to come. I would suppose that the attempt to explicate the relevant sense of ‘having a specifiable effect’ would be the attempt to work out whatever idea of causation in the human sphere that the explicator presupposed. This would seem to suggest that the term ‘historical interpretation’ points to a set of related issues which clearly calls for philosophical clarification. Oddly enough, given the rather widespread use of the term, there is not really very much on the subject in the philosophical literature.’ Although I have

collected no data on the subject, my observation is that all manner of different activities are called ‘interpretation’ even if they in no way resemble what is talked about above. For whatever reason, the contexts in which it seems to a writer that he may be well served by calling the activity he is attempting to describe ‘interpretation’ turn out to be quite varied, though this would likely be less the case if we were to have available a good account or explication of historical interpretation properly so-called. Murray G. Murphey’s recent book, [9], on philosophy of history

contains a chapter called “Historical Interpretation and Historical Theory,” and one turns to it in the hope that at least a beginning might be made in resolving this difficulty. Far from being a term which slips conveniently into his writing, interpretation is an idea to which Murphey has given much explicit thought, so much that he has emerged with a list of five explicit characteristics which he at-

tributes to historical interpretations. [t seems to me that it will be useful to quote in full the paragraph in which this is done:

' There is rather considerably more on the subject of hermeneutics, which encompasses the historical activity of interpreting literary remains of past times and places, but that is not the kind of historical interpretation I am talking about here.

120 NOTE ON HISTORICAL INTERPRETATION If one examines what historians do, one quickly finds that they do not propose theories: they propose what they call “interpretations.” What is such an interpretation and what is it supposed to do? First, an interpretation always contains a causal model—whatever else it may tell us, it always claims to tell us why certain events or sequences of events occurred. Second, interpretations are generally proposed, not as interpretations of evidence, but of “facts” antecedently established from evidence. Third, an interpretation tells us the “significance” or “meaning”

of the events which it purports to explain. Fourth, interpretations are generally presented in narrative form, particularly when the subject dealt with is a change of some sort. And fifth, the interpretation is alleged not to be exclusive of other and equally legitimate interpretations. Each of these points requires some discussion. ([9], p. 102)

What Murphey seems to have done is include within the sense of interpretation he intends to characterize a number of the ways in which he finds that the term is used in the historical literature. ‘This might seem to support my own observation as to the varieties of its use, for it is not by any means obvious that the five characteristics Murphey presents constitute a logically neat package. But that is not really what I want to talk about here, and I think that what I ought to begin with are just some remarks about his list so as to be sure they are fully understood. It is obvious that Murphey does not ignore what I took historical

interpretation to be, for it is included by him in point three. But apparently he takes it that interpretation is an overall sort of activity that includes both the determination of significance and the offering of explanation, and the latter is alluded to in his formulation of point three and explicitly stated in point one. One may suppose that inasmuch as it is listed first, Murphey must take explanation by way of

an appropriate causal model to be the most important feature of historical interpretation. And it would seem, then, that for him inter-

pretation is not something historians do in addition to explaining how it is that events or sequences of them come to be, but is an over-arching activity which includes it. The fourth point recognizes both that historians frequently report the outcome of their research in the form of narratives, and that one of the functions of a narrative account is to attempt to make sense of something by showing how it emerged in the course of events, and how it led to what followed. Thus, point four may be related to both points one and three. And

point five recognizes that an interpretation may be from a given perspective and that the historian whose perspective it 1s may be

NOTE ON HISTORICAL INTERPRETATION 121

fully willing to recognize that there are other perspectives which are

as plausible or as legitimate as his own. But there is no reason to think that this is always the case—do Marxist interpretations allow equal legitimacy to their nvals?—and I wonder if Murphey would want actually to say that the account of a writer who was not tolerant about such matters was not an interpretation in spite of that. I should suppose the question of interpretational tolerance, its range, its limits and its legitimacy, might itself be the subject of explicit investigation.

It is, however, to the second of Murphey’s five points that I want especially to direct attention. He tells us that an interpretation 1s not of evidence but of facts established on the basis of the evidence: by ‘antecedently’ he obviously means antecedent to the task of interpretation. It is rather important to be clear about the difference between the two tasks mentioned here. Every systematic discipline begins with

certain “data” it is called upon to render intelligible, and for history the data are the historical evidence. In Murphey’s view, “Historical facts are not established from pure data—they are postulated to explain

characteristics of the data” ([9], p. 64). What Murphey intends by this is that the facts we come to believe in with respect to the human past are not directly drawn from the documents and other kinds of evidence by any form of logical inference. Rather, they are hypothesized along lines familiar to us from the sciences in order to make sense of the evidence. Since the evidence has such-and-such characteristics only a reconstructed historical event—-or fact—that makes its having those characteristics intelligible could be taken seriously as historically plausible.* He goes on to say that “The function of historical theory is to explain the evidence; and the facts are hypotheses

introduced into that theory because they have explanatory value” ({(9], p. 64). The only thing new here is the use of ‘theory’. There is a kind of historical writing which attempts to evoke the character of an age, and Murphey takes that kind of writing to be the construction of a theory in terms of which specific facts of human behavior * T have no particular quarrel with this kind of historical conventionalism, having myself defended such a view on several occasions: [1], [2] and [3]. And while I still

incline to agree that the position is correct, I think that by itself it is far from adequate. It tells us about the relations between reconstructed historical events and historical evidence, placing particular emphasis on the logical function of the former in the context of historical inquiry. But it leaves completely obscure the character of the intellectual enterprise by means of which the historical past—the historical facts which explain the historical evidence—comes to be known.

122 NOTE ON HISTORICAL INTERPRETATION reconstructed by the historian make sense. ‘Thus, individual instances of human behavior which historians of colonial America are able to tell us about are rendered plausible and intelligible because Perry Miller has produced “a work devoted entirely to the delineation of

the patterns of belief characterizing” colonial New England, and Edmund Morgan has written a book in order “to provide a set of generalizations about the family as an institution in this society which would explain the behavior of its members” ([9], p. 84f.; italics added). These general descriptions of past epochs Murphy takes to be social theories restricted in scope to single societies. It might seem at first blush that Murphey has confused explaining facts and explaining data, and that these theories restricted to single societies do the former— explain specific human acts within the given societies—but Murphey has been quoted earlier in this paragraph to the effect that historical theory explains evidence. And Murphey 1s entirely correct about this.

The historical theories he has in mind are not generalized from observed behavior, they are reconstructed on the basis of historical evidence by historians in just the way that specific instances of human behavior are. Thus both the broad evocation of an age and the specific reconstructions of individual actions are epistemically identical in being

directly the outcome of the historian’s attempt to determine what the human past was like on the basis of thinking historically about the historical evidence. What Murphey says in his second point about historical interpretation is that what has just been sketched in the previous paragraph falls outside of that activity. What he says is that interpretation is not part of that activity which is concerned to make sense of historical

evidence. That means that it is not part of the activity which is concerned to tell us what the human past was like, for it is this activity which explains the evidence: we believe that the human past had such-and-such features because the evidence we have is the way

it is. Interpretation is a subsequent stage of the historian’s work, a stage in which he attempts to find the significance of the “facts” which were first established to explain the evidence. “So this is what actually happened,” we may imagine the historian musing to him-

self, “now what does it all mean?” The attempt to answer this last question is the attempt to interpret the facts and determine their significance.

One way to determine what a thinker has in mind 1s to pay some attention to what he thinks exemplifies the idea he is trying to exphi-

NOTE ON HISTORICAL INTERPRETATION 123 cate. Fortunately, Murphey illustrates at some length his conception of historical interpretation by way of various opinions with respect to the significance of Bacon’s Rebellion, which took place in Virginia

in 1676. I do not know anything about that rebellion other than what I read in Murphey’s account, and I have no independent views to express. But that is of no consequence since our present concern

is to use Murphey’s own understanding of it, and his own understanding of the way in which different historians have dealt with it, in order to further our understanding of what he takes historical interpretation to be. He presents the views of three historians, but it will serve our purposes well enough if we consider only the first two,

T. J. Wertenbaker and W. Washburn. On the face of it, Murphey’s choice of these historians’ views of Bacon’s Rebellion seems a happy one, indeed. ‘Their respective con-

ceptions of its significance appear remarkably unlike. The general situation seems essentially as follows. An Englishman named Nathaniel

Bacon arrived in Virginia in 1674 and established a plantation there. For whatever reason—that is one of the issues which divide Wertenbaker and Washburn—he became a leader of a revolutionary movement which succeeded in driving off the governor, Sir William Berkeley, who yet managed to retain control of naval and military forces. Before the matter could be settled on the basis of military and strategic considerations, Bacon died suddenly’ leaving his followers demoralized and easily defeated by Berkeley’s forces. In Wertenbaker’s view, though the occasion for the rebellion was an Indian war, there were persistent, underlying factors which created the situation that made revolt virtually

inevitable should an occasion for it present itself. These persistent conditions were “the low tobacco prices, the high taxes, the corruption and oppression which characterized the rule of Berkeley” ([9], p. 103), and Wertenbaker takes it that the rebellion was the reaction of Virginians to the imposition of the Berkeley regime upon their rights as free men. During the bref period of Bacon’s sway, the legislative body of the colony passed a series of laws, known as Bacon’s

Laws, which Wertenbaker regards as democratic legislation. As to the basic significance of Bacon’s Rebellion, Wertenbaker sees it as * Apparently, another of those unmotivated occurrences, entirely intolerable in fiction, which make it impossible to accept the view of people like W. B. Gallie, [4],

that history is a species of the genus story and must satisfy the literary criteria of what makes a story acceptable.

124 NOTE ON HISTORICAL INTERPRETATION

two-fold, yet both elements of it are clearly related intimately. “It is only when one recognizes that the ‘real’ causes of the revolt were the economic and political oppression visited upon Virginia by England

and the determination of the Virginians to defend their rights that the course of the revolt becomes intelligible” ([9], p. 103f.). And this, in his view, points to fundamentally determining Virginian attitudes

which will again be given expression a century hence under conditions which result—this ttme with obvious success—in the American Revolution. In Bacon’s Rebellion, then, we see an early expression of the Virginia democracy’s determination not to be put upon by oppressors, and “I'he more one examines the movement which Ba-

con headed, the more its kinship with the American Revolution becomes apparent, for both had as the main principle the defense of American rights” (Wertenbaker, quoted in [9], p. 104). The character of Bacon’s Rebellion as it emerges from Wilcomb Washburn’s account of it is remarkably unlike the one we have just been considering. ‘The Indians, far from providing an occasion for the expression of something deep-seated are precisely what stimulate the revolt, for in Washburn’s view its motive is not the high ideals of Wertenbaker’s account but, rather, a greedy desire to despoil them. “Indeed, Washburn’s book presents Berkeley as the hero and Bacon as the rogue, for Bacon emerges as a hot-tempered, avaricious man chiefly interested in destroying the Indians so he could claim their land” ({9], p. 105f.). What Washburn sees at work here is the “aggressiveness of the frontiersman” ([9], p. 106), the very formulation

of which leads Murphey to propose that there are psychological principles—which he attempts to formulate ((9], p. 106f.)—which underlie Washburn’s account and which Washburn applies to the specific circumstances of colonial Virginia: “The generalizations involved are universal psychological principles applied to a particular context—there is no Spint of ’76, no peculiar destiny of America marching through Washburn’s pages” ([9], p. 107). Murphey’s attempt

to specify what the psychological principles are which inform Washburn’s account is an attempt to make explicit the causal model, which he says in point one of his characterization of historical interpretation, any interpretation will have. What is of greater interest to us here is the way in which Washburn’s whole conception of Bacon, the man, his goals, the character of his opponent Berkeley, and the

nature of the revolt itself, differs from the account of these same matters according to Wertenbaker.

NOTE ON HISTORICAL INTERPRETATION 125

We have, then, if Murphey is right, two different interpretations of the same historical event, but what I should like to ask at this point is: With respect precisely to what do the two accounts disagree? If Murphey is nght, the answer should be easy to arrive at, for the two should disagree with respect to the five points dealt with earlier when I presented his account of the nature of historical inter-

pretation. Actually, it need not be the case that to disagree with respect to the interpretation of historical events means to disagree with regard to each of the five points. Yet, if historical interpretation is what Murphey says it is, we ought to be able to locate the specific disagreements of historians in conflict over interpretation in the various

elements of his schema. Disagreement over interpretation is certainly not an uncommon experience in intellectual life. Physicists and philosophers of science disagree over the interpretation of quantum mechanics, and I suppose this means that there is some agreed-to body of mathematically formulated theory the significance of which is in dispute. Or two critics disagree over the interpretation of a work of art, and this means that they disagree, not with respect to the present reality of the work in question, but as to its esthetic significance. We must ask ourselves what kind of disagreement we find in the quarrel between Wertenbaker and Washburn, at least so far as it 1s presented to us in Murphey’s book. It is possible for historians to disagree over the significance of some

past event concerning the actual character of which they are in agree-

ment. Say there was a national election in the United States, the long range consequences of which are taken to be significant. Take, for example, the election of 1932 which brought Franklin Roosevelt and his New Deal to power. Some may consider that this election

was significant as the point in which the United States explicitly accepted the principle that government must be responsible for the social and financial well-being of its people, and abandoned the view that what government does is badly done, and so it must do exceedingly little. Others may think that the 1932 election was significant as the crucial moment in time when the American people abandoned their political independence for the sake of momentary melioration. No doubt, this is too simple minded, yet it will serve as an example. Both sides to this dispute are to be understood as interpreting the same event having reasonably specific characteristics: the voters went to the polls; so many voted for the winner, so many for the loser;

126 NOTE ON HISTORICAL INTERPRETATION the states and sections voted along these or those lines, etc. We have,

thus, to deal with an event for interpretation which is about as determinate as the equations of quantum theory or the work of art the critics dispute about.

Presumably, Murphey would want to say that the example he presents for discussion is precisely like this: it is a concrete historical event, namely, Bacon’s Rebellion. But when we look closely at the actual character of the dispute between Wertenbaker and Washburn, this conception of the matter simply cannot be sustained. It becomes clear that Bacon’s Rebellion is not an historically settled something with sharply delineated features. On the contrary, if Wertenbaker is right, then at least part of what Washburn says took place in colonial Virginia could not have taken place, and vice versa. It is really not difficult to show that this is the case. Thus, one presents Bacon as a

revolutionary defender of citizens’ rights, the other as a ruthless despoiler of the Indians. One describes Berkeley as a corrupt and oppressive tyrant; the other as not that at all. At least some of the important facts which make up Bacon’s Rebellion, on Wertenbaker’s view of it, are other than some of the important facts which make up

Bacon’s Rebellion on Washburn’s view, other in the strong sense that they are not compossible in the same historical universe. Clearly, then, if there are historical interpretations in what the two historians

write about Bacon’s Rebellion, they are not interpretations of the same event understood as made up of the same set—or for the most part of the same set—or of cotenable sets—of historical facts.* In the second of the five points which Murphey offers as characteristic of historical interpretation, he tells us that interpretations are about historical facts, not about historical evidence, and the presumption which seemed to follow from this was that disagreement in interpretation must be disagreement with respect to the significance of

facts, not over the meaning of evidence. Now we discover in the very example that he presents to illustrate the character of disagreement in historical interpretation, that the historians are in disagree* This situation 1s far from unique. A careful reading of the essay Christopher Hill calls “Recent Interpretations of the Civil War” in [5] makes it clear that what Hill is reporting on is not differences of opinion among historians of seventeenthcentury England over the significance of the English Civil War—what were its consequences and implications—but rather what the facts were which make it up. In virtually every disagreement Hull examines, the point at issue seems to be what the weight of the evidence and the available techniques for dealing with the evidence require that we believe took place.

NOTE ON HISTORICAL INTERPRETATION 127

ment not over what the historical facts mean, but, rather, what the actual historical facts are. The disagreement between Wertenbaker and Washburn is precisely at that stage of historical inquiry which Murphy, in point two, calls antecedent, the stages in which the facts have to be established. In his historical-conventionalist view, the historical facts as they emerge from historical inquiry have the task of explaining the evidence. ‘Thus, what Wertenbaker and Washburn are quarreling about is just in what facts it is possible to believe, given the evidence and historians’ ways of dealing with the evidence. One thinks the evidence makes best sense on the hypothetical assumption—

this would surely be Murphey’s way of saying it, and mine, if not actually that of the historians—that Sir Wilham Berkeley was a corrupt tyrant, and the other that the evidence points to a hypothetical assumption of an opposite sort. Both are presumably engaged in the task of explaining the evidence. In Murphey’s list of five characteristics of historical interpretation, he has affirmative declarative statements only, with but one exception. That exception is in point two where he tells us that interpretation does not deal with the evidence. But now we see that it is precisely at the stage in which what the evidence requires that we believe is at issue, that the dispute intended

to illustrate disagreement over interpretation takes place. How are we to account for so unexpected an outcome? For the most part, there is very little philosophical attention paid to the intellectual processes whereby the historical past comes to be known. Philosophy of history has been much exercised by the question of historical explanation, which means the question of how historical

facts already known are to be explained. And there is increasing interest in historical narration which turns out to mean how historical

facts already known are woven into connected stories. Historical conventionalism, which purports to say something about the way in

which historical events as reconstructed in historical research are related to historical evidence, may seem to be an exception to this, but in actuality it is not. What it does, as I have indicated above in footnote 2, is tell us about how reconstructed events are related to historical evidence—that the logical function of such events is to explain the evidence—but it has nothing to say about how it is that such events emerge from the course of historical research. Thus, it, too, does not provide the philosopher of history with the motive or incentive actually to examine the way in which the historical past comes to be known.

128 NOTE ON HISTORICAL INTERPRETATION

Murphey has obviously not made the characteristics of historical reconstruction a theme for investigation. ‘That is why he can present a dispute which is in fact located at that stage of historical research

as something else. ‘This is not the place to present an account of what the elements are which enter into historical reconstruction nor to attempt to say anything about the logic of the way in which the historical past emerges from the course of historians’ thinking about historical evidence.’ I would only say without discussion here that the kind of inference which is involved in such work is neither deductive nor inductive. It is rather something along the lines that certain logicians have called abductive.® I dare say that no one has had more

to say about abduction than C. S. Peirce. Murphey is the author of a very important study of Peirce’s philosophy (see [8]), and I cannot at all doubt that he would have discovered this for himself had cnly

he turned to the subject. How we come responsibly to know now what happened in a then we can never know by acquaintance is a subject worth turning to. No outmoded restriction of philosophy to the context of justification only’ should keep us from attending to the philosophical excitement of history’s context of discovery.®

° I have attempted to do this in a paper I have recently written but not yet published called “Toward a Logic of Historical Constitution.” (In this volume pp. 171-206.)

© Murphey does have a brief criticism of the way in which historians are said “to distill the ‘facts’ by a critical analysis of documentary materials,” but it is clear that that is less an account based on examination of historical practice than on what is said ‘about that practice in handbooks, such as the classic one produced by Langlois and Seignobos in 1898 (p. 58f.). I have discussed the limited value of such hand-

books in the paper referred to in footnote 5, and it would not be appropriate to present that material here. I would only say that recent scholarship on sixteenthcentury French historiography makes it clear that on the very day that the Langlois and Seignobos handbook came from the press it was already more than three centuries out of date; cf. [7] and [6]. Thus, the conception criticized does not accord with the way in fact that historians emerge with the histoncal past on the basis of historical thinking about historical evidence. ’ A restriction which must keep anyone who accepts it from advancing beyond the limitations of historical conventionalism; cf. footnote 2 above. 8 In a letter to me dated August 19, 1974, Murphey claims that his account was intended to be “descriptive of what historians (or most historians) say mterpretations are” and is not asserted on his own behalf. In view of this it would appear that my criticistns in this paper are directed not at Murphey, but at those ‘historians whose own accounts are the bases of his.

NOTE ON HISTORICAL INTERPRETATION 129 REFERENCES

[1] Goldstein, L. J., “A Note on the Status of Historical Reconstructions.” Journal of Philosophy 55 (1958): 473-479. (In this volume pp. 95-101.) [2] Goldstein, L. J., “Evidence and Events in History.” Philosophy of Science 29 (1962):

175-194. (In this volume pp. 3-27.) [3] Goldstein, L. J., “Collingwood on the Constitution of the Historical Past.” In Critical Essays on the Philosophy of R. G. Collingwood. Edited by M. Krausz. Oxford:

Oxford University Press, 1972. (In this volume pp. 312-336.) [4] Gallie, W. B., Philosophy and the Histoncal Understanding. London: Chatto and Windus, 1964.

[5] Hill, C., Puritanism and Revolution. New York: Schocken Books, 1964. Chapter 1. [6] Huppert, G., The Idea of Perfect History. Urbana, Illinois: University of Ihnois Press, 1970. [7] Kelley, D. R., Foundations of Modern Historical Scholarship. New York: Columbia University Press, 1970. [8] Murphey, M. G., The Development of Peirce’s Philosophy. Gambridge, Massachusetts:

Harvard University Press, 1961. [9] Murphey, M., Our Knowledge of the Historical Past. Indianapolis and New York: Bobbs-Merrill, 1973.

EPISTEMIC ATTITUDES AND HISTORY In the course of trying to understand the intellectual enterprise whereby

historians purport to characterize a human past they can never know by acquaintance, I have increasingly found it necessary to criticize and impugn an orientation toward the historical enterprise which may be called historical realism.’ I call it an orientation—rather than a doctrine or theory—because it is rarely explicitly owned,* much less systematically defended, yet it is quite pervasive in the ways in which people think about history and our knowledge of the historical past. One feature of historical realism is the tendency to think about historical knowing in perceptual terms, the tacit expectation that claims to historical knowledge must satisfy, say, the same criteria of reference, factuality, and objectivity as do claims made concerning the perceiv-

able world in our natural present. This is no doubt owing to the fact that such criteria for those things that we have are the outcome of sustained philosophical reflection on our knowledge of the external world, the world of present-day common sense and its refinement by the natural sciences. Having on the basis of such reflection arrived at seemingly satisfactory criteria of reference, factuality, and objectivity, one expects that any claim to empirical knowledge must satisfy them:

if history is to be taken seriously as an empirical discipline, its claims to knowledge can be no exception. To say that historical realism approaches the claims of historical research in perceptual terms is to say, in effect, that it treats the past as if it were present. Such a charge has always the ring of paradox

about it, but it is quite easy to show that it is true. Consider, for example, the once hotly fought controversy over verification and meaning. There is no need to review that controversy here. It will ' See my “Collingwood’s ‘Theory of Historical Knowing,” History and Theory, vol. 9 (1970), pp. 3-36, (in this volume pp. 273-311.) and “Historical Realism: The Ground of Carl Becker’s Skepticism,” Philosophy of the Social Sciences, vol. 2 (1972), pp. 12131 (in this volume pp. 106-118.); my most sustained attack on historical realism is to be found in Histoncal Knowing, Austin and London, 1976. 2 C. Behan McCullough—“Historical Instrumentalism,” History and Theory, vol. 12 (1973), pp. 290-306—calls himself a realist (cf. p. 291), but the claim with which he ends his paper—that written history “is first and foremost a description of what an historian usually has good reason to believe really happened” (italics added)—is certainly

not a realistic one.

EPISTEMIC ATTITUDES AND HISTORY 13] suffice simply to recall that one argument offered against extreme formulations of the thesis that identified the meaning of a statement with its verification, was that such a view rendered meaningless all statements about the past including those purporting to deal with the historical past. ‘Io meet the force of such objections, verificationists introduced the weaker criterion of verification-in-principle, thus making it possible to treat historical claims as meaningful even under conditions of less than full—or even partial—actual verification. No present-day

writer could actually verify the claim that Caesar was stabbed to death on the Ides of March, but the claim that he was is meaningful because it is, after all, verifiable-in-principle. And what is verificationin-principle? It is that anyone who happened to be located—spatially

and temporally—so that he could witness an occurrence if it took place, would witness it. There may be people who think that this move does, in fact, save history, but those are people who clearly know nothing at all about the ways in which claims to historical knowledge are made, justified, and verified. ‘To talk about the verification-in-principle of historical

statements by means of a procedure which no historian can—in principle—use, is surely very strange. What we have is not a procedure for verifying——even in principle—claims to historical knowledge, but,

rather, a procedure for witnessing events and things which are contemporary with the witness. But no one who could witness an event could be its historian. History is a way of knowing necessitated by our desire to know about things we cannot witness. Historical hypotheses may be formulated, and once formulated they are subject to

verificational tests. To say of an historian’s hypothesis that it is verifiable-in-principle ought to mean that there are—or could be— historical techniques for determining whether or not it is reasonable to believe what the historian claims might have taken place at some time and place. ‘To determine what verification-in-principle must mean

in a context such as this requires that attention be paid to the actual character of historical knowing: how historians determine what sorts of evidence go together with what other sorts, and the whole procedure

of thinking historically about historical evidence in the context of prevailing ideas of what is humanly and socially plausible in order to emerge with a believable account of what transpired in the time and place with which the historian is concerned. The historian’s account

will be written in descriptive, perceptual language and will sometimes be called a description, and this will strengthen the inclination

132 EPISTEMIC ATTITUDES AND HISTORY

many have to think of historical accounts in the perceptual terms required to satisfy the sorts of criteria referred to above. ‘This, however,

is a mistake. The use of descriptive or perceptual language is inevitable. It appears that that is all we have with which. to offer accounts of things happening or taking place whenever it is that they happen or take place.

The more crucial matter is to determine what we are saying when we Call the historian’s account a description. What must be emphasized

is that there is a fundamental epistemic distinction to be made between the descriptions of witnesses and those of histonans. ‘The former

are accounts of things present to the perceptual field of the witness. The latter emerge in the course of determinate intellectual activities which are in no way at all like perception.’ It is trivial and truistic to say that the objects of historical description do not present themselves to the historian’s perceptual field, yet there is nonetheless a marked inclination in philosophical writing to treat them as if they are epistemically like the descriptions of witnesses. ‘The fact that we call them “descriptions” and that we present them in language entirely like that which we might use to present an account of something we actually experienced—perceptually—contributes to the seeming plausibility of the verification-in-principle idea mentioned above. But that is an idea as to how a witness, not a historian, might verify a claim to knowledge.

All this has to do with historical events, with the epistemic status of claims with respect to what is supposed to have taken place in the

human past. As concerns that, the whole thrust of the preceding remarks is that realism and the primacy of the perceptual attitude are inadequate. But the question arises as to how radical we dare be

in the rejection of realism in history. One eminent scholar in this field,* reacting to a draft of the paper on Collingwood cited in note

1 which he had seen in advance of publication, wondered how I would treat the relation of the historian to his evidence in view of my opposition to realism in history. Even if historical events are not capable of being experienced perceptually,’ the evidence with which * There is one qualification that might be made here, that concerning the epistemic attitude the historian has toward those reaha which are his evidence. But that is the subject of the present paper and will be dealt with in the sequel. * Professor Maurice Mandelbaum; unfortunately, I can no longer locate his letter and cannot produce his query in his own words.

’ I suppose that I am expected to qualify this by adding that they cannot be

EPISTEMIC ATTITUDES AND HISTORY 133 he deals are certainly things which fall within the historian’s perceptual

field. It does not seem to me that this is a problem that was particularly pressing in the context of the paper on Collingwood inasmuch as its antirealism emerges in the context of attending to historical events. Those events emerge in the course of historical research, and they are never zdentifiable in perceptual terms as are the sorts of entity which are typically discussed as exemplifying the kinds of claims realists make. But the question about historical evidence is reasonable on its own terms, and I should like to approach the problem by wondering what is actually wrong about a position with respect to it which seems to be so reasonably night.

The position in question may be found formulated in M. G. Murphey’s recent book on historical knowledge.® Murphey attempts to construe historical reconstructions as rather like scientific theories.

Without tracing the parallel in detail, which would not serve the purpose of the present discussion, what is relevant here is to note that in analyzing these reconstructions as theories we must not only recognize that they refer to “objects and events in the past” but must attend to “the consequences of these statements as well” (p. 26). ‘To talk about the consequence of a theory is usually to refer to those possible experiences which will tend to confirm it, and it is precisely this that Murphey has in mind: “And since those consequences which are now testable refer to evidences which exist in the present, some statements of history are statements about presently existing objects and events. It is those objects and events which can be observed by us, and it is these observations which provide the sensory experience upon which all historical knowledge rests” (p. 26f.). I am not sure what is meant here by “events,” but what is of interest for the moment is the suggestion that historical knowledge rests on sensory experience. What will emerge from the course of our discussion is that precisely because historical knowledge does not rest on sensory experience that a realistic account of the relation between histonans and their evidence will not work. But first, I should like to take just a bit more space to

show that Murphey’s advocacy of the view that it does is both experienced by the historian but are by those who participate in them. But one who participates in something does not know that something in the historical way, and inasmuch as I am concerned here—as elsewhere—with history as a way of knowing, I use “historical events” to refer to events which are known in that way. © Murray G. Murphey, Our Knowledge of the Historical Past, Indianapolis and New

York, 1973; page numbers given parenthetically in the text will be to this volume.

134 EPISTEMIC ATTITUDES AND HISTORY pervasive in his thinking about history and not unreasonable given the way it emerges in his various accounts. It is not unreasonable because it emerges from his attempt to treat historical reconstruction on the model of scientific theorizing. If that can be done successfully, then history’s claim to be taken seriously as

a licit way of knowing is surely advanced, for what would be done would be to assimilate history to a mode of knowing the epistemic credentials of which are generally recognized as superior to its own. Thus, for one concerned to defend the epistemic claims of history, to discover the ways in which its practice resembles that of natural science

is clearly to strike a blow on its behalf. ‘This emerges quite clearly in the following passage in which Murphey summarizes a conception of

the nature of a disagreement among historians which he had just been discussing (p. 112):’ ... They did not first establish a set of facts, and then come to different interpretations because they had chosen different facts to start with. They started with certain documentary remains, and, in seeking to account for them, they elaborated a structure of fact and hypothesis which, taken as a whole, made sense of the documentary remains. What is being accounted for is the character of these remains; what accounts

for that character is the whole structure of “fact” and “law.” It is as arbitrary to separate historical “fact” and historical “interpretation” as it would be to separate physical constructs and physical theories. Interpretations explain facts in just the way physical theories explain the behavior of subatomic particles. ‘The particles are not something given which the theory then explains, and historical “facts” are not something given which is then interpreted. Subatomic particles are constructs which are part of physical theories, and historical facts are constructs which are part of historical interpretations. Jn both cases, what is ultemately explained are the observations which we make here and now through our sense organs.

Later he will add, that “Theory gives observations their meaning, and observations give theory its warrant, in history as well as in every other domain of empirical knowledge” (p. 127).

In both kinds of case we have theories which contain or embody constructs and have the function of explaining certain data characterized as observable. On the whole,’ I do not wish to quarrel with Murphey’s conception of the matter, and I certainly do agree that it is the logical function of historical reconstructions to make sense of ’ Italics added in last sentence. ® I do not agree in detail with everything contained in the passage quoted at length, but to discuss those points would be irrelevant here.

EPISTEMIC ATTITUDES AND HISTORY 135

historical evidence. In this regard, it does parallel the function of theory in science to make sense of—explain—the data of experiment and observation. I would note in passing that to know this about the logical function of either of these is still not to know how historical facts or events are actually reconstructed or how theories of science are constructed,? but however they are they do indeed have the task

that Murphey points out. The parallel continues even to the extent that the constructs of the two—the subatomic particles and historical facts—are in principle nonobservable,'” but Murphey does not make anything of it and neither shall I. What I want to attend to here is his claim that the parallel continues even further to the extent that what it is the function of either mode of theorizing to explain are data which provide occasion for sensory experience. Obviously, I can take issue only with the claim that historical evidence provides such occasion. ‘That the data of the natural sciences provide such seems too well established to be challenged here. ‘The character of the sensory experience involved may be less simple than was once thought, and it may well have a large conceptual admixture, yet it seems hardly doubted that when it comes to testing the theories of the natural sciences we must reach a point at which the testimony

of the senses must be taken into account. Qua a datum of science, any object must provide occasion for sensory experience. What I want to claim is that gua historical evidence, it does not. ‘The epistemic

attitude appropriate to historical evidence is not perceptual but hermeneutic.

At first blush, this may seem to raise problems. The things which are the historian’s evidence must be experienced perceptually by him or else he would not know that they were there. ‘he document, old coin, excavated tool, or what have you must be perceived before the

historian can do anything with it or about it at all. I do not mean to disagree with so obvious a truism, yet I do want to say that qua object of perceptual or sensory experience the thing is not historical evidence. And I want further to say that historical knowledge does

not rest on sensory experience even though its point of departure may involve perceivable things. [ want to say something which I ose this, see Norwood Russell Hanson, Patterns of Discovery, Cambridge, Eng., '° The kinds of things—events and behavior—which are historical facts are, of course, observed all the time, but qua historical events and behavior cannot be observed.

136 EPISTEMIC ATTITUDES AND HISTORY | admit has a paradoxical air about it: the things which are historical evidence must be perceivable, but the way in which they are historical evidence is not a perceivable quality. to anyone for whom the object in question is only an occasion for a sensory experience the object does not function as historical evidence. Perhaps this formulation offers a bit of a clue. When we talk of the thing as historical evidence

we are concerned with a certain function it has, its role in a determinate mode of investigation. For anyone who turns his perceptual apparatus in its direction, it may serve as an occasion for a perceptual or sensory experience: the perceptual is the epistemic attitude appropriate to it. But the features, qualities, or characteristics attributed to it gua historical evidence are not perceptual,'' and the epistemic

attitude appropriate to it as such is hermeneutic. To think that historical evidence occasions perceptual or sensory experiences, is to think of the evidence as being the particular realia that may be handled and looked at. It is almost as if to say that the thing is historical evidence as a characteristic of its own. Experiencing—perceiving?—the thing makes apparent to us its quality as his-

torical evidence, and should we manage to make use of it in an historical investigation, the knowledge which emerges from that investigation may be said to rest on our experience of 7¢ among other

things. If this were a sound approach, it ought to make sense to consider the possibility of using the rubric “historical evidence” as a classificatory device in terms of which to order the things of the world. In this respect, it would resemble the way in which other perceptual qualities may function in that way, colors and shapes, for example.

And this would seem to fit nicely into the way in which historical evidence is handled in standard handbooks of historical method. The

tendency of such handbooks is to divide the domain of historical evidence into classes of things—documents, coins, seals, and the like—

and provide recipes or instructions on how to deal with each class.” This way of approaching the matter suggests that there are things in the world which are historical evidence and other things which are '' There are physical and chemical tests which may be applied to objects which facilitate their being located in past time—and these involve perceptual reports of the scientist—but that location is in past time understood physically, not historically, and it is entirely independent of whether the objects prove to be useful to historians aan still informative handbook of the sort in question is Allan Johnson, The Historian and Historical Evidence, New York 1926.

EPISTEMIC ATTITUDES AND HISTORY 137 not. And if we are really talking about things and their characteristics, we seem indeed to be talking about the occasions for perceptual or sensory experience.

I do not wish to quarrel with the handbooks. It is no doubt useful when introducing students to the field of history to present in some orderly and systematic way the sorts of evidence they are likely to encounter during their professional work. Time enough for subtle clarifications at a later stage. But if the handbooks have their pedagogical value, it need not follow that we are required to accept the conception of historical evidence that they may seem to suggest. Nor does the practice of history—and the historical development of that

practice—support such an acceptance. It will prove to be more in accord with the nature of that practice to think of evidence somewhat along the functional lines I indicated above, rather than along lines that see historical evidence as things of certain sorts. Collingwood’s

view seems by far the more promising: “In scientific history anything is evidence which is used as evidence, and no one can know what is going to be useful as evidence until he has had occasion to use it.”!”

And he adds that “nothing is evidence except in relation to some definite question.”'*

What makes dubious the claim that historical knowledge rests on the sensory experience of historical evidence is that the character or “quality” of being historical evidence is nothing that can be sensed. If Collingwood is right, not even an historian can “sense” it in a bit of evidence before it becomes clear to him that it 1s, indeed, evi-

dence relevant to the problem he is attempting to deal with. The very same thing becomes in one context what it was not in another, a bit of historical evidence. There is no point to asking what about the thing is the sensible or perceptual quality which makes it historical evidence or to pretend to go through the stages of looking for it. On the one hand, there clearly is no such quality, and anyone who said he was looking for it would only convict himself of failure to understand what historical evidence 1s. On the other, the very idea of the historical is an idea having historical and cultural moorings,’ ‘5, R. G. Collingwood, The Idea of History, Oxford 1946, p. 280; italics added.

* Ibid, p. 281. '- To be explicated perhaps along the lines suggested by Rodney Needham for the concept of “belief” in his Belief, Language, and Experience, Oxford (Basil Blackwell), 1972.

138 EPISTEMIC ATTITUDES AND HISTORY and for anyone to recognize something as historical is to approach or apprehend that thing in terms of that idea rather than for him to be able to read its historical character out of the thing itself." If anything may be historical evidence, if its being historical evidence

rests upon the possibility that it may be useful in dealing with an historical problem, then clearly its being historical evidence rests upon

none of its perceivable qualities. And in this regard, I do not think one can distinguish between evidence which is such because of the writing or inscription it contains and other kinds of evidence. I say this because I can imagine some wishing to maintain that what I have been saying seems plausible with respect to the latter, but not with respect to the former. Written texts, it might be claimed, are always historical evidence; being historical evidence is an inherent characteristic of the written. There is a difference between written texts and other sorts of things, namely that the former are written texts and the latter are not, but that is not a difference that counts for anything here. Let us treat briefly each of these “kinds” of evidence

in order to see that this is the case. It does not seem difficult to understand that there is nothing perceivable about a thing that we miss when we fail to see that it 1s historical evidence. Without the discipline of history there would be no such thing as historical evidence, and without a curiosity about a past we cannot know by acquaintance nothing would serve as a clue

to what that past was like. In the Babylonian Collection at Yale there is half of a cuneiform tablet which was discovered by archeologists in the ruins of Dura Europos—in eastern Syria—when a brick

from the second or third century A.D. fell and split open. The half tablet served as the core of the brick which was built up around it. And it is said (I do not actually know if it is true) that the monks at Santa Katerina—in the Sinai desert—used to use old manuscripts to start fires. Whatever that ancient bickmaker or those less ancient monks saw when they looked at the tablet or the manuscripts, they did not see historical evidence. One cannot doubt that those things provided those people with occasions for sensory experiences, but it is equally not to be doubted that they did not apprehend them with '6 The anonymous reader of my paper for this journal is likely correct in suggesting that being scientific evidence is no more a perceivable charactenstic than being historical evidence. Yet it is the perceivable characteristics of the scientific evidence which enables it to be used in the way it is, and this is not the case with respect to historical evidence.

EPISTEMIC ATTITUDES AND HISTORY 139 the appropriate hermeneutical attitude in terms of which historical evidence is identifiable as such. When the first Assyriologists in the nineteenth century began to discover cuneiform collections buried in the sands of Mesopotamia, it may well be that they were elated over all of the new historical evidence that they were uncovering, but | think that we had best modify that interpretation of what they had found. That they could think of themselves as having uncovered historical evidence is, of course, owing to their having an historical sense—unlike those monks referred to above. Suspecting that those seeming chicken scratches on the tiles were actually human writing, of course they believed that what they had was of historical importance. Yet strictly speaking, it was not historical evidence. If we asked what it was evidence for, we might have been told evidence for some

aspect of the human past. And while that seems quite correct, it really conveys no information. It tells us about no specific problem concerning which any of the tablets might be relevant evidence. To be sure, until Assyriologists learned to read the tablets—a stage beyond their initial discovery—they could not be used as evidence for anything. That they—and we—might think of them as being historical evidence

from the very start is only because we have the idea of history, the sense of the historical. All that those tablets actually were at the time

of discovery was potential evidence. In which case, gua historical evidence the tablets do not occasion perceptual or sensory experiences. What is given to the sensory or perceptual apparatus of anyone who comes into contact with those tablets are such of their aspects

as color, shape, texture, and so forth. Only those who approach the texts in light of an historical orientation or with a particular historcal problem to solve are capable of apprehending the tablets as historical evidence.

What I have been saying about written texts is exactly—and in the same way—the case for things which have no writing or inscription

on them yet come to serve as historical evidence. At first blush, this

may seem a more dubious claim than that made with respect to written remains, for while many will agree that the writing has to be understood—as distinct from being percetved—there does not seem to be

anything about nonwritten remains apart from what provides occasion for sensory or perceptual experiences. I think that this objection can only mean that there are no sensory or perceptual characteristics which may be ascribed to a thing other than those which provide occasion for sensory or perceptual experiences. And with this,

140 EPISTEMIC ATTITUDES AND HISTORY

of course, I fully agree. But to say this is still not to attend to the object as a piece of historical evidence. For that, one must apprehend the object from the standpoint of an epistemic attitude which is not perceptual at all. Our traditional empiricisms tend to be comfortable only with things the nature of which is exhausted by their perceptual

characteristics, and those who subscribe to them are likely to see these remarks as mystifying question begging. But to identify empirical

knowledge with knowledge which may be reduced to the reports of the senses without remainder has never proved fruitful to those who are seriously interested in objects of knowing other than as things inertly there presenting themselves to our perceptual experience.'’ Neither the historicity of a thing nor its use as historical evidence is a perceivable characteristic. One could imagine some family in an off-the-beaten-track part of Europe occupying a house that had belonged to it for generations and which contains any number of objects

which are very old, even centuries old. The family uses them in whatever ways they may be used: some may perhaps even be useful for the tasks they were originally intended to serve, others may simply be decorative or treasured in whatever way. They are objects with which the family lives in its present. Yet for someone else, whose sensory experiences of it may be entirely like that of any member of

the family, some particular object in the house may be historical evidence, something which bears on the history of crafts or tools or

what have you. Only for those with the relevant preparation is it possible for the object to be apprehended as historical evidence. The epistemic attitude in terms of which one apprehends historical

evidence need not be limited to things: documents or clay tablets, tools, coins, or ruins. It is true that discussions of historical evidence are typically about things of that sort, yet even patterns of behavior might be apprehended in the nonperceptual way we are talking about.

Think of a Roman Catholic family in an out-of-the-way place in Spain, or on an adjacent Mediterranean island, which has the custom of turning to the wall their pictorial representation of Jesus each week before sundown on Friday—the eve of the Jewish sabbath. It seems an odd custom the point of which is quite lost to those whose custom

'’ While this is no place to discuss such matters, to make the perceptual attitude the only licit epistemic attitude is to render unintelligible esthetic, social, and inter-

personal apprehension. With respect to the latter, cf. my “Why the Problem of Other Minds?” Philosophical Forum, vol. 2, n.s., (Winter 1971), pp. 271-277.

EPISTEMIC ATTITUDES AND HISTORY 14] it is, yet for one who Anows—as distinct from merely sees in the perceptual sense'*—that pattern of behavior is evidence that the family in question is descended from Marranos. Since that the pattern may be evidence is not something which could be apparent to everyone,

it is clear that its being historical evidence is not knowable on the basis of sensory or perceptual experience.

In sum, what we have to deal with is an approach to things and behavior patterns to which their perceivable characteristics as such, those characteristics of them which become thematic when philosophers argue for and against realism in epistemology and when they seek to make distinctions among the realisms—say, critical or naive—

are not relevant. Those are not the “characteristics” of things qua historical evidence. At the risk of oversimplification, it may be said that at least one of the basic problems which enters into the debates over realism is that of the subjectivity of knowledge as against the self-subsistence of the world. Realists of whatever sort are concerned

to claim that it is not the case that knowledge is subjective in any self-defeating sense and try to show that it is possible to make responsible and veridical statements about the universe and its parts which are true in the sense of not being dependent upon the characteristics of the knower. The different realisms differ over just what

sorts of thing may be said in that way, but at least they are all agreed on that fundamental point, that the world has knowable characteristics which are independent of anything else. Ihe examples that realists choose to discuss are determined by their respective conceptions

of what it is that is independent in the required way and how it comes to be known. Without entering into a survey of the differences

with respect to these issues, I think it may be said that in all cases what is independent comes to be known in ways which fall within the sensory-perceptual mode of apprehension, which is the mode common to humanity for the apprehension of its common world. But the historical is not part of the common world of humanity,

and nothing gua historical may be apprehended in that mode of apprehension common to all. We have already seen that being historical evidence is not a characteristic of things which prove to be ‘5 Of course, merely perceptual seeing is not adequate for the recognition of behavioral patterns—as distinct from behavioristic movements—but we need not bother about that here.

142 EPISTEMIC ATTITUDES AND HISTORY historical evidence; it is not the independent sort or characteristic we have seen that realists look for. On the contrary, we have seen that the existence of historical evidence depends upon such things as the cultural existence of an historical sense or attitude, the existence of an historical mode of inquiry, and the recognition by historians that something is relevant to the solution of an historical problem. ‘Thus, qua historical evidence the relationship of the thing to the historian is not realistic at all. ‘To think that only a perceptual epistemic attitude obtains between subject and object is to make impossible the understanding of how evidence actually enters into the considerations of

historians as they seek to know in their present a human past they can never know in the perceptual way.

HISTORY AND THE PRIMACY OF KNOWING I

I very much appreciate the opportunity that P. H. Nowell-Smith’s paper, “Constructionism in History,”' provides for me to deal with issues he raises with respect to matters I did not deal with explicitly in Historical Knowing. | might have said that the opportunity 1s provided

for me to turn to the philosophical underpinnings of my book. But J have avoided that way of speaking in order not to beg the question at the very start, and that has led me to the realization that it is a mode of expression I should have wanted to avoid in any event. Nowell-Smith takes it that my book contains a sound account of the way in which historical knowledge 1s acquired, but that my presentation

of it is intertwined with an erroneous philosophy. What he wants to

do is to rescue the former from the clutches of the latter so as to make it acceptable to a larger number of philosophers, historians, and even, I suppose, ordinary people, than those who would be willing

to be called idealists of an extreme sort; constructionism as a philosophy, it would appear, is just an extreme form of idealism. For me to speak of dealing with the philosophical underpinnings of my book—more particularly, of its construal of historical knowing—would

be to reject the suggestion that that construal is open to both of us, to Nowell-Smith and to me: I would be saying that it is not independent of my particular standpoint in philosophy. And that would seem to foreclose from the start the central goal of Nowell-Smith’s effort. In addition, part of what ails epistemology in our day is the tendency

of philosophy to impose criteria of knowledge from outside rather than to seek the articulation of such criteria on the basis of attending to the actual character of the successful acquisition of knowledge. This 1s, to be sure, a complicated matter, and the successful achievement of knowledge may not be obvious prima facie. Yet the notion that philosophers may spin out such criteria by determining the logical

consequences of their pet assumptions and expect that those who ' History and Theory, Betheft 16: The Constitution of the Historical Past, vol. xvl, Nov. 4,

1977; pp. 3-29.

144 HISTORY AND THE PRIMACY OF KNOWING

practice the merely empirical disciplines will acquiesce is surely too much.? We ought to begin where the problems are—not with com-

mitment to realism or an alternative—and then see where we are led. I can say for myself that I did not begin as an opponent of realism. I have come only in the course of time to see that realism cannot make sense of historical knowing. I speak of an alternative to realism, and the question arises as to

what we are to call the alternative to which I subscribe. NowellSmith takes it that my conception of how historical knowledge is acquired is “constructionist,” but he holds that I go wrong in not seeing that constructionism ought to be a methodology only and thus allowing myself to become a constructionist-as-philosopher. I am not sure that I fully comprehend what such a term implies and am rather

reluctant to adopt it as my own, a reluctance reinforced by my knowledge that it is a term coined by one who wants to impugn the position. In addition, “constructionist” may suggest active and intended cognitive effort; but the broad philosophical stance that I should wish to oppose to realism, were I presenting a fully-articulated theory

of knowledge, does not require that instances of knowing be always directed toward some purpose. I would say only that mine is a belief in the primacy of knowing, not in the primacy of reals that are independent of knowing.

I do not doubt for a moment that the universe is a somewhat solid sort of thing, that there are things and people in it, that all of these are the outcomes of courses of happenings or processes which

have roots in a past. I do not believe that the cognitive efforts of human beings are creative in the way that the Western religious tradition takes it that God created out of nothing. And I also believe that we had ancestors who lived and acted in some sort of real past time. Which means, of course, that we are not free-floating in the universe with only the products of our minds to give us solid support. * Although I find myself in sympathy with very little that T. W. Adorno contributes to the well-known volume on The Positwist Dispute in German Sociology (London,

1976), there is one passage I can quote with unqualified agreement: “The epistemologies, as they were developed and handed down by the great philosophical tradition since Bacon and Descartes, are conceived from above even by the empiri-

cists. They have frequently remained inappropriate to the living tradition of knowledge ... By no means the least of the necessary tasks of epistemology . .. would be to reflect upon the actual process of cognition instead of describing in advance the cognitive achievement in accordance with a logical or scientific model to which... productive knowledge in no way corresponds” (p. 111).

HISTORY AND THE PRIMACY OF KNOWING 145

Then why am [ not a realist? Because when we attempt to determine what the world—natural and social—in which we live 1s like, realism

is factually vacuous. It may pronounce that there is a world, but it can say nothing more than that. It may say, with Nowell-Smith, that if the scientist has reasoned truly, then the world really is ike what he

says it is like. But to speak in this way is to confound conceptual truth with determinate factuality: there is no way to reach outside the scientist’s discipline in order to determine that he has or has not reasoned truly. There is no comparing what he says—what the practice of his discipline makes it reasonable to believe—with something brutely real independent of human cognizing activity. To say “if he has rea-

soned truly,” is to add nothing to whatever factuality emerges from the disciplined acquisition of knowledge itself; it is only a grasping for a greater sort of solidity than the acquisition of knowledge 1s capable of giving. Realism is a philosophy of the loss of nerve. To believe in the primacy of knowing is not to eschew reality: I do not think it is inconsistent for me to take the position I do here and yet be interested, for example, in the nature of social reality.

What is to be clarified is the approach to the real. ‘Vhere is no real , anything given to us except by means of an approach to it. Nothing

is added to what we know when we add that if we reasoned or thought or perceived truly about the reality in which we are interested, then we know the brutely real as it actually is. I know the real as it is given to me to know it by means of the procedures available for knowing it.

It is clear, I think, that this position is intended to be entirely general. It is not restricted to the sort of knowledge that is acquired by the systematic disciplines, but claims to be true of whatever common sense knows as well. It is worth emphasizing this inasmuch as

realism often purports to be the philosophy of common sense. It may well be the point of view of what some have called the plain man, yet what the plain man knows about the world in which he lives, what he learns by walking in the woods and chatting on the corner, are what he has come to know; his knowing them is owing to his having come to know them. Of course, it is possible to treat this as truistic, but that would only help miss the point. The point is that even the most everyday, simple-minded kind of knowledge is acquired in whatever way it is acquired, and there is no getting to the object of that knowledge except by means of some procedure whereby we come to objects of knowledge of that sort.

146 HISTORY AND THE PRIMACY OF KNOWING

I cannot expect that in these comparatively few words I have resolved the doubts that a position of this kind raises in the minds of readers, much less that I have effected the conversion of all those who started reading as fully-persuaded realists. But we are not here concerned with epistemology in general; our concerns are with phi-

losophy of history. I have said what I think is enough to give the reader an idea of the orientation to expect. What I hope is that by the time we have come to the end, it will seem reasonable to believe that at least so far as the acquisition of historical knowledge is concerned, that orientation seems plausible and realism does not. II

Even if one could develop a thoroughgoing philosophical position in terms of the primacy of knowing, one could not deny that presentday objects enter into our experience in ways that are determined by their being present-day. My distinction (HA, xiv, 100-102, 134-137) between the account of a witness and that of an historian is intended as an expression of just this difference. The object of the witness’s report is relevant to that report in that it is that object which in the end provides the occasion for the report. Its being there—in whatever

way it is there for the witness—is the first stage in the course of events the outcome of which is the production of the report. If the object is the sort of thing which persists in existence—not an occurrence or event—it is relevant in still another way; namely, it may be

looked at again so as to test the adequacy of the report; it is the touchstone against which to determine the truth or falsity of the report.

I want in this section to consider a number of issues which are implicated by the previous paragraph and which are dealt with at length in Nowell-Smith’s essay. In my book, I make it clear that I take the real past to be irrelevant to the practice of history, and it is obvious that Nowell-Smith finds that incredible. Related to this is the question of what it means to speak of a touchstone for the truth or falsity of historical claims. Finally, the way in which Nowell-Smith

deals with this question rests upon the cogency of the distinction between methodology and philosophy, the point being that since I have not attended carefully to the distinction, I have permitted myself to confound methodological grounds with philosophical conclusions. The distinction as we have it in Nowell-Smith, however, is simply a

HISTORY AND THE PRIMACY OF KNOWING 147 consequence of the realism to which he subscribes, and, as will emerge

in the course of what follows, I find nothing compelling about it; indeed, I reject it. Nowell-Smith seems to think that the real past is relevant to the historian’s enterprise in that it is to the events of the real past that the results of the historian’s work, as summed up in the account he writes, refer. It would be strange, indeed, if the point of the enterprise were not itself relevant to the enterprise. It would be like saying, in considering the Ming vase of Nowell-Smith’s example, that the vase was not itself relevant to what purports to be a description of it. Gould we imagine seriously that in considering the adequacy of

the description we could be forbidden an occasional glance at the vase on the ground that it is the description we are interested in and not the vase? Surely not. But do I not seem to be saying something perilously similar when I insist that the real past is not relevant to the historian’s account? Is there not a genuine parallel with respect to every salient feature of both situations? Can we not admit that just as (if I have described the vase truly) the vase 1s what I have described, so, if the historian has reasoned truly, the real past event is what he has described? And is it not a feature of both descriptions that if they are correct it is because they correspond to the reality of their respective referents?

If I could take more space than in fact I dare, I would try to argue for some kind of parallel by assimilating the vase case to the historical-event case. I would attempt to show that even with respect to presently-presented objects what we can say about them is tied irrevocably to a way of knowing and that the adequacy of our accounts are determined within a framework of knowing. But I shall not do so here. I shall, in effect, allow Nowell-Smith to continue to think that I am willing to be a realist with respect to present objects, because even if realism is an appropriate philosophical stance with respect to our knowledge of them, it is not so with respect to our knowledge of history. By making no attempt to assimilate our knowledge of the natural present to our knowledge of the historical past, I shall be able to emphasize what then seems to be stark differences

between them so that the parallel sketched in the previous paragraph simply breaks down. The key difference for our present purpose is the occasional glance

we may be permitted of the vase. (We can take no such glance at the historical past.) It is, of course, observations of this kind that lead

148 HISTORY AND THE PRIMACY OF KNOWING

to the introduction, by realists, of the distinction between methodological grounds and erroneous philosophical conclusions, but for them

to inject that here would be to beg the question. Hence, I prefer to go on with the case I am trying to make, and we shall see later on whether the realist can introduce the distinction with effect. The point to be made is that there is nothing that corresponds in history to the

object of the glance at the vase. The only way that Nowell-Smith gets that sort of object for history at all is with his if-he-reasonedtruly phrase. But that is factually vacuous. If an historian wants to determine whether a colleague has reasoned truly in the only sense of that phrase that can be interpreted by a methodologist of history, he must simply go through the thinking himself. That enables him to determine whether, given the evidence and the ways in which historians think about evidence of that sort, the conclusions he is examining are sound. What I am trying to say is that all that we

ever come to know about the human past we come to know by means of the application of the methods of historical inquiry.* What we come to test our claims to historical knowledge against 1s never the real past—the counterpart to the vase—to which realists say our

accounts refer; we have no access to that past. There is, in other words, no way to determine whether the historian reasoned truly in the realist’s sense.* It is hard to doubt that there was a real past— or to formulate such a doubt in intelligible language—but I cannot see what role we are to find for it in the practice of history. This is the context for my saying that the real past is irrelevant to the practice of history, and that it is not the touchstone against which to determine the adequacy or not of an historian’s account. NowellSmith takes up the question of touchstone in a discussion of verification, but much of that discussion is not really germane to the issue.

It may seem to be germane, because much of what he says about the example he chooses to discuss, whether there are delphiniums in his garden, seems reasonable. But that example in no way resembles > I omit from consideration claims to such knowledge based on reports of Scriptural revelation. Such claims are not established in the historical way and do not fall within the purview of an account of historical knowing. * This problem is by no means limited to history. What are we to make of the realist’s reality of the world in view of the changing character of natural science’s conception of the world? For the physicist to reason truly means, I should suppose, to work correctly and effectively within the framework of his discipline. If NowellSmith’s view is correct, it may be possible that Newton reasoned truly and it may be possible that Einstein reasoned truly, but it is not possible that both of them did.

HISTORY AND THE PRIMACY OF KNOWING 149

examples from history. (It is a bit disconcerting to be praised for discussing in detail real examples of historical thinking and then be called on to consider the existence of some delphiniums.) The point that Nowell-Smith wants to make is that there is a logical distinction between the existence of something and our grounds for knowing it. He wants to drive a wedge, presumably a logical wedge, between “Is

it true that there are delphiniums in your garden?” and “How do you know that there are delphiniums in your garden?” Related to this is the desire to emphasize that there is a logical distinction between

a description and its referent, particularly in view of my claim that in history this distinction does not exist. But my point is neither logical

nor ontological; I am not put off by the thought that things actually took place in long-gone days. My point is epistemological, and, in the immediate sequel, I shall try to show what it involves.

The choice of examples plays no small part in the character of philosophical discussion, and there is really something to be said for limiting ourselves, in philosophy of history, to genuine examples from the discipline of history. Much of Nowell-Smith’s discussion of verification and the question of the touchstone for truth is clearly owing to the way he thinks about the example he chooses to discuss, the presence or otherwise of delphiniums in a garden. However complex

the character of the experiences which bear on our answer to the question whether there are such flowers in a given garden, and whether or not some sort of realistic account is adequate to the situation, it is not possible to doubt that there is something brutely and compulsively given to the perceiver as he faces the garden. It is that which creates the prima facie plausibility of Nowell-Smith’s logical wedge. That brutely and compulsively given cannot be confounded with the assertion that there are delphiniums in the garden: the assertion is one thing; the perception and its attendant aspects in terms of which the assertion is justified is something else again. In history, there is nothing which parallels the brute-and-compulsive givens of the garden and its flowers. The importance of this is not understood by Nowell-Smith because he has been driving a logical wedge, and the Jogical distinction that he wants to make between an assertion and its ground seems unaffected by it. And he can always find a referent for the assertion by the device of assuming that the historian reasoned truly and giving a realistic interpretation to that. But if one sets aside conceptual distinctions and attends only to the

context within which knowledge is being acquired, the duality of

150 HISTORY AND THE PRIMACY OF KNOWING

assertion and ground in the realistic sense advocated by Nowell-Smith

and others’ just does not add up. It is, of course, still possible to insist upon the distinction between assertion and ground, but in history

this has none of the characteristics we find in the delphiniums-inthe-garden case. There are no brute and compulsive givens; the procedures whereby we come to know about the human past remain always cognitive-constructive; the testing of our conclusions usually requires that the procedures of historical thinking-inquiry be done again, and the status of the past we come to know has no affinities with the kinds of objects that realistic epistemologists discuss when

they deal with examples. This is what I mean when I say that the real past cannot serve as the touchstone for the truth of historians’ claims. It does not seem to enter into the work of historical investigation at any point. It is at this point that the distinction between methodology and philosophy is supposed to be relevant. Discussion of how historical knowledge comes to be had is relegated to the former, and while considerations of methodology are not to be denigrated, they are nevertheless not to be confused with philosophical conclusions. Thus,

it appears to be a matter of methodology to characterize the way in which historians come to know the historical past. Such a characterization will have no place for mention of the real past, and in itself that is not objectionable. However, should one observe that the characterization in question implies that the real past is irrelevant to the practice of history, one is charged with having drawn a philosophical conclusion from a methodological ‘discussion, and a conclusion which

is mistaken into the bargain. But the particular kind of conclusion that I am trying to reach is epistemological—it comes out of an attempt to understand how historical knowledge is acquired and justified——and I do not see that it 1s affected by the considerations we have been examining. These considerations are conceptual: logically true, but factually vacuous. We are told that if the historian has done his job properly, that is, reasoned truly, the account with which he emerges is nothing less than a description of a real past event. But what can that mean in epistemological terms? If I have spoken truly,

what I have said is true, but this tells us nothing at all about the > Say, Maurice Mandelbaum in his now classic The Problem of Historical Knowledge (New York, 1938) and J. L. Gorman in his review of my book in History and Theory 16 (1977), 66-80.

HISTORY AND THE PRIMACY OF KNOWING Fey

truth or falsity of my assertion. We have the right to demand that the assertions of historians be subject to the test of truth. But that means subjecting their assertions to the sort of testing techniques which

function within the actual discipline of history. The if-he-reasoned-truly argument seems plausible because the logical distinction between an assertion and its referent seems reasonable; but that distinction fails to penetrate the epistemic peculiarity of history. The distinction as used in this context by realists seems to justify a leap from within the context of historical investigation to the real past. The leap, however, leaves us with a real event which has no determinate content. It effectwely says no more than that we had real predecessors who actually walked on the earth and did real things. It cannot say anything about the factual content of their lives;

to the extent that we can say anything about such matters, we are back within the practice of historical inquiry. ‘Thus, if we attend rigorously to the actual character of historical knowing, these logicalconceptual, but factually empty distinctions—between methodology and philosophy, assertion and referent, as we find them in NowellSmith’s paper—have no force at all. It is hard to avoid the conclusion that so far as the epistemology of history is concerned, these distinctions

must be rejected.

Ii] Given the general importance of examples in philosophy of history, as in most branches of philosophy, one example treated by NowellSmith is worth our attention, because understanding its character will enable us to consider a number of related issues from within the framework that I am trying to defend here, namely, that of the pri-

macy of knowing. The example is that of a murder witnessed by Peter, reconstructed by a detective whose reconstruction is accepted by a jury which convicts the accused. Subsequently, the accused is sentenced to a suitable punishment. ‘The problem for me, as NowellSmith sees it, has to do with the event which is the act of murder. On my view, Nowell-Smith wishes to emphasize, there is a differ-

ence between the real act of murder and the reconstructed act of murder. In addition, a reconstructed act of murder has no victim, so surely it must be a horrible miscarriage of justice if the accused is made to pay a penalty for such a crime. I should like now to show

152 HISTORY AND THE PRIMACY OF KNOWING

that this example is irrelevant to considerations of the sort before us.

Nowell-Smith knows that the act of murder witnessed by Peter and the act of murder reconstructed by the detective are the same. It is, after all, his example, and it possesses just those features he chooses to impart to it. What is important to ask about the example is whether it resembles the knowing situation in which historians typically find themselves; and the answer is: no. Nowell-Smith looks

at the situations which make up his example from the outside: he “observes” the witnessing of the act of murder, and he “observes” the detective reconstructing it. The purpose of the example is to suggest

that the detective knows—in the end—what Nowell-Smith knows, but the example cannot show this. With respect to the situations of the example, Nowell-Smith occupies a position which God is said to occupy with respect to what goes on in the cosmos. The vantage

point of God would enable one to know that some event in the actual human past had precisely the features that some subsequent historical reconstruction imputed to an event in the human past; and for God, knowledge of the human past would not have to be mediated by the discipline of history. But what features an historian felt

impelled to impute to the event he was reconstructing would be determined by an historian working only within the framework represented by the detective’s standpoint in the example, not by that represented by Nowell-Smith’s standpoint. We want to say that what the witness saw and what the detective reconstructed was one and the same event. And by parity of consideration, we want to say that what those who were with Caesar on that fateful day when he crossed the Rubicon saw® and what historians who reconstruct or constitute that event characterize is, like-

wise, one and the same event.’ We do not, of course, want to say that there were no determinate human events in ages gone by. It is exceedingly difficult to doubt that real things were being done by © It is a little more complicated than it may seem, masmuch as crossing the Rubicon

is not simply a behavioral event in which the subject begins on one side of a river and ends on the other. The crossing carries with it decisions with respect to policy which may not have been all that visible to everyone with Caesar on that day. But we shall not bother with those complications which can only raise problems for realism, not for me. ’ Is there actually evidence of the crossing itself—which would have been visible to the people with Caesar—or do historians conclude that he crossed the river because there is evidence, or reason to believe, that once he was on one side and later on the other? Could such a “crossing” as the latter ever be witnessed?

HISTORY AND THE PRIMACY OF KNOWING 153

real people. But what things? Again, we are forced to recognize that only the discipline of history enables us to find answers. What we find it exceedingly difficult to doubt proves to have no determinate content. ‘Thus, the example of the murder case is simply misleading. The accused is convicted because the evidence leads first the detective and then the jury to the conclusion that the accused committed the crime. Their conclusions are arrived at later in time than the date of the act of murder, yet the conclusion, when it is arrived at, is to the effect that the crime was committed at the earlier date. ‘The conclusion may rely heavily on the testimony of the witness, but it is not that testimony; the conclusion may seem to assert the factuality of the

same event as that reported by the witness, yet it is not the same report. The standpoint of Nowell-Smith as inventor of the example does not actually enter into the work of real attempts at solving crimes. And the standpoint of God does not enter into the work of historians

attempting to constitute the human past. It is the standpoint of the attempt to know that hes back of what I have been trying to say, rather than anything inherent about past or present as such. It is true that I write—and wrote in my book— as if it were the peculiarity of historical events being located in the past that leads to the special epistemological difficulties of presentday historians. And I certainly think that a past event is rather more difficult to deal with than a present event. But it is because they are not available for observation that past events raise problems that present occurrences do not, not because of anything that has to do with the nature of the past as such. By choosing examples for analysis which are appealing to common sense, realists like Nowell-Smith seem to be able to undercut the position that emerges when the standpoint of knowing is placed in the forefront. The reference in his paper to political figures whom we know about in non-historical ways is intended to create the impression that on my view we must have many Franklin Roosevelts and many Gerald Fords; that the Franklin Roosevelt who was president of the United States until April 1945 must be other than the Franklin Roosevelt I now remember as I am writing these words;

that the Gerald Ford whose name confronted me on the ballot in November 1976 is different from the Gerald Ford who followed some policy in 1974, so that to vote for or against the former Gerald Ford

in November 1976 on the basis of the policy of the latter Gerald Ford in 1974 really makes no sense. It is really quite absurd. But the

154 HISTORY AND THE PRIMACY OF KNOWING

point of the realist critic 1s, I suppose, that if he can make me see the absurdity of the position, I shall give it up. But the absurdity 1s not consequent upon the primacy-of-knowing position. It is, rather, the result of the way in which realists interpret such positions. What I must do in what immediately follows, is to draw out the basis of the absurdity in order to draw a sharp contrast between it and what I am trying to say. The absurdity emerges from the view that the events of human history are located in the past. It depends on taking literally the metaphor of temporal location. That the real past is inaccessible would seem to mean that there is something® that separates us now from

what was previous to the present time. But it is a long temporal span that precedes the present. If events are located within it, they are presumably not spread out so as to fill all of it. Franklin-Rooseveltbeing-elected-in-1932 occupies one span of time; Franklin-Roosevelt-

being-inaugurated-in-1933, another span of time; and so on until Franklin-Roosevelt-dying-in-1945, which occupies still another. And, of course, Franklin-Roosevelt-being-remembered-by-me-in-1977 would

be something else again. How are we to effect any sort of unity among all of these Franklin Roosevelts?? In my view, the “unity” of all these disparate Franklin Roosevelts is simply a consequence of

the fact that it is one career which emerges from the attempt of historians to deal with the relevant period of American history. In fact, there is no problem of unity. ‘The so-called problem is an artifact of dividing the past into temporal locations into which different events

are deposited. Once it is clear that the past is not to be treated as a receptacle, there are no problems of continuity—whether of persons or of long-term occurrences such as the transformation of Metz in the Middle Ages. ‘There is no reason to hold that what historians

reconstruct are successive moments in time. What is closer to the truth is that they constitute a course of events or the course of a life. The continuities are built into the historical constitution itself. I can imagine, however, that there is yet one matter, suggested by what Nowell-Smith writes, which may still seem to be unresolved. Even if what I say seems reasonable in that an historical account of 8 Likewise a metaphor for which I hope I shall not have to pay. ° On the problem of identifying individuals in the historical record, especially in the case of people who are not numbered among the great, see Identifying People in the Past, ed. E. A. Wrigley (London, 1973), particularly the editor’s “Introduction” and Ian Winchester’s “On Referring to Ordinary Historical Persons.”

HISTORY AND THE PRIMACY OF KNOWING 159

the life and career of Franklin Roosevelt would have the continuities built in, and that the so-called problem of the unity of the different time spans of that career is only an artifact of a conception of time which is not implicated by the primacy-of-knowing point of view,

that life and career came to an end in 1945. There are, however, people still living who remember Franklin Roosevelt, in fact, even knew him personally and worked with him. They did not know and work with the historical Franklin Roosevelt (as the term “historical” is being used here; obviously, there is another sense in which he was an historical person and was known to be such in his own lifetime) but with the real Franklin Roosevelt whom they still remember without the need for historical reconstruction.

In a matter of decades, this problem will cease to exist, and it does not exist with respect to Caesar or George Washington. Because it is of such limited interest, I overlooked it in a book which was intended to deal with the special character of historical knowing and with little else. But the question does raise an issue of principle and needs to be examined. ‘The past-present distinction seemed serviceable enough given what I was doing in my book; and since the examples I have preferred to discuss have been of historical events long past, there was no occasion to consider the present question. Thus, I lost an opportunity to show that what distinguishes my views from those of historical realists is the primacy-of-knowing orientation

rather than anything that may pertain to past or present as such. The methods of historical research are needed whenever the object of investigation is not present in the sense of being able to provide

an experienced sensory compulsiveness. In that respect, even the remembered Franklin Roosevelt—what is remembered about him— may require some methods of inquiry for confirmation or rejection,

although not everyone who remembers may care to bother about putting his memories to the test. Elderly citizens who remember his

campaigning for office in their home towns simply accept the veridicality of their memories. There may be little enough to those memories beyond the fact of the president’s presence—his smile, his mellifluous voice as he spoke his words. Precise time, other elements in the context of the rememberer’s life, even which of Roosevelt’s four presidential campaigns it was which brought him to the remembered place, may be quite blotted from the recollection of this or that rememberer. It is, to be sure, the same Franklin Roosevelt who is remembered as the one who did all the other things ascribed to

156 HISTORY AND THE PRIMACY OF KNOWING

him. The recollection in question could be placed into some wider context by a qualified scholar to whose attention it is called and who may think it worth doing. ‘The memory itself, whatever it may mean to the one whose memory it is, does not point to its referent without some framework of knowing. I remember reading, though I no longer

remember where, that King George [V became so used to telling how he led the charge at Waterloo that he came ultimately to believe it. What he remembered had no status as fact—save, of course, the fact of his remembering—because there is no acceptable framework of knowing into which it can be inserted. In general, one’s memories fit into the framework created by the context of knowing—which gives them a claim to be taken seriously—or they do not. The defining

characteristics of that context do not, however, include the brutely real of the compulsively given; they do not include a real Franklin Roosevelt or a real battle of Waterloo independent of such contexts. Put in another way, the referents of historical statements or statements of memory are not to be found outside of the framework of knowing. There is no way to reach outside of such frameworks; the knowing situation is not compatible with such a reaching out; both the statements which refer and the objects to which they refer are constituted within the framework of investigation. There is no epistemic way, as distinct from conceptual word-play, to reach out from within that framework to realistic objects.'° And only such devices as the factually vacuous if-he-reasoned-truly or the belief that the God’seye view of Nowell-Smith’s detective story has relevance to the knowing

situation as it actually is may lead to the expectation that such a reaching out is possible. IV

The long chapter on “Disagreement in History” (ch. 4) in my book

was intended to illustrate something about the way in which the historical past comes to be known, to tell something about the way in which the discipline of history is practiced. If I understand Nowell'0 I would not even exempt the case of the delphiniums in the garden from this, because there is more to being a delphinium or a garden than the brutely given. To be either is to realize a concept, and that immediately plunges us into a context of knowing—even if it is only common-sense knowing—in which such concepts emerge. (See C. I. Lewis, Mind and the World-Order [New York, 1929].)

HISTORY AND THE PRIMACY OF KNOWING 157

Smith’s reaction to that chapter, it would appear that he likes it as a contribution to the methodological clarification of the practice of history, but he seems to doubt that I have a nght to it given certain philosophical views that he attributes to me. In his view, the whole notion of disagreement with respect to what we have reason to believe

the human past was like is unintelligible in a philosophy which is “constructionist” and which rejects historical realism and restrictions imposed by the real past upon claims to historical knowledge. It seems to me, however, that Nowell-Smith’s conclusions rest less on the actual

character of the views I hold than on certain views he would attribute to me: specifically, a belief in incredible entities. In the present section I wish to say something about disagreement in history and to

show why my view on the subject of disagreement is in no way incompatible with my non-realistic philosophical position. Logically, the point of departure for historical research is a body

of evidence. Given the evidence, the histonan attempts to come to some conclusions as to what a span of the human past was like. How shall we understand disagreement among historians? There are a number of ways in which historians may be in disagreement. ‘They

may disagree on how to explain an historical event concerning the character of which they are not in agreement. Or they may disagree over the significance—the meaning—of an historical event concerning the character of which they are not in disagreement. I was not concerned with disagreements of these kinds. My book is concerned with history as a way of knowing, and the sort of disagreement which could underscore the nature of that could only be disagreement over

what the human past was like, given the evidence and the ways in which historians think about evidence of the given sort. Through the medium of a detailed examination of an example—namely, the state

of the scholarship on the Dead Sea Scrolls in the middle to late 1950s—some conclusions were arrived at. The example was of a period of Scrolls scholarship prior to the emergence of a scholar’s tradition with respect to their place in history, and that made it possible to emphasize certain considerations. There was still no agreement as to what other bodies of known evidence the Scrolls were to be joined with. The determination of what bodies of evidence belong together

is part of the problem; we do not first pile up the evidence and then determine by means of some inferential procedure what the past was like which produced that pile. On the realist view, there is an event in the real past which provides the touchstone against which to test

158 HISTORY AND THE PRIMACY OF KNOWING

the truth claims of different accounts, and this real event functions for historians in the way that something perceived functions for the accounts of eyewitnesses. This view was challenged in the discussion

of the chapter in question in the following way. To begin with, the three accounts of the three different scholars whose books constituted the example were so unlike one another as to make the comparison implausible. But more important, what emerged from the discussion was the way in which historical accounts are developed from within the practice of history, with no reference to anything external. From

the standpoint of realism, this must inevitably lead us to historical skepticism, but whether or not it does is precisely the point at issue."' For me, there is no difficulty in looking forward to a time when the weight of historians’ considerations will lead to the formation of an

historians’ tradition concerning the general character of the period in question (cf. HK, 90f.), and my impression is that so far as the

Dead Sea Scrolls are concerned there has already developed an overwhelmingly-supported view that they were produced by Essenes or Essene-like sectarians.

How I can presume to say the sorts of things with which the previous paragraph comes to an end is precisely what Nowell-Smith cannot understand. In his view, since I have dispensed with the real past, I ought to be left with historians subjectively constituting historical events, a subjectivism rendered inevitable precisely because I do not have the real past to determine where the truth of the matter

is. He thinks that realists may say that “constructs” which are rejected in the history of histonography are false, and he wonders what

I can say. I seem to say that the historical event is what is reconstructed; there is no external event which the historian’s reconstruction or account is said to describe. An account or description may be false, but it is surely not an event.’? Thus, I would seem to be left with all manner of entities—reconstructed historical events—with no way of getting nd of any of them.

Indeed, it seems to be suggested that I am worse off than that other “constructionist,” R. G. Collingwood, who is quoted to the '' Cf. the interestingly opposite accounts of the basis of the skepticism of the historian Carl Becker in Nowell-Smith’s What Actually Happened (Lawrence, Kansas, 1971) and my “Historical Realism: The Ground of Carl Becker’s Scepticism,” Philosophy of the Social Sciences 2 (1972), 121-131.

'2 Nowell-Smith’s argument depends on treating my notion of the constructed historical event as if it had the properties of his notion of the real past event.

HISTORY AND THE PRIMACY OF KNOWING 159

effect that the historical past must be logically consistent. Clearly, those entities which make up my historical universe do not collectively form a consistent whole; of the three reconstructions which make up my example, at most only one could be correct; and whichever one that is, the other two could not be compresent with it in a common universe—yet I have no way of getting nd of them. But why should I be saddled with views such as these? ‘I. W. Adorno says of Popper that he “seeks the essence of criticism in the fact that progressive knowledge abolishes its own logical contradictions,”'’ and

in the sense that in the course of continuous work on intellectual problems we come to increasingly preferred solutions, that certainly seems to be true of history as of any systematic discipline. For whatever reasons, some historical constructions will seem to the community

of scholars better than others that have been proposed, and these latter will simply be dropped from consideration. ‘There can be no doubt that Nowell-Smith believes this, too. The interesting question is why he does not think that this obvious solution is open to me. Of course, we have seen that one reason 1s that in his view only the accessibility of the real past provides the ground for determining which of several conflicting accounts is true.'* [ am not saying here what precisely can be meant by “accessible”; I shall not impute to realists views stranger, epistemologically speaking, than what they choose for themselves. The point is that in Nowell-Smith’s view there is an epistemically licit and non-vacuous way to distinguish between the real past and the variety of reconstructions that historians produce, and inasmuch as the primacy-of-knowing orientation rejects that possibility, it forecloses the possibility of choosing between reconstructions or of allowing that some particular historical account is rejected, and hence no longer clutters up the universe. This way of restating my views and restricting the options open to me seems to receive confirmation from the way in which NowellSmith construes two words which appear in my book, “experience”

and “knowing.” That I use them appears to him to give away my hand, so to speak. His argument supposes that there 1s a markedly limited range within which these words may be used, and that he is right about what they must mean when they appear in writing which '5 Adorno, 16. '* How this is done is never made clear. Obviously, the if-he-reasoned-truly argument cannot conjure up a real past capable of resolving the conflict.

160 HISTORY AND THE PRIMACY OF KNOWING

purports to be philosophical. I have long found it remarkable how frequently members of the philosophical school which identifies the meaning of words with their use are unwilling to permit others to use words in order to say what those others mean to say. In any event, I find nothing compelling about the restrictions Nowell-Smith

| would put upon the functions of the words in question, and what he says on the basis of his construal of them neither captures the intent of my usage nor points to a limitation on the possibilities open to the primacy-of-knowing point of view. It is strange to see a writer insist that “knowing” is a neologism, with no obvious sense, yet at the same time be so certain of what it

must mean and to what it must commit one who uses it. In any event, for realists, we are told, there is such a thing as historical thinking or historical reasoning: it is not a way of knowing but a way of coming to know. I do not see why a way of coming to know

is not a way of knowing; but at most, if at all, this may be an admission that I am less sensitive to the compulsions of the English language than are some others, not that | am committed to a philosophical thesis. For Nowell-Smith, however, one speaks of “knowing” because one comes to think of it as a form of experience—hke those activities referred to by those other verbs ending in “ing” which he cites—to which he responds: “To know something is not to have a special’? kind of experience; it is not to have an experience of any kind at all.” The discussion of “knowing” comes at the end of Nowell-Smith’s essay; that of “experience” around the middle. Yet for all that, they are really part of the same argument, and I feel no hesitation about dealing with them together. I talk about “historical knowing” as the way historians come to know about the human past, and I called my book Historical Knowing so as to emphasize that it deals with that side

of the historian’s work and not with such other questions as explanation in history, the historical narrative, the meaning of history or meaning in history, or anything else. For Nowell-Smith, there is more to my choice of the word “knowing” than meets the eye, and its use convicts me of subscribing to the view that history is a special form

5 “Special” is a key word, and there was nothing in his previous remarks that

| required Nowell-Smith to use it. But special experiences need special entities to be their objects, and in this context that role is played by the eternal constructions that non-realism is not permitted to get nd of.

HISTORY AND THE PRIMACY OF KNOWING 161

of experience.'® And it cannot be denied that the notion of the historian’s experience is to be found in my book; Nowell-Smith correctly quotes my insistence that in doing philosophy of history we must “attend to the experiences of historical constitution” (AZA, xxii; “experiences” is not emphasized in my text as it is in his quoting of

it). And this leads him to make a distinction which I do not, and which I reject, namely, between historians’ experiences and the infrastructure of history. I distinguish between the superstructure of history-—the finished literary product by means of which an _historian may present his conclusions to date—and the infrastructure of history—the methods whereby the historian thinks historically about

the evidence before him and emerges with a conception of what some part of the human past was like. ‘he infrastructure has nothing to do with the historian’s subjectivity. Rather, it concerns the methods which define the discipline of history, and I have tried to emphasize that while historical knowledge is relative to the discipline of history, in the same way that any sort of knowledge is relative to

the disciplined way in which it is produced, it is not relative to the subjectivity of historians." What we find in section V of Nowell-Smith’s paper is an attempt to force a distinction between the historian’s experience and the infrastructure of history. The passage he quotes (/7/K, 5) to the effect

that Arthur Danto “is not able to do full justice to the sorts of experiences historians have which enable them to constitute the historical past” (here, too, “experiences” is not emphasized in my text though it is in Nowell-Smith’s quoting of it), ought to be interpreted in the context of what I was trying to do, namely, to deal with what historians do when they try to constitute the human past. ‘This puts us precisely in the midst of attention to the infrastructure. Not too many pages later, I make reference to “the intellectual experiences of historical knowing” (7K, 13; emphasis added here), and in my final chapter I observe that one thing that would not be of use to us in the attempt to solve the problem of historical objectivity “is a phenomenology of the historian’s experiences qua historian in any subjective sense” (HA, 212f.). Thus, when Nowell-Smith says that my '© And not merely in the sense that one works at being an historian in ways which are different from the ways in which people work at being physicists or psychologists.

'7 Or need not be; obviously, I cannot deny that there exists sloppy, incompetent, and tendentious work in history as in any field.

162 HISTORY AND THE PRIMACY OF KNOWING

“new terminology,” that is, my use of “experience” or “historian’s experience,” suggests “something different from attending to the infrastructure of their thought,” we must wonder what leads him to draw away from what is the obvious sense of my words for the sake of a clearly strained interpretation. His interpretation has it that the experiences in question are of a more personal kind than the application of the techniques of the infrastructure; they would prove to be of greater interest to psychologists than to methodologists of history. ‘To determine what an historian’s experiences were, say those of Sir Ronald Syme when he wrote The Roman Revolution, one needs an investigation of another sort than that which would lay bare the infrastructure of his inquiry. We would have to reconstruct those experiences historically, arguing back to them from his text, in some way other than the way we use his text and its underlying scholarly apparatus, in the effort to show how it is that he came to have the view of Augustus and his age which is reported in his book. All this puts so great a strain on the passages in which the word “experiences” appear in Historical Know-

ing that we are bound to wonder about what led Nowell-Smith to retain the discussion in his paper even though the section which contains the discussion of “experience” ends by wondering whether or not the account presented is “too literalistic.”!® But he decides to stick to his position; he thinks it not “unfair to press the word ‘experience’ as closely as I have done because Goldstein’s anti-realist

position and all the problems that... constructionism as a philosophical thesis gets us into depend precisely on the formulation in terms of experiences.” Why should this be the case? I am not sufficiently familiar with Nowell-Smith’s writings to be able to offer an account of his conception of experience, but from what he says in “The Constructionist ‘Theory

of History” we can learn something about it. Whatever he takes experience to be, apparently cognitive activities cannot be experiences.

If they could, there would be no way to make the distinction we have just seen him try to make. Being what experiences seem to be, they must have direct objects. Perceptual experiences have immediate objects of experience, and so for any other sort of experience one '8 Being “too literalistic” suggests that Nowell-Smith’s use of the term is correct and mine not. I do not agree to that, but I do not wish to waste space by discussing what “experience” really means. He is not too literalistic; simply wrong.

HISTORY AND THE PRIMACY OF KNOWING 163 could imagine. If I believe that we may speak of historians’ experiences,

it must be that I believe in special objects of such experiences, and here we return to those special objects of knowing—incredible entities—

that we met above. And it will not do for me simply to deny that I believe in them. For Nowell-Smith, such a denial would presumably mean that I do not understand what my anti-realist position entails. It is, he believes, precisely because it entails the sorts of special entity he claims it does—conflicting constructions’? which can never be chosen among because there is no real past to provide the ground of choice—that he is justified in his “too literalistic” reading of my words.

But if Nowell-Smith’s refutation of my anti-realistic position is an example of the widely-used method of dismissing views as absurd because they entail the existence of strange or special entities, it remains for me to indicate how the primacy-of-knowing point of view is entirely reasonable and is not subject to this kind of criticism. I want my remarks to be general, since I believe that the primacy of knowing is a generally sound epistemological stance, though I do not want to stray too far from philosophy of history. For me, it 1s difficult to understand how that wide-ranging and complex “something” which is natural science from antiquity to our own day could have taken the manifold directions it has over the centuries, pursuing natural reality first under one fundamental set of assumptions, then under another—were it not that natural reality is mediated for us by the procedures of knowing. No one would wish to claim that during the past two and a half millennia the natural world has changed as frequently and as radically as our theories have concerning its nature. What we take nature to be like has certainly changed most radically during the course of time. It is difficult to see how this could be so if a straightforward epistemological realism could provide an adequate account of how we come to know what we know. By “straightforward” I mean a realism that did not fall back on verbal tricks. If we can assume that in the comparatively brief period of time since the pursuit of scientific knowledge began nature has not undergone radical

change, then, given what the history of science seems to be, it is clearly impossible to take seriously the claim that our knowledge of nature comes to us unmediated by the procedures of knowing. I think that such mediation is a fundamental fact of knowing. If one believes '° It would take some doing to show how constructions become the immediate objects of historians’ experiences.

164 HISTORY AND THE PRIMACY OF KNOWING

that such mediation is incompatible with genuine knowledge, that genuine knowledge is possible only on the assumption that the objects

of it are given truly, then one must, perforce, become a skeptic.” Presumably, Nowell-Smith ought to hold that to say what I have just been saying must entail that those intellectual constructions which are the views of nature of Aristotle, Newton, and, say, the Copenhagen school, are eternal entities among which there can be no grounds for choice. But it is precisely the making of such choices that punctuates the history of science. And those choices are made within the frame-

work of the enterprise of knowing itself. At no point does nature break into our consciousness—if that makes any sense—to present itself unmediated by the methodologies of knowing. Choices are made,

theories are accepted and disregarded, thinkers and theorists come to definite conclusions as to the character of natural reality, and no verbal sleight of hand about reasoning truly brings the unmediated into account. (And if one 1s willing to forgo the unmediated, I cannot

see why one should bother to define oneself as a realist.) There are two considerations which may make it seem that the realist view may have some merit as an account of our natural knowl-

edge. The first 1s that in the world of the natural, present nature seems to impose itself brutely upon our awareness of it. The other is that observation of nature plays a role in the determination of whether

or not theories are acceptable. To be sure, both of these are very complex matters. Perception of things is not simply a matter of apprehending the brutely given, yet for all that the element of the brutely

given is part of the perceptual situation, or so a great many would argue. And while there is a good deal of sophistication manifested these days about the relation of observation to the confirmation of theories and about what is to count, in the end, as an acceptable observation, that observation of what is present to sense plays a signifi-

cant role in the testing of theories is hardly a matter for dispute. Neither one of these considerations obtains in history, though the general features of the knowing situation—the development of claims to knowledge by means of systematic methodologies into which unmediated reality cannot intrude itself—obtain in history too. It might 0 See my HAistoncal Knowing, ch. 2, and “Historical Realism: The Ground of Carl Becker’s Scepticism” (in this volume pp. 106-118); for the view that Collingwood’s anti-realism in history 1s an attempt to overcome skepticism, see my “Collingwood’s Theory of Historical Knowing,” History and Theory 9 (1970), 3-36. (In this volume pp. 273-311.)

HISTORY AND THE PRIMACY OF KNOWING 165

be possible to argue, though I shall be satisfied merely to suggest it without attempting to prove it, that an examination of the way in which claims to historical knowledge are made provides a purer case on the basis of which to construct a theory of knowledge than other more usual cases. There is a touch of irony about such a suggestion _ when one realizes how often history is made to bow its head shamefacedly for its inability to satisfy epistemic criteria which philosophers

have developed on the bases of their construals of natural science. V

In the present section, I want to reconsider the question of reference in a way to emphasize the difference between Nowell-Smith’s position and mine. In Historical Knowing, the idea of reference is tied to that of verification, and for that I am accused of failing to grasp the distinction between methodological and philosophical conclusions.*' I have already indicated that I take the distinction as presented to rest on the prior acceptance of historical realism, and so to press it against

me seems only to beg the question rather than to introduce a new consideration. Tying the two together—certainly, discovering that there

is some relation between them—seems to be inevitable in a philosophical examination of how knowledge is actually acquired and justified which does not take its point of departure from conceptual distinctions

which are abstract and removed from real quests for knowledge. Reference—the way in which an assertion speaks about its object— and verification—the way in which claims concerning objects of 1n-

terest are shown to be true, or at least reasonable to believe—are certainly distinct notions. Yet acknowledging this merely conceptual dis-

tinction does not entail anything about if or how they are bound together in the actual framework of knowing. But before I attempt to deal here with the issue of historical reference, I should like to acknowledge one issue which seems to emerge from any effort to tie reference and verification together. I have already noticed it, but given the focus of my attention in that earlier section,

I did not deal with it on its own terms. It is the question whether a view which ties reference and verification can avoid the seemingly strange conclusion that no one referent can be the object of more 71 Gorman is particularly insistent on this point.

166 HISTORY AND THE PRIMACY OF KNOWING

than one method of verification. Nowell-Smith raised the question in his discussion of the witness-detective example and, of course, it came up frequently in those years during which operationism was being much discussed. When I dealt with Nowell-Smith’s presentation of the witness-detective example, I was interested in making the point

that the example was introduced in such a way as to falsify the knowing situation: that with respect to the situation characterized by the example, Nowell-Smith occupied a position which may be analo-

gous to God’s view of the world but was surely not analogous to that occupied by working historians and detectives as they seek to learn something about the objects of their interest. The actual force of the question was not really faced. The question is a general one, not limited to problems of historical knowing. Actually, if it were generally conceded that historical knowing could serve as the model for the acquisition of knowledge of any sort, as I suggested at the end of the previous section, it could

be argued that my handling of the matter as sketched in the previous paragraph is all that is needed, for there would be no way for multiple perspectives on an identical referent to be introduced in the way that Nowell-Smith introduces the perspectives of the witness and the detective. Only that of the detective or the histonan or the producer—not the observer—of things known would be possible. Since it is not generally conceded that historical knowing could serve as such a model, I shall pursue the question of one referent for several methods of verification somewhat further. And inasmuch as it Is a general question not limited to the subject of our present concerns,

I do not feel the need to work out a detailed solution. I want only to say enough to indicate the direction I think a solution would take.

This will enable me to get the question out of the way. An absolutely rigorous empiricism could well insist that far from creating a problem, different operations, different ways of knowing,

must have different objects or referents, and that there is no epistemically responsible way to avoid this conclusion. But such a position must rest on some sort of epistemic atomism which 1s by no means compelling.” It suggests that every test, every operation, every technique designed to elicit knowledge of some sort, is isolated from

2 Tt should be apparent that I do not hold the “atomic conceptions” attributed to me by Gorman, 79.

HISTORY AND THE PRIMACY OF KNOWING 167

every other such one, and that there cannot be, within the framework of the knowing situation, grounds for thinking that different techniques are directed at the same referent. Some may think that the only way to know that the two or more referents are really one is to be able to extncate oneself from the lmitations of the framework of knowing: occupy God’s perspective, or that of Nowell-Smith in the witness-detective case; or be able to argue oneself out of that framework by such means as the if-he-reasoned-truly ploy, or some

others if they can be conceived of. But to say that the referents of the various techniques are really one is to make a claim to know something; it does not result from the impact of the unmediated real

upon our awareness. Thus, there is no reason why it cannot be analyzed and understood within the epistemic framework of the relevant discipline. It would be useful if this could be illustrated by the

consideration in detail of some examples, but these would not be examples taken from history. hey would come from natural science—

say, the way we know that the various objects of measurement of the different kinds of barometer is the same air pressure. I am_ not going to present an account of such an example nor am I going to pursue the matter further here. I think I have said enough to show that the issue need not be particularly troublesome for the primacyof-knowing orientation or, indeed, for any view to the effect that there is a close tie between reference and verification. To return to the subject of historical reference, how do we know what an assertion—or an entire account—is about? How can we determine what precisely it refers to? It might be said that if one understands the assertion as an expression in some determinate natural language and can analyze the assertion into its parts, one could then discover just what part of the assertion is its subject, that about which

the assertion is made. Not infrequently, such a procedure receives the support of common sense, our ordinary understanding of the world providing a framework within which the assertion is explicated and understood. There are, of course, everyday assertions for which such an approach 1s entirely warranted. But to take such a view with respect to determining the referents of all assertions, without further regard to specific epistemic contexts, is to give primacy of place to

ordinary language and common sense. There are many who think that this is precisely what we ought to do; perhaps they would wish to insist that reference 1s a perfectly ordinary thing, that we refer all the time. I’o adopt some other approach to the analysis of reference,

168 HISTORY AND THE PRIMACY OF KNOWING

they might further argue, is to give primacy of place to something else, and that would be no less arbitrary. This, however, need not be the case; it may seem to be so only if the question about historical reference is asked abstractly, removed from any specific context. When we ask about historical reference, we are asking specifically about the way in which assertions of a determinate class refer to their referents, and it cannot be the case that the answer is to be given in terms of the analysis of ordinary sentences entirely removed from the epistemic context of that determinate class. I should think that if we want to know how any class of assertions of some particular sort refer, we have to attend to the ways in which those who formulate assertions of that sort use them and test for their adequacy. It is not obvious a priori that all classes of assertions will prove to be identical in these respects. Nor is it obvious a priori that any particular characteristic may be ascribed to all assertions qua referential. Whether it may or not would have to be determined by examination of the various classes of assertion seriatim. One might read in a text of Roman history, “On January 10-11, 49, Caesar led his troops across the Rubicon”; the 49 would be B.C., but the context would make it unnecessary to state that explicitly. What I want to do now is to discuss this assertion qua referential. It is obvious enough how a great many people would answer the question: To what does the given assertion refer? It refers, they would say, to a complex real consisting of the man Caesar, the troops who followed him, a determinate place between Italy and Cisalpine Gaul, a specific span of time, and something that the people involved— Caesar and his troops—did. It refers directly to that complex; if the assertion is true, its truth is in virtue of the character of that complex itself entirely independent of how we come to know about the complex or ascertain whether or not the assertion is true. All this seems to suggest that the reference takes place outside—if I may use a spatial metaphor without being accused of intending it literally— the framework of historical knowing, as if an assertion has the capacity to reach out, as it were, from within that framework to the object of its reference. What keeps me from acceding to this common-sense view of reference is that I simply do not understand the nature of the reaching out. It seems clear that everything that we can come to say about the historical past emerges entirely within the

framework of historical knowing. Every attempt to subject to

HISTORY AND THE PRIMACY OF KNOWING 169

verificational test the claims that historians make requires that the procedures which led to the claims in the first place be repeated. There seems to be no way to the referent of an historical assertion except by means of the procedures of historical constitution themselves. If there is anything to be said for the view that the truth or falsity of a claim to knowledge is determined by the character of its referent, and if in history there is no referent except as it 1s constituted

or constructed by means of the techniques of historical investigation—which is the point of my saying that in history the distinction between a fact and the assertion of a fact does not exist (HK, 42)— then it is hard to see how reference in history can be achieved by a reaching out from within the framework of knowing to the facts and events of an unmediated real past. Put another way, if the reference of an assertion is determined by attending to what is looked for by those who are qualified to determine its truth, the realist view of the matter seems to have little in its favor. One possible solution to the problem this poses for realism might be to force a distinction between the referent of an historical assertion and the way in which it refers. The former would presumably refer to the real event in the real past, and the latter would involve considerations of technique and methodological procedure such as | have just been talking about. In effect, what we would have is the restoration of the distinction with which the realist criticism of the primacy-of-knowing point of view begins, namely, the distinction be-

tween philosophy and methodology. But in the end, this will not work. To be sure, the distinction is intelligible; the two are conceptually

distinct. Yet when we seek to put it into the cognitive framework within which historical knowledge is actually acquired, there is nothing

we can do with it. The referent remains entirely empty of factual content: whatever can be said of it depends on the outcome of the application of methodological procedures. The whole notion of a referent in the sense required by this distinction is parasitic: on the one hand it gets its factual content from the sphere it would treat as merely methodological, and, on the other, it gets its seeming intelligibility from the common-sense notion of a real past. But all that one

is committed to by the latter is the factually vacuous belief that we had predecessors on earth and that they did the things that they did. We seem, then, to come full circle. Without playing conceptual tricks or making common-sense assumptions, there seems to be no way to break out of the framework of knowing in order to refer to

170 HISTORY AND THE PRIMACY OF KNOWING

historical facts and events which are real in the sense of being unmediated by our ways of knowing. Yet, for all that, we have, thanks to the discipline of history, a rational and intelligible conception of what it is reasonable to believe took place in the human past. Why should

one want more than that?

TOWARD A LOGIC OF HISTORICAL CONSTITUTION I

The question which it is my intention to explore in the pages which follow is, What makes an historical reconstruction acceptable? But before I begin, it seems appropriate to say something about the term ‘historical constitution,’ particularly since so far as | know I am the only one who uses it. Such perverse idiosyncracy ought not to be indulged, yet in the present case some justification can be offered. I am not wedded to the term for its own sake, but it does seem to suit my purpose admirably. What I am trying to do when I use it, is to avoid using the established alternatives, because those alternatives carry

along with them suggestions about history and historical knowing which I should want to reject. Much philosophical effort has been expended in recent decades over the problem of historical explanation, but that will not concern us here. Prior to the effort of any historian to explain some historical event or course of events, must be the attempt to establish that some-

thing-or-other did in fact take place, that very something-or-other which will subsequently have to be explained. And it is at that prior stage that the epistemological problems of how the past is known arise. But whatever those problems are, and whatever solutions to them seem best, at that stage the historian is engaged in a variety of activities the goal of which is to make known what it is reasonable to believe happened at some time in the human past. It is by that set of activities that the historian constitutes the historical past, and the end-product of what he does during that stage of his work might be called the “constituted historical past” or the “historical past as

constituted in historical research.” | But why use such terminology when there seemingly exist perfectly ordinary ways of referring to the sort of activity in question? Thus, in the very opening sentence of this paper I use the common term ‘historical reconstruction,’ and one frequently sees the term ‘historical description.’ In my view, both of these expressions carry with them connotations which may tend to distort our understanding of historical knowing. And when one attends to how much of what

172 TOWARD A LOGIC OF HISTORICAL CONSTITUTION

passes for epistemological inquiry these days begins not with the ways

quests for knowledge are actually carried on, but with what certain words, deemed relevant to whatever issue lies before the writer, mean,

it becomes particularly pertinent to avoid using words which may have such an effect. It is perhaps not always possible to dispense with them entirely—as witness my opening sentence—but one function

of the present discussion is to make explicit what they cannot mean should they from time to time appear in what I write. Let us begin with the expression ‘historical description.’ In itself it is harmless enough. We could agree that historians’ accounts of what

it seems reasonable to believe took place in the human past might be called by that expression. Unfortunately, even if some of us were willing to allow the meaning of the expression to be exhausted by its extension, there are others for whom the overtones of the ordinary

word ‘description’ would suggest what I take to be a mistaken conception of historical truth. Typically, a description is an account of something observed or witnessed. ‘The truth or falsity of a descrip-

tion in that typical sense is determined by the extent to which it accords with what has actually been seen or witnessed. If one reads or hears a description which seems for whatever reasons implausible, one may sometimes even have the opportunity of testing its truth by setting out oneself to observe the object of the description. The main point here is that in this typical sense descriptions involve perceptions, and the truth of descriptions involve the concordance of what might be perceived with the account given of it. Of course, no one thinks that historians perceive the events they purport to describe. But many have no difficulty with thinking that historical descriptions must satisfy

the same conditions of truth as do descriptions of the typical sort, that is, they must accord with what might be perceived by anyone present at the time who might be in position to witness the event. Yet this is a condition totally irrelevant to the practice of history. Historians—as distinct from witnesses present at the time—do make

judgments of truth and falsity with respect to the accounts of past events proposed by their colleagues, and if philosophy of history has as one of its tasks the clarification of the bases for such judgments, then it wants to attend to what historians—as distinct from witnesses present at the time—do when they make and justify such judgments. There are some who would wish to distinguish between the grounds for truth of descriptions of all sorts, and the particular ways in which truth is established under varying circumstances. But that distinction

TOWARD A LOGIC OF HISTORICAL CONSTITUTION 173

is utterly gratuitous, for there is no way to show that historical descriptions satisfy the conception of truth in question. If we are interested in the actual practice of the discipline of history and the way in which historians-at-work make judgments of truth and falsity,

we ought to attend to those intellectual—in no sense perceptual— activities in terms of which historians come to have some idea of what we may have reason to believe once transpired. And this makes

it clear that the accounts they produce are not descriptions in the usual sense of that word. Put somewhat differently, if the way in which an account is put together—its genesis and function within some mode of inquiry——is relevant to the determination of its epistemic

status, it obviously follows that descriptions of the typical sort are epistemically different from accounts produced by historians. ‘To call

the latter ‘historical descriptions’ without having attended to the considerations which have occupied us in the present paragraph is to risk being mistaken about the conditions of truth in the discipline

of history as well as about the actual character of history as a discipline.

The term ‘historical reconstruction’ is perhaps less of a problem than ‘historical description,’ presumably because ‘reconstruction’ has

a less obvious epistemological sense than ‘description.’ The term suggests a putting together of what once was but is no longer, and although there are or were philosophers who took their task to be rational reconstructions of one sort or another, the basic sense of the term suggests an activity which is not intellectual. I do not actually know what the origins of the term ‘historical reconstruction’ are, and some may think that it is simply metaphorical. But it does seem to make sense if it arose in something like the following way. It may have to do with a not uncommon view of the way in which historians put together—literally reconstruct—historical events out of evidence.

On this view, a classical statement of which may be found in the well-known handbook of Langlois and Seignobos,' the historian amasses his documentary evidence—the only kind of evidence discussed by Langlois and Seignobos—subjects it to critical scrutiny, and emerges

with a collection of discrete affirmative assertions each one of which embodies an historical fact. This is followed by the synthetic activity of the historian whereby he determines which facts ought to be put

together so as to produce a more or less complete account of the ' Ch. V, Langlois and C. Seignobos, Jnétroduction to the Study of History (London, 1898).

174 TOWARD A LOGIC OF HISTORICAL CONSTITUTION

event in which he is interested. ‘The historical event which is presented

in his account is constructed or reconstructed out of the facts which emerge from the first or ‘analytical’ phase of his work. But this conception of Langlois and Seignobos is woefully out of date; indeed, at the time of its first publication it was already out of date by not less than three centuries.’ If historical accounts could only be put together out of such antecedently established building-block facts as those required by the Langlois-and-Seignobos conception of historical synthesis, very little of the already-established historical knowledge at our disposal would in fact exist.* This will become more apparent as

we proceed. But for the moment, we may simply note that if the idea of an historical reconstruction does implicate their idea of historical synthesis, the term is clearly misleading. Having considered the two standard expressions for the activity which will be the focus of our attention in what follows, I could now

actually use them freely with the expectation that those who read my remarks will be careful not to impute to me the suggestions associated with those expressions with which I do not wish to be associated. There is, however, no assurance of this. All too frequently, philosophical critics have preferred to find the sense of an opponent’s

view in what the words he uses really mean, rather than in what he uses them to say. The whole character of J. L. Austin’s attack on A. J. Ayer’s phenomenalism—in the former’s Sense and Sensibilia—is

of that sort. ‘hus, it would seem best to use what I take to be an appropriate term. The word ‘constitution’ I borrow from the phenomenologists, but

whether or not they would approve of what I do with it is not for me to say. The problem of constitution for phenomenology is the problem of how an object or the intersubjective world is put together by experiencing subjects.* It is this focus upon the problem of putting something together on the basis of materials of a sort unlike ? This is made patently clear by recent work on sixteenth-century French historiography; cf. Donald R. Kelley, Foundations of Modern Histoncal Scholarship (New York and London, 1970); and George Huppert, The Idea of Perfect History (Urbana, Chicago, London, 1970).

> For a suggestive and stimulating conception of historical synthesis remarkedly unlike that of Langlois and Seignobos with a somewhat Kantian orientation, see Henri-Irénée Marrou, The Meaning of History (Balttmore and Dublin, 1966). * A handy guide to the development of Husserl’s interest and work on this matter, as well as references to his relevant writings, is Robert Sokolowski, The Development of Husserl’s Concept of Constitution (The Hague, 1964).

TOWARD A LOGIC OF HISTORICAL CONSTITUTION 175

the finished product which seems to me to be strikingly relevant here. That the world of objects is put together out of materials unlike the objects themselves is, of course, not peculiar to phenomenology. Hume taught that objects were built up out of impressions, and Russell that they are logical constructions out of sense-data. ‘Thus, whatever the

particular contributions of phenomenology to these matters, when those who work within the philosophical boundaries of phenomenological philosophy talk about the constitution of an objective world out of subjective materials, they are dealing with a problem or a range or problems known to the rest of us as well. But those problems are not relevant here. All I wish to note is that this process of world-

apprehension or world-making is called ‘constitution’ by the phenomenologists. Why does it seem to me a term worth importing into our present context? The focus of the problem of constitution 1s the process whereby the subject constitutes a world. ‘The way I have just formulated the matter may suggest that what interests those who work on the question

is how the world is constituted by the idiosyncratic subject, and | have no doubt that no small part of the phenomenological literature on the question assumes that this is the case. But it need not be. In the view of Alfred Schutz, in its most fundamental givenness the world is intersubjective,’ and it is hard to doubt that had he taken up the question of constitution he would have attempted to show how an intersubjective world is constituted by means of certain soczal-

cognitive processes. It is important to note that the problem of constitution need not be rooted in subjectivism, since the position I intend

to develop in the course of the present paper requires that a somewhat analogous distinction be made between the historian—the idiosyncratic subject who pursues whatever historical investigations he is moved to pursue—anid the discipline of history—that organized body of knowledge and techniques for the acquiring of knowledge which keeps the historian from falling into the abyss of idiosyncratic subjec-

tivism. In any event, certain things begin to emerge from all this, not least of which is the way I conceive of how to go about answering the question indicated in the very first sentence of this paper. I said there that the problem we are to deal with is what makes an historical ‘reconstruction’ acceptable. It is now clear that I shall be moving > See Alfred Schutz and Thomas Luckmann, The Structures of the Life- World (Evanston, 1973).

176 TOWARD A LOGIC OF HISTORICAL CONSTITUTION

in the direction of attending to the character of the discipline in terms of which the historical past is known. The historical past is constituted by historians in the course of historical research. It might

seem reasonable to say that whether or not the outcome of a piece of historical research is acceptable must depend on the extent to which it accords with what the part of the human past with which it deals was actually like. And that would point us again in the historical-description direction. But the difficulty with taking that direc-

tion is that it leads nowhere. There is no way to know that an historian’s account accords in that way, because we know nothing about

the human past other than what has emerged from historical research.° From this, it clearly follows that there is no way to test the adequacy of an historian’s account against a past known independently of historians’ accounts. The view of this paper is that the only possibility open to us at all for the justification of historians’ accounts and the determination of their adequacy must be rooted in attention to them as the outcome of a certain epistemic procedure. One might think, with Dilthey and Marrou, of a critique of historical reason. But

one must mean by that not the quest for immutable categories of the historical consciousness,’ but rather the attempt to understand how the actually-practiced discipline of history is carried on, with particular reference to the character of making claims to knowledge— as well as criticizing and justifying such claims—within the discipline

itself. It appears, then, that the whole of our present task may be located within a perspective which understands history as that disciplined way or set of methods whereby the historical past 1s constituted in historical research. Thus, my use of the term ‘historical constitution’ is not a linguistic idiosyncracy which ought to be

replaced by the more usual terms discussed above. Rather, it 1s explicitly chosen because using it indicates something about my © To avoid unnecessary digressions, I am willing for the moment to say that we know nothing of that part of the human past which lies beyond memory other than what has emerged from historical research. If it were pertinent to raise the matter here, I would point out, however, that our memories have no presumption of being veridical, and that historians of the most recent past, even if they were participants in the events they deal with, may not simply transcribe their recollections into their texts, but must treat them critically as if they were documents. I have some more extended observations on this subject early in the chapter called “The Narrativist Thesis’ of my book Histoncal knowing (Austin, Texas, 1976). ’ T have tried to say why in the chapter entitled ‘Historical Objectivity’ in the book referred to in the previous note.

TOWARD A LOGIC OF HISTORICAL CONSTITUTION 177

conception of the character of the objects of historical knowing as constituted in research, not as objectively there (gegenstdndlich) waiting

to be discovered.

I The task before us is to characterize that intellectual process out of which the historical past is constituted. But why do that in a paper which purports to be philosophical? Would it not be more appropriate

to do that in a handbook of historical method, a work which takes as its task the instruction of the neophyte on how to be an historian?

But the fact of the matter is that the handbooks never raise the question, much less point to an answer. What one typically finds in

books of that sort are instructions on how to deal with historical evidence so as to derive therefrom the maximum amount of information. They tell us about the varieties of evidence,® though they tend to be preoccupied with written evidence, and offer recipes for deriving historical facts from it. Beyond that, they tend to be unsat-

isfactory and cast precious little light on how historians produce accounts of the human past on the basis of the evidence as it emerges

from the applications of their critical techniques. According to the classic handbook of Langlois and Seignobos, the outcome of the application of critical analytical techniques to evidence is that the historian is left with a collection of facts—one fact embodied in each of the simple declarative sentences of his documentary evidence—

and from this collection he chooses the facts he needs in order to produce the account he wishes to produce. This, as I shall try to indicate shortly, is entirely false to the procedures of historical constitution it purports to describe. Marrou has a strong feeling that what Langlois and Seignobos say about the intellectual activity we are concerned about here cannot be correct, and is most unambiguous in his belief that what is needed is an account of the historian’s mental

or intellectual processes in order that we may understand how he actually moves from his evidence to the events he purports to present

to his readers, but he does not seem to be able to say precisely and in detail what such an account would be like. If we are to have rules ® Allen Johnson, The Historian and Historical Evidence (New York, 1926), contains a

useful and informative survey of such things.

178 TOWARD A LOGIC OF HISTORICAL CONSTITUTION

for becoming an historian, I suppose they ought to be produced by

people who are themselves historians. But, so far as I have ever managed to ascertain, no such rules have been formulated. The handbooks tell the beginner how he is to handle his evidence, but they tell him exceedingly little about how he is to derive the historical past from it.

But if we have no account of how to derive the historical past from historical evidence, we do have all manner of writing in which that is done. Thus, we do have a body of claims to knowledge which may provide occasion for philosophical scrutiny. Our concern, then, is not to produce a how-to-do-it account to serve in place of those the handbooks fail to provide, but to attend to the character of the intellectual procedures implicit in the production of the kinds of claim to knowledge which are the outcome of historical investigations. If handbooks actually contained explicit instructions for deriving the historical past from historical evidence, the answer to the question this paper attempts to deal with would be comparatively easy to arrive

at. One would point to the way in which some historical account was derived in accordance with the rules and could discern with relative ease why it was—or was not—an acceptable ‘reconstruction.’

But instead of having so easy a path, it will be necessary for us to look ourselves for the direction in which the answers lie. For this, a look at some simple examples of historical constitution may be of some help. Let me begin with the simplest example I can think of. In Collingwood’s ‘Roman Britain” we are told that in fifth-century Silchester

“a tombstone was found... written in the Irish, as distinct from the British, form of Celtic.” From this bit of evidence, Collingwood derives—which is not to say deduces logically—the following historical

fact:'° “An Irishman who died in Silchester and left friends able to make him an epitaph in his own language must have been a member

’ Collingwood’s contribution to R. G. Collingwood and J. N. L. Myers, Roman Britain and the English Settlements, 2nd ed. (Oxford, 1937).

'° In calling it an “historical fact,” I mean only that it is the outcome of historical research or an act of historical constitution. This does not commit me to belief in its truth or its being reasonable to believe in any other way. (Oddly, this usage is to be found in Langlois and Seignobos, p. 205, where they speak of “discordance between facts” established according to different methods.) This matter is discussed in greater detail in the chapter called ‘Historical Facts’ in the book referred to in note 6.

TOWARD A LOGIC OF HISTORICAL CONSTITUTION 179

of an Irish colony in the town.”'! Collingwood’s conclusion seems entirely reasonable, but that does not make it correct. Controversies in history are frequently conflicts over what it is reasonable to believe in view of the evidence and the ways in which historians think about

evidence. Such conflicts are apt to be the opposition of what seems entirely reasonable to what seems entirely reasonable, thus not everything which seems reasonable will prove to be historically true. ‘lo speak of something as seeming reasonable may seem to carry with it the suggestion of subjectivism: it seems reasonable to me, but it might

not to you. But the whole point of our inquiry is to find an answer to the question of why an historical account seems reasonable to believe, thus what we want to have are reasons which move us beyond the merely subjective to the sorts of consideration which would be cited in rational defense of a claim to knowledge. With respect to

the example from Collingwood, it is not clear to me why it seems reasonable, but we may note from the start that two things which might make the assertion of an historical fact seem reasonable may not be said of it. One of them I have already noted in passing: while Collingwood may be presumed to derive his historical fact from the evidence, the derivation is not an instance of logical deduction. | shall elaborate somewhat on this point in the immediate sequel. But first, I would like to say what the second thing is, namely it cannot be claimed that Collingwood read the fact out of the evidence. Our evidence contains no sentences informing us of the existence of such a colony. Collingwood asserts that it existed only because the existence of the tombstone becomes intelligible if there were such a colony. Thus, reflecting on his evidence leads the histonan to propound the existence of something for which he has no direct documentary evi-

dence. It is here that we find at work what Collingwood himself thinks of as the historical imagination, and Marrou and others as historical reason.

I want now to pick up the point about logical inference. It may seem obvious to some that with the critical establishment of the evidence, the historical facts ought to be derived logically from it, and one even finds the relation of historical facts to historical evidence

formulated in that way.'* It is not necessary for us to introduce a '' Collingwood and Myers, p. 316. '? Thus, in discussing some views of the origins of certain Jewish festival lights, T. H. Gaster (Commentary, December 1952, p. 536) says: “Whether any of these

180 TOWARD A LOGIC OF HISTORICAL CONSTITUTION

formal symbolism in order for it to be seen clearly that Collingwood’s

statement about the Irish colony does not follow deductively from the one about the tombstone. But some may wish to say that it must presuppose some statement or other which taken together with that about the tombstone would in fact make it possible to derive deductively the one about the colony. I do not wish to take this matter up in detail, but I do wish to indicate something about why it is a most implausible view. ‘To begin with, it is a point of view altogether wooden

in its application: any conflict between historians as to what the evidence makes it reasonable to believe would immediately be con-

strued to be an implicit conflict over what is presupposed. Those fields of historical inquiry which are much worked on and in which unsettled questions are reflected in much controversy would find themselves haunted by all manner of ghostly assumptions. In addition, if we are interested in knowing something about the character of the thinking which results in the constitution of historical events, we are

not helped by being told that any constitution rests on the sort of assumptions that would, in logical fact, make it possible to derive the conclusion deductively from the evidence. Since Collingwood has not

derived his statement about the colony from the evidence deductively, we want to know why it seems reasonable. To say that a deduction 1s possible in spite of what the historian did, is merely to insist upon some point of principle and, at the same time, to display a lack of interest in what the character of historical constitution is actually like.

I have, in effect, said two things about the relations of the past fact constituted by Collingwood and the evidence upon which the constitution is based, namely, that he did not derwe the former from the latter deductively and that the relation between the two—however he actually derived the one from the other—zs not deductive. There is, however, more that may be said about this issue, and while I do not think that this is the place to deal with it further in detail, I do wish to indicate something about what sorts of consideration are involved. Basically, what is involved are the two inter-twined themes of the evidence relevant to some historical problem and the character of disagreement with respect to historical constitution. With respect to the first of these, while Collingwood’s account makes it theories is nght cannot now be known; all of them are based on deductions from fragmentary and inconclusive evidence.”

TOWARD A LOGIC OF HISTORICAL CONSTITUTION 181

clear that his belief in the colony depends on the existence of the tombstone and that without the tombstone he would not have mentioned—perhaps even thought about—the possibility of such a colony,

there is really more to it than that. Colhmgwood knows not only about the tombstone but about a host of other evidence which go into the constitution of the history of fifth-century Britain. Thus, it might be argued that Collingwood’s statement about the colony does indeed have a deductive relation to the evidence, but the evidence is to be construed rather more broadly than simply the tombstone. But

even if there were merit to this suggestion, it must again be noted that the thrust of such considerations would be in the direction of a post-constitution justificationary reconstruction of the historian’s activity

along the lines of certain preferred logical principles. This would be of no help at all if we are interested in the actual character of historical constitution as the intellectual activity in terms of which the historical past becomes known. But the idea of relevant evidence neatly held together in virtue of being the evidence relevant to the constitution of the same bit of the historical past is altogether too abstract and very far from the actuality

of historical practice. If one attends in detail to the character of controversy in some field of history in which the basic issues still concern the way in which the past is to be constituted, one discovers quite early on that one central feature of the controversy is disagreement over what the relevant collection of evidence is which is germane

to the problem. Part of the point at issue is to determine which documents, for example, belong with what other documents, artifacts, ruins, and the like, as conjointly being the evidence for the same narrowly-construed swath of the historical past. In the first decade

of research instituted by the discovery of the Dead Sea Scrolls, when scholars were far from agreement as to the general character of the

texts and the community which produced them, and when arguments were being put forward linking the scrolls to such diverse groups

as the Essenes, the Holy Congregation of the very early Rabbinic tradition, and the Sicari of the Jewish revolt against Rome nineteen centuries ago, one finds that different scholars associate the scrolls with different bodies of surviving ancient evidence, and where they may sometimes refer to the same evidence, they will use it differently in different ways for different purposes. Nor have we to do with an arbitrary, subjectively-tinged kind of activity. It is not the case that

each historian piles up his evidence neatly and in order and then

182 TOWARD A LOGIC OF HISTORICAL CONSTITUTION

proceeds to the task of constitution. Were that the case, the logicaldeduction view might seem to be plausible, differences among historians being understood as owing to the fact that the different piles of evidence have different logical consequences. The decision to order the evidence in one way and to re-order it in another is part of what happens in the course of the historian’s attempt at constituting that

part of the past on which he works. As he proceeds, he comes to think that evidence which antecedently seemed to have no relation to other evidence is, indeed, directly pertinent to it, and things he thought belonged together might not. It depends on how satisfactorily it seems to him the results of what he believed to begin with are turning out to be. As he moves along certain lines, it occurs to him that what he is doing may be furthered by what he once discovered in old documents, or what he has been learning about the past illuminates documents he had no antecedent reason to think had any

relation to the matters of his inquiry. There is no point here to pursuing this further. ‘he point to be made is that the logicaldeduction point of view seems no more to make sense of the historian’s activity of historical constitution in the typical complex cases than it

did in the simple case of the tombstone and the Insh colony. III

Having thus dismissed two views which may have seemed antecedently plausible, namely, that historical facts are read out of the evidence once the latter has been treated critically, and that the histori-

cal past is constituted deductively by logical inference from the evidence, what have we left? It is exceedingly difficult to answer this without further attention to what the actuality of historical constitution

is. In addition, such attention may well be required for those for whom the initial plausibility of the views dismissed in my previous discussion has not been entirely removed. ‘There is, after all, a certain abstractness to that discussion: it is rather more argumentative than illustrative and, consequently, cannot show that historical constitution

is like this and not like that. Presumably, the way to correct this would be to become immersed in the writings of historians, but it must be noted at once that not all kinds of such writing would be of service to us. A good deal of such writing does nothing more than present historians’ accounts of what various parts of the human past

TOWARD A LOGIC OF HISTORICAL CONSTITUTION 183

were like. My “nothing more” is not intended as denigration. It is only intended to indicate that there is something else, and it is that something else which concerns us here. ‘Those accounts which most of us who are not historians usually read when we read history are the historians’ reports to date of what they take the human past to have been like. They are the finished products of historical constitution, but they do not typically present us with the way in which that constitution was carried out. In a well-worked-on field, the historian typically begins with what the scholarly tradition has established, and on the basis of new thinking and new evidence produces such modifi-

cation as he takes to be required and supportable. He presents, not infrequently in narrative form, his account of what the past was like,

and for the most part that account is not interrupted in order to show how historical thinking about historical evidence led him to the conclusions he presents. He will, of course, make reference to evidence, but this will typically be informative only to his colleagues who know

how the evidence is preserved and published. For the rest of us, it will provide only the assurance that the work is one of responsible scholarship, not an arbitrary figment of the historian’s fantasies. Thus, in order to get some idea of the nature of historical constitution, it is usually useful to examine instances of httle-worked-on periods. In dealing with these, the historian is rarely far from his evidence—in some instances the non-historian reader gets a stronger dose of evidence than he can possibly want—and it is there that we might hope to see just how the conceptions of the past which are written up in history books are actually formed. In what immediately follows, I shall present a sketch of an example of this. Even if it should prove difficult to extract from the sketch an answer to the question with which this paper was begun, it should make it clear that the cognitive activity which is historical reconstruction is rather more complex than the views dismissed above would have it. The point of departure for the inquiry we are about to sketch was a stone inscribed in runes found encased in the roots of a tree in Kensington, Minnesota, late in the nineteenth century.'? The first response of scholars was to reject the inscription as a forgery. What would such a medieval text be doing in the heart of North America, '° The discussion which follows is based upon Hjalmar R. Holand, Norse Discoveres and Explorations in America, 982-1362, (New York, 1969, Dover Publications ed.), part IT.

184 TOWARD A LOGIC OF HISTORICAL CONSTITUTION

and was it not an unusual coincidence that a Scandinavian text is purported to be discovered precisely in that part of the United States

in which there was a heavy settlement of immigrants from Scandinavia? One may note, then, that the very first question which seems to arise here is whether or not what purports to be historical evidence for something medieval is, in fact, historical evidence for something medieval. ‘The historian whose work we are considering, H. R. Holand, deals with the problems raised in ways which yield what seem to him to be satisfactory and acceptable results: Given

the kind of tree the roots of which encased the stone, it could not have begun to grow before a certain time, which time was before the arrival of the immigrants from Scandinavia. Those who were bothered about the character of the language in which the inscription was made tended to compare it with the sagas of the eleventh century, whereas the inscription itself contains a fourteenth century date, and

anyway, a modern forger who was intent on fooling the scholarly world—tfor whatever hard-to-fathom motive—would be equipped with

the latest grammars and dictionaries and would produce a text the scholars would have no difficulty about accepting. Thus, the very defects of the text argue in favor of its genuineness. But we are still left with the problem of how to account for a runic inscription in

the heart of North America almost a century and a half before Columbus is supposed to have discovered the place. Of course, one does not simply account for an inscription. One has to account for an inscription having certain features, and among

those are the things it asserts. This one talks about an expedition seemingly of both Swedes (Goths) and Norwegians and a bloody massacre, presumably at the hands of Indians. It mentions ships and it tells of things being determinate distances from other things, which distances are reported in a unit—“days-journey’—the length of which is not familiar. It may be noted that one of the problems that scholars found with accepting this text as authentic is precisely that they could find no instances of Swedes and Norwegians participating together in a joint expedition. Holand thinks that this last is once again owing to their tendency to compare the text with those which have come down to us from the eleventh century, but if one takes seriously the claim the text itself makes to be from the fourteenth century, there is, in fact, one known instance of an expedition in which represen-

tatives of both of these Scandinavian peoples took part. And that was an expedition under the leadership of Paul Knutson which was

TOWARD A LOGIC OF HISTORICAL CONSTITUTION 185

supposed to have set forth in 1355 at the behest of King Magnus Eriksson for the purpose of rendering aid to settlements in Iceland which were then beset by Eskimos, and which were, in consequence,

breaking up with the settlers lapsing into paganism. Given that the Kensington Stone—which is what the runic text we are dealing with is called—is dated 1362, and taking into account the character of the sailing vessels in use at that time and the distances which would

need to be traversed, to find remnants of an expedition from Scandinavia arriving in Minnesota seven years later is not the least bit implausible. There are other things stated in the inscription that Holand must look into. There is, for example, reference to a camp at a lake'* having two skerries located one day’s journey from the stone on which the text is inscribed. Holand scours the lakes within what he thinks it reasonable to believe a day’s journey is, does, indeed,

find one with two skerries and discovers within it or at its shore evidence that might be the remains of an early encampment of the sort involved. Step by step, Holand constitutes an expedition that began in Northern Europe in 1355, passed along one or another plausible route, and arrived in Minnesota in 1362. ‘The reasonableness of the account is further supported, or so Holand argues, by the discovery in the region of tools and weapons which, it is claimed,

rather more resemble the sort produced in Scandinavia than that manufactured by the Indians of the Minnesota area. ‘There are other matters discussed in Holand’s account,’ but I think I have presented enough to go on to some general observations. When Holand’s work is completed, we know something about an expedition we never knew before, about a complex historical event

never before described, never recorded in our historical sources. Holand’s account is not made up out of statements found in the evidence. We did, apparently, know of an expedition that was thought

to have left Scandinavia for Iceland in 1355 even before Holand produced his book, but we know precious little about it and about what befell it in the course of its travels, if it ever actually started out. I am in no position to say that Holand’s account is the last

'* Actually, the word for lake is not contained in the text, but given what skerries are and that we are dealing with an area in the midst of the continent and not on its coast, it is obvious that a critical analysis of the text requires that it be read so as to refer to a lake. '° Including the claim that the Mandan Indians have Scandinavian features.

186 TOWARD A LOGIC OF HISTORICAL CONSTITUTION

word on the subject or even that it is the most acceptable account, given what is known. But I do want to say that it is an attempt to show why it is reasonable to believe something about the human past which no one, presumably, thought about prior to this effort at historical constitution.'° We have before us the outcome of Holand’s constituting activity,’’ but is there anything we can say about what it consists in? And is it possible to say why it might seem like a reasonable, even acceptable piece of work? Before we try to consider what might be said about this, I should

like to make what might be considered to be a relevant digression.

It is relevant in that it does deal with a matter of the historian’s procedure, and, in fact, clarifies by reference to the specific example of Holand’s work a point discussed somewhat more abstractly above.

Yet for all that, it is somewhat digressive in that it does not deal immediately with the questions propounded at the end of the immediately preceding paragraph. In any event, I did argue above against the idea that prior to the historian’s activity of constituting the historical past he had first to put together the evidence relevant to his

problem and then proceed with the job in hand. I said there, that part of what emerges from the course of historical work is a view of what the evidence is that belongs together as relevant to the solution of some particular historical problem. There is no way to tell from the evidence itself that it belongs naturally with some other evidence and must obviously be used together with it for the constitution of ‘© Should it turn out that other scholars were working in the way that Holand was, and that there was already discussion of the possibility that remnants of the Paul Knutson expedition had ended up in the region of the Kensington Stone, it would not be difficult to amend my statement so as to take account of it. The personal contribution of Hjalmar R. Holand is of no particular relevance to the matter under discussion. '’ The usefulness of Holand’s book as an example of the sort of intellectual work we are interested in in no way depends on its being acceptable to the community of scholars in the long run. That the outcome of his constituting activity is acceptable obviously depends, among other things, on the authenticity of the Kensington inscription, and it appears that the overwhelming majority of philological and runological opinion rejects it. If this opinion is sound, then it must prove to be the case that the inscription is a modern hoax. An interesting account of the evidence for that may be found in Theodore C. Blegen, The Kensington Rune Stone (St. Paul, Minn., 1968), a contribution, as it were, to the constitution of an event rather different

from what we find in Holand’s book. And one which would have served our purpose as readily as his, though one which would have been rather more difficult to present in brief compass. But on the assumption that the inscription 1s what it purports to be, Holand’s use of it and other data provides a perfectly serviceable example of

the sort of constituting endeavor which concerns us in this paper.

TOWARD A LOGIC OF HISTORICAL CONSTITUTION 187

the same historical event or course of events.'® Allowing that the Kensington Stone is an absolutely essential piece of evidence for the inquiry which Holand undertook, on the ground that without it would not have occurred to anyone to think of the possibility that Norwegians

and Swedes had reached Minnesota in the fourteenth century, can we say that it has some obvious connection with the evidence in terms of which historians constitute the beginning of the Paul Knutson

expedition, and that the stone and the other evidence clearly—and antecedently—belong together as evidence for the same course of events? This hardly seems to be a reasonable view. To begin with, at the very moment of the discovery of the Kensington inscription what was most reasonable was to suspect its authenticity—somewhat analogous to the way in which it was at least not unreasonable to be dubious about the results of the Michelson-Morley experiments when

they were first reported—and this would make no sense if its connection with the Paul Knutson evidence was obvious and natural. It is only when it becomes clear to Holand that the arguments against the inscription’s authenticity cannot be sustained, and that the question

of how a stone with a runic inscription became encased in the roots of a tree in Minnesota becomes a serious one, that the possibility of finding a connection between the two sets'” of evidence becomes something to consider. There is, of course, nothing necessary about even this. An inscription dated 1362 and discovered in 1898, has more than five centuries to be moved about, and with sufficient freedom of imagination I suppose one could come up with all manner of hypotheses as to how it got to where it was found. Still, once the historian is satisfied as to its authenticity, it is not unreasonable to '8 For a somewhat opposing view, see Maurice Mandelbaum, The Problem of Historical Knowledge (New York, 1938), p. 252. In its context Mandelbaum’s account seems reasonable, but that is only because he has chosen to discuss an example from very recent history where the classification and ordering of evidence is much less of a problem than in other periods. ‘9 To talk of two sets, even though one consists only of the Kensington Stone, is

not unreasonable. Even though Holand begins with the stone, by the time he is finished he makes reference to other objects found in Minnesota which he takes to be Scandinavian in origin. It could easily have happened that some scholar took note of the existence of a collection of artifacts found in America which seemed more to resemble objects undoubtedly of medieval Scandinavian provenance than anything else known from North America. In that event, when Holand got started he would have been able to begin, not with one seemingly-out-of-place inscription, but a whole body of seemingly-out-of-place artifacts. His problem, in the end, would be the same, namely, to constitute a swath of the historical past that made sense of the seemingly-out-of-place.

188 TOWARD A LOGIC OF HISTORICAL CONSTITUTION

think that it may well have been inscribed in the area in which it was found, particularly if that is an area of relative wilderness rather than one over which much civilization has passed. Having reached that stage—accompanied by the obvious recognition that if Norwegians and Swedes were so far from home in the middle of the fourteenth century there must surely have been some not inconsequential expe-

dition which resulted in their being there—the historian discovers that indeed there was a recorded expedition of the sort in question, the very first in recorded history in which both Swedes and Norwegians

collaborated. It is only at this point that it makes sense to entertain the possibility that the two kinds of evidence may be brought to bear on the same subject. IV

The pieces fall into place. A motley of odds and ends are made to fit together so that a reasonably coherent picture emerges. What starts out as implausible, namely, that there could be authentic, fourteenth-

century Scandinavian remains in Minnesota that were not brought there at a much later date, turns out to be not all that implausible after all. In one sense, what we are looking for here is an explication of the concept of historical plausibility, and whatever it may turn out

to mean, I suppose it is clear that it is not a logical idea in the formal sense. An acceptable historical constitution must present us with an account which is plausible, and that seems to mean one that it is reasonable to believe given what we know. As long as there was no reason to believe that Europeans had reached the Americas before the voyage of Columbus, the existence of the Kensington Stone was certainly perplexing. Even when it became evident that Scandinavian explorers and settlers had appeared rather early on the northeastern shores of North America, Minnesota is surely a long way from there. In any event, we have some notions which seem relevant to the matter which has been our concern throughout the present paper,

yet if one looks at them closely, one cannot really claim to have characterized historical constitution. One may say that in order to be acceptable, an historical constitution has to make something historically plausible. And one may say that historical constitution is the result of thinking historically about historical evidence. But we seem not to know how to explicate historical plausibility except in terms

TOWARD A LOGIC OF HISTORICAL CONSTITUTION 189

of the idea of seeming reasonable to believe. And we do not know what to say about historical thinking about historical evidence other than that it is not deductive thinking. One might suppose that if the relation between historical evidence and historical events as constituted in historical research is not deductive, it must be inductive. It is difficult to speak with certainty about a subject on which the specialists themselves, the logicians, do not agree and concerning which there seems to be so much argument.

Yet my strongest suspicion about the matter is that the relation between the two cannot be inductive. The logicians tell us that in inductive reasoning we widen the scope of our knowledge inasmuch as inductive conclusions take us beyond what is asserted—explicitly or implicitly—in the premises of an argument. Inasmuch as it is plainly the case that historical accounts say considerably more than what 1s contained in the evidence which is the historian’s point of departure, it might seem that historical thinking is precisely a form of inductive

reasoning. And when one attends to the variety of different things which are discussed under the rubric ‘induction,’ it might not seem all that troublesome to add historical thinking to the list. But there are at least two reasons for not doing this. Whatever induction is, logicians speak of it as a kind of inference: Barker’! takes it to be a mode of ‘nondemonstrative’ argument or ‘nondemonstrative inference’ and Salmon, in the work just referred to, tries to show im the case of every one of the different sorts of induction he deals with how it looks as a form of inference having certain logical characteristics.

In addition, the kinds of matter which is typically discussed when logicians are talking about induction, the kinds of example introduced to make clear what they have in mind, indicate that inductive infer-

ence involves classes of things having properties of sorts, making predictions about the characteristics of populations on the basis of partial samples, and such lke. Neither one of these may be said to characterize the sort of thinking which goes into historical constitution.”” It is exceedingly hard to discern a line of inference in either *° A clear and concise survey of the area may be found in Wesley C. Salmon, Logic, 2nd ed. (Englewood Cliffs, 1973), ch. 3. “1S. F. Barker, Induction and Hypothesis (Ithaca, 1957), pp. 3 and 10. “2 C. Behan McCullagh (History and Theory 12 (1973), p. 453) says that “the nor-

mal form of historical inference” is the statistical syllogism, but I find this claim completely mystifying and have no idea as to how it could be made out. I suppose some clue to what he has in mind is to be found in his earlier discussion, p. 441;

190 TOWARD A LOGIC OF HISTORICAL CONSTITUTION

of the examples that we have attended to in this paper. What is the

inferential path that takes Collingwood from the tombstone in Silchester to his conclusion with respect to the Irish colony? And how can one ever find the direction of inference that takes Holand from the Kensington stone, and the other evidence which the course of his investigation proves to be relevant to the work he is doing, to the final result which is the constitution of an historical course which begins with the Paul Knutson expedition and ends with the massa-

cre of part of its remnant and the inscribing of the runic text in Minnesota? Nor is it possible to say that historical thinking is like arguing from the character of a sample to that of a whole. It makes no sense to say that back of Collingwood’s conclusion le correlations of the sort discussed by logicians when they deal with induction. Historical-evidence-in-hand is in no way like a sample of a larger population and is not treated as if it were. It one wants to identify historical constitution as a specimen of a determinate kind of logical thinking, we could say, following Peirce, that it is a kind of abduction, but in the end this will not prove to be clarifying. There are a number of discussions of abduction in Peirce’s Collected Papers, and I think that the following characteriza-

tions of it are typical of what he says. In one place he tells us that abduction “consists In examining a mass of facts and in allowing these facts to suggest a theory. In this way we gain new ideas; but there is no force in the reasoning” (8.209). Elsewhere, he says “Deduction proves that something must be; Induction shows that something actually 1s operative; Abduction merely suggests that something

may be? (5.171; italics in original). It is abduction, not induction,” but that discussion seems to involve statistical regularities between classes of historical evidence and kinds of historical events the like of which has never figured in the

attempts of historians to constitute the human past. “8 About induction, Peirce says such things as the following: “Induction consists in starting from a theory, deducing from it predictions of phenomena, and observing those phenomena in order to see how nearly they agree with the theory” (5.170; italics in original). Elsewhere, he says, “The third way is induction, or experimental research. Its procedure is this. Abduction having suggested a theory, we employ deduction to deduce from that ideal theory a promiscuous variety of consequences to the effect that if we perform certain acts, we shall find ourselves confronted with certain experiences” (8.209; italics in original). On previous occasions I have argued that the logical role of an “historical reconstruction” is to explain the historical evidence in the sense that the “reconstruction” 1s an hypothesis from which the evidence ought to be deducible (cf. ‘A Note on the Status of Historical Reconstructions,’ Journal of Philosophy 55 (1958), 473-79; (in this volume pp. 95-101); and ‘Evidence and Events in History,’ Philosophy of Science 29 (1962), 175-94, (in this volume

pp. 3-27) parts I and IJ). I still subscribe to this view. Thus it would seem that I

TOWARD A LOGIC OF HISTORICAL CONSTITUTION 19]

which in Peirce’s view is the source of new ideas which fertilize sci-

ence and give it new direction, and he goes so far as to say that “every single item of scientific theory which stands established today has been due to Abduction” (5.172). Finally, Peirce claims that “Any-

thing which gives a rule to Abduction... puts a limit upon admissible hypotheses” (5.196), and this seems to suggest that we cannot expect to be able to offer a formal characterization of what abduction is inasmuch as such a characterization would presume to put some limit on the application of intelligence and imagination to the data which set the problem to be dealt with.” Even our meager sample of two instances of historical constitution makes it clear that to think of historical thinking as a sort of abduction does not seem unreasonable. Here, too, what emerges from the course of the historian’s investigation results from thinking historically

about the data before him in order to emerge with a conception of the human past which is richer by far than that data. What thinking about those data suggests is only what may have been, surely not what must have been, there being, as he says “no force in the reasoning.” If historical thinking is, indeed, a form of abduction, then here, too, one has to say about our knowledge of the historical past what Peirce says about scientific thinking, namely, that all that we have of it is owing to this way of thinking. And then we are forced to face the last of the passages quoted, that abduction cannot be subscribe to a view with respect to the relationship between evidence and events in history which Peirce would surely call inductive. The constituted event functions as

an hypothesis, and from it one must not only be able to deduce the evidence in hand which suggested it, but derive some conception—I hesitate to use “deduce” here—-of what additional supportive evidence, or kinds of supportive evidence might be hoped for. But this relation between evidence and event can only be discovered after the event has been constituted. Our problem here concerns the nature of that constitution. And that is precisely what Peirce himself says in the second of the two passages just quoted when he indicates that a theory is first suggested by abduction and then the inductive task of confirmation is undertaken. ** For the sake of those who would want to know, I may point out that volume VII of Peirce’s Collected Papers contains a chapter called “The Logic of Drawing History

from Ancient Documents’ and a section of that chapter (7.136-143) is concerned with abduction. In spite of having a title which suggests its relevance to what I am attempting here, I did not find Peirce’s discussion of particular use. There are two reasons for this. His conception of drawing history from ancient documents deals with ground for believing or rejecting ancient testimonies, and it betrays no awareness of the problem of how the historian constitutes a past he does not copy out of surviving texts. And he thinks that the only way to avoid idiosyncratic subjectivism in determining what it is reasonable to believe is to be able to apply the calculus of probability, though his own practice (cf. 7.235-236) is not always consistent with this belief.

192 TOWARD A LOGIC OF HISTORICAL CONSTITUTION

limited by a rule. ‘This means that being told that historical constitution

is a form of abduction is not to be told anything about the actual character of this mode of thinking—if it is only one mode of thinking. We are, in effect, left with the question with which this paper was begun, What makes an historical reconstruction acceptable? rather like Peirce’s own “What is good abduction?” (5.197). Not infrequently

when seeking to deal with abduction, which he sometimes called “retroduction,” Peirce would refer to or even discuss Kepler’s work on the orbit of Mars, which work he highly esteemed (cf. 1.71ff.), but calling it abduction or the “greatest piece of Retroductive reasoning ever performed” (1.74) doesn’t actually leave us with an idea of its nature as being told Euclid’s Geometry is deductive in character or that Spinoza’s Ethics attempts to use deductive techniques to establish metaphysical conclusions does about those works. In order to know just how Kepler reached his conclusions, N. R. Hanson was required to devote a large part of a substantial work to examining it in detail.* But what he emerges with is certainly not the logical form of abductive inference to which any other instance of abductive reasoning—including historical constitution if in fact it is abductive reasoning—might be expected to conform.

Vv

The outcome of an instance of historical constitution is a claim to knowledge which the claimant—and typically others as well—takes to be reasonable, and we have still to determine what being reasonable can mean here. I think that it must be clear by now that the reason that this is a perplexing question is owing to the increasingly apparent fact that between the claim to knowledge—the historian’s constituted event——and the data, the historical evidence which leads to his constituting activity, there 1s what might be called an epistemic

gap. I think it may be said in general that whenever abduction or abduction-like activity is called for it is because there is no direct way to bridge a starting point and a conclusion. If the relation between historical evidence and historical events were a deductive one,

clearly there would be no gap. Nor would there be a gap—or, at any event, so large a gap—if any sense could be given to the idea *° Norwood Russell Hanson, Patterns of Discovery (Cambridge, Eng., 1958).

TOWARD A LOGIC OF HISTORICAL CONSTITUTION 193

that historical events resemble historical evidence. Were we to have

half a dozen or so portraits of someone, say, from the eighteenth century, that is, before the availability of techniques for photographic

reproduction, and if the portraits resembled one another, we could presumably believe that the subject of the portrait resembled his portraits in the way that they resemble one another. We would thus conclude that we have some zdea as to how he looked. Quite obviously,

there is no such resemblance between our idea of some historical event and the evidence on which that idea is based, though it may be that those who tend to think of history as drawn from documents are trying to think of something analogous to that. Documents, of course, are not irrelevant. But they do not contain the complex and continuing events which emerge from the historian’s act of constitution. Nor does it make sense to think of those complex accounts which we find in books of history as generalizations, as it were, of what is found stated as particulars in the documentary evidence. A. M. Maclver may be suggesting such a relation between the two when he speaks of contrasting “the ‘generality’ of historical

statements with the individuality of the facts on which they are based.””© He goes on to say that an historian’s report to the effect that “The Normans defeated the English at Hastings in 1066” is the sort of historian’s generalization he had in mind, and that what it generalizes is all of the specific and detailed facts which go to make up the battle, its course and its outcome. The unfortunate suggestion of Maclver’s claim is that there is some way in which those particular facts actually figure in the historian’s work, whereas the actuality of the situation is that he does not know them at all. He has reason to believe that the battle took place and the outcome was what is asserted, but whatever that reason is it is not because he knows who shot whom, and so on. What Maclver has done 1s confuse the expli-

cation of the concept of battle with how an historian knows that some particular battle took place. The example of the Battle of Hastings, however, 1s not really helpful

to us here. It would not be the least surprising to learn that there are actually extant documents which make reference to that battle, and anyone who takes seriously the view that the historical facts are dug out of the documents would likely to be able to use that example © A. M. Maclver, “The Character of a Historical Explanation,’ Amistotehan Society Supplementary Volume 21 (1947), p. 38.

194 TOWARD A LOGIC OF HISTORICAL CONSTITUTION

to make his point. So it is best to drop it, taking note only of the fact that contained in the way that Maclver makes use of his example is a hint of a sound point. And that is that the historical statement is related to data of a sort unlike itself. It is not, of course, related as a generalization—in his special sense—to specific factual data. One may suspect that he had a glimmer of something, a sound intuition which was not made clear and which, in consequence, he expressed incorrectly in terms of a poorly chosen example. More helpful will be the following observation by Marrou:?’ In his thesis,” J. Schneider studied the development of an original social class, the patriarchal society of Metz. He based his findings partly on a very carefully constructed, criticized, interpreted and exploited file of some two hundred legal documents concerning financial operations, principally real estate, carried out by the bourgeois of Metz between 1219 and 1324. It would be wrong to imagine that those elementary

“facts” are more concrete, more real and more historical than the phenomenon of the collective fact—the transforming of an urban oligarchy into a landholding aristocracy. But in a still more obvious way

than in the case of small individual facts, the role of the operative procedures selected by the historian appear to have a decisive role in determining these ‘facts’ of a global type.

Marrou then goes on to discuss the problem of determining rates of variation in mortality for any given region of rural France in the seventeenth and eighteenth centuries, the accumulation of ‘elementary “fact” ’—presumably corresponding to the “concrete” facts of the quotation above-——and the determination of techniques appropri-

ate to the extraction of ‘global’ facts—the ‘collective’ facts of the quotation—from the elementary facts.” What sets Marrou’s problem is that the global fact is not contained in the elementary facts individually or taken together. If one follows the procedures of Langlois and Seignobos and other handbook methodologists who think along their lines, and sets out to extract

from each of the two hundred odd legal documents referred to in the long quotation from Marrou all of the declarative sentences it contains, subjects them to the most critical examination, and emerges from the enterprise with a collection of unimpeachable assertions each of which is said to embody a fact, or a concrete fact,’ one still does “7 Marrou, op. cit., p. 314. *8 Jean Schneider, La ville de Metz aux XIIF et XIV’ stécles (Nancy, 1950). 9 Marrou, op. cit., p. 314f. %0 The concrete or elementary facts that Marrou mentions are clearly intended to

TOWARD A LOGIC OF HISTORICAL CONSTITUTION 195

not have the collective or global fact of the transformation in medieval Metz constituted in Schneider’s book. Nor would there be any way of ordering the concrete facts so as to alter this state of affairs.

Marrou is terribly sensitive to this situation and it is the main point of his book to deal with it. Something happens when these concrete facts are subjected to the scrutiny of historians, and that something leads to the emergence—constitution—of collective facts. The passage quoted above makes reference to historians’ “operative

procedures,” and the second chapter of his book is called “History and the Historian are Inseparable.’ Marrou knows that the only way

we are going to be able to understand how historical events are constituted is to attend to those ‘operative procedures’ in ways that the handbooks seem never able to do. One may note that Marrou’s inclinations seem somewhat Kantian, and the reader expects that at anytime he will attempt to provide a transcendental deduction of the categories of historical reason. Even if one has not much sympathy for that sort of conception of the problem and doubts if it is the least

bit feasible, one must at least recognize that Marrou has a better sense of what the problem is and where the solution lies than do the

authors of typical handbooks. I should say myself that there is no way in which the quest for categories of historical reason can be successful,’ but there is surely some relevant sense of historians’ ‘operative procedures’ which is what we want to consider. In other terms, it is the elements which enter into the historian’s abductive or abduction-like procedures which must be examined.

It should not be necessary to insist that our interest is not in anything that might be called the thought processes of historians understood in some psychological sense. It is by no means obvious that there are such special thought processes, and our concern here is with the sorts of things which enter into historical thinking. Our focus is on the work historians do—work which is always intellectual and consequently always involves thought—the thought which produces it and the things thought which emerge from it, and that only. There is no need to conjure up special thought processes in order to deal with historical constitution, but it is usually a sound precaution to try to make that clear since it is virtually inevitable that someone be the facts contained in the purified declarative sentences which emerge from the Langlois and Seignobos methodology. *! T attempt to show why in the final chapter of Historical Knowing (see note 6), but

there is no need to present that argument here.

196 TOWARD A LOGIC OF HISTORICAL CONSTITUTION

will try to foist such a reading on what I am trying to do here. What

must be attended to are the kinds of intellectual or theoretical or conceptual conditions which enter into the constitution of the historical past. VI

We have seen that historical constitution is an abduction-like mode of thinking in which the historian thinks about historical evidence and emerges with a conception of what some part of the historical past was like. Obviously, the account which emerges must make sense of the evidence, but beyond that it says considerably more. We must now try to come to some comprehension of the sorts of thing which enter into the constituted account. To begin with, we might speculate about what sorts of consideration might have entered into the writing of history at a time prior to the development of the systematic, social-scientific study of human affairs. It is hard to doubt that for an historian of such a time what is plausible or not is determined by the sophisticated common sense of his tme and place. Readers of Herodotus know how frequently, after reporting some—for him—hard-to-believe account, he observes that what he reports is what he was told and that he himself accepts no responsibility for its truth or falsity. Herodotus’s own implicit doubts

about the truth of these things is clearly rooted in the fact that no educated Greek of his time could lend credence to them. They could

not be real, and no account of what purports to be a piece of the historical past could include them. When Herodotus tells us about those things he is not really working as an historian. He is simply telling us interesting things he has learned about the Egyptians, the Cimmarians and others in the course of his travels. If Herodotus could actually have been the historian of any of those peoples in the sense of constituting the course of their affairs in generations earlier than his own,” he would have to take rather a stronger stand with respect to those reports than he actually allows himself to take when he is only repeating what he had been told. In that case, even if his documents contained reports of those happenings, those reports would

have to be taken as evidence for the possible beliefs of people but 32 Which, of course, he could not, for want of suitable historical techniques.

TOWARD A LOGIC OF HISTORICAL CONSTITUTION 197

not for what actually took place. In sum, what is reasonable to believe is constrained by two factors, the evidence and common sense, and the imperatives of neither may be gainsaid. In recent times, there has been a great deal of effort expended on psychological and social scientific studies. While it is not easy to determine the extent to which these researches have affected what most people have come to think about what is possible in human affairs, it is hard to doubt that the community of scholars among whom historians must be included has been affected and influenced. A point made by John W. Hall, in a report of a conference dealing with a certain period of Japanese history, seems to be relevant here: “Significant reinterpretations of major portions of a nation’s history occur periodically as historians combine new findings with new approaches or as they pursue their inquiry with markedly different conceptions of the nature of the historical process.”*’ That new findings, things which may function as evidence, affect history’s concep-

tion of an age is rather a truism, and we need not be concerned with it here. It is the “new approaches” and “different conceptions” that are relevant to our interest at the moment, for it is these which affect the ways in which historians think about what the evidence suggests and what can be admitted into their accounts as plausible. In effect, what this means for the practice of history during the time since the emergence of the social sciences as systematic ways of com-

ing to grips with the nature of social reality and possibility, is precisely that the sophisticated common sense in terms of which matters were settled by Herodotus is no longer the key to determining what

is or is not acceptable in an historical account. Actually, these last words require some qualification. One cannot claim that all historians take the findings of the social sciences and psychology into consideration when they set out to write their accounts of whatever swath of the human past they are writing about. Thus, it is no doubt the case that any number of presently practicing historians are still affected by their own sophisticated common sense of what is socially and humanly plausible. But that sophisticated common sense is rather markedly unlike that of Herodotus and the educated Greeks of his day, for no small part of what makes 33 John W. Hall, “The Muromachi Age in Japanese History: A Report on the Conference Held in Kyoto, August 27-September 1, 1973,’ Social Science Research Council Items 27 (December 1973), 41-46; esp. p. 41.

198 TOWARD A LOGIC OF HISTORICAL CONSTITUTION

it up——no small part of what informs it—comes from the social sciences

and psychology. An historian need not be explicitly committed to the utilization of those disciplines in doing his work for him to be affected by the ideas which those disciplines are disseminating into the intellectual atmosphere. Examination of the works of individual historians not particularly associated with the endeavor to make history scientific by means of the explicit effort to introduce into it the methods and theories of the social sciences—not, that is to say, people like Lee Benson, Robert Fogel or Charles Tilly—would presumably

make it possible to discover just what some of the elements are. Unfortunately, that cannot be done here. But it does not seem unreasonable to believe that the common sense of educated people in the twentieth century has been importantly affected by the results of work in the social and psychological disciplines. It is entirely like the way in which our common sense with respect to the natural world is affected by the results of modern science even if we are not scientists and do not have much actual knowledge of natural science. Some indications of the uses of social-science theory in the constitution of the historical past have been given by Robert Berkhofer in his well-known study of the relevance of such matters to the historian’s tasks altogether.** Interestingly, he does it in such a way as to point

to the distinction I made earlier between the descriptions of eyewitnesses and historians had his interests led him to want to draw that distinction explicitly. For what he does is contrast the situation in which social scientists—who as observers of the social behavior of the groups they seek to study are like eyewitnesses—list what people do and say they will do because “all behavior would be available to the researchers in life” with that of the historian whose “data restrict

him to only partial aspects of cultural and social behavior, so he must reconstruct the other aspects from those he possesses.” It is not clear from this last passage whether or not Berkhofer thinks that those partial aspects are taken directly from the data, 1.e., the evidence,

without the application of reconstructive techniques, and if he does I should certainly wish to take issue with him. But I should want to say, in broad agreement, I think, with the spirit of his account in the

chapter relevant to our present discussion, that what an historian ** Robert F. Berkhofer, Jr., A Behavioral Approach to Historical Analysis (New York

and London, 1969), ch. 5. > Ibid., p. 99.

TOWARD A LOGIC OF HISTORICAL CONSTITUTION 199

can derive from the data which he confronts must surely be affected by what the theoretical social sciences have to tell us about the range

of human and social possibility. With that said, we may go on to agree with Berkhofer—for this is what I take him to be saying—that historians might well have reason to believe more about the human past than may be constituted directly by thinking historically about their evidence. Should what can be directly constituted in the usual way turn out to be an action or event of the sort that social scientists have related theoretically to actions and events of other specific sorts, even though the evidence does not permit the direct constitution of

instances of the latter sort, we may conclude that such did indeed take place. If we have theoretical reasons to believe that events of kind A, an instance of which may be constituted in the usual historical way by scholars working in some given field of history, are associated

in some determinate way with events of kind B, for which we have no evidence or not evidence sufficient for an instance to be directly constituted historically, our account of the tme and place in question might still make some general kind of reference—not an explicit description—of an event of the relevant kind if our confidence in the social science theories involved was adequate to the occasion. ‘Thus, it would appear that the social and psychological sciences have two roles to play in the task of historical constitution:*° they define the range of human possibility within which the constituting activity of the historian is carried out, and they make possible the hypothetical postulation of possible events which might not otherwise be constituted given the limited character of the surviving historical evidence.

vil Some may find the discussion of the present section rather unusual, yet since it continues our consideration of the way in which historical constitution is affected by intellectual commitments as to what is socially and humanly plausible, it is certainly not irrelevant to the

main themes of the paper. Our point of departure will be some observations made by Carl E. Braaten on the subject of “The Historical

*© Of course, they must figure centrally in attempts to explain historical facts, but

that subject is not our concern here.

200 TOWARD A LOGIC OF HISTORICAL CONSTITUTION

Event of the Resurrection.”*’ The question may obviously be raised as to what history might have to say about the resurrection of Jesus given how central the claim concerning it is to Christian belief. ‘The presumption has been that Christians intend that the claim be afforded no less degree of historical factuality than is accorded other claimed events of antiquity, say Caesar’s crossing of the Rubicon. Nor is it obviously clear how it could have the central significance it has for Christian belief and hope if its factuality were in any sense impugned or if it were treated as factual, if you like, but in a sense of factuality significantly unlike the factuality of things ordinary or mundane. Yet for a long time, there have been intellectual tendencies of precisely that sort, and increasingly we have witnessed the sharp separation

between faith and factuality. No doubt one element entering into this divorce is the growing realization that the traditional proofs for the existence of God are of doubtful logical validity, a realization which tended to support among the faithful the view that their faith ought to be a freely given commitment coerced by neither fact nor

logic. It is perhaps not fortuitous that one of the most trenchant critics of the proofs for the existence of God, David Hume, was also the author of an influential essay, ‘On Muracles,’*® which in effect argues that what is historically believable is what fits into the regularity

of our experience. Yet essentially the same attitude was taken by a thinker so unlike Hume as F. H. Bradley, who says, “For everything that we say we think we have reasons, our realities are built up of explicit or hidden inferences,’ and “that an inference . . . is justified solely on the assumption of the essential uniformity of nature and the course of events.”*° Thus it is evident that a number of intellectual streams went into the forming of what Braaten calls “the world view of positivistic historicism” the control of which over historical method has resulted, in his view, in its being “understandable that theologians would become hostile or merely indifferent to it.”*! Given the seemingly unchallengeable rupture of fact and faith,” the appearance on the scene of New Testament scholarship of exis“7 Carl E. Braaten, History and Hermeneutics (New Directions in Theology Today, vol. II) (Philadelphia, 1966), ch. 4. 38 Which is chapter 10 of his Enquiry Concerning Human Understanding. 39 F. H. Bradley, The Presuppositions of Critical History (Oxford, 1874), p. 10.

Ibid., p. 15. *! Braaten, op. cit., p. 43.

* Except, presumably, among fundamentalists in religion who are not or not much affected by secular intellectual currents.

TOWARD A LOGIC OF HISTORICAL CONSTITUTION 201

tential hermeneutics,** initially inspired by the work of Rudolf Bultmann, who was in turn influenced by the philosophy of Heidegger,

may well have seemed to Christian theologians as a promising way to make faith relevant to more than itself. Without entering into the details of it, which I am not qualified to do in any event, this new hermeneutical direction was less concerned—indeed, perhaps not concerned at all—with the factuality of the events of Easter than with the existential significance of faith in the Easter event, or, more particularly, what existential truth the assertions by the Gospel writers on the resurrection of Jesus contain. To take this direction 1s obviously not to be all that much concerned with the resurrection as an event like other events, and is to be impervious to the question of

whether or not such an event is possible. But can one even talk about the existential significance of faith if the faith in question takes the form of asserting that something took place which could not have taken place? Can one actually avoid the question of the factuality of the resurrection and still seek to explicate its theological significance?

That the answers to these questions must be in the negative seems precisely the point of Wolfhart Pannenberg’s assertion that, “If one assumes that the dead cannot rise, that an event of this type can never happen, the result will be such a strong prejudice against the truth of the early Christian message of Jesus’ resurrection, that the more precise quality of the particular testimonies* will not be taken into consideration in forming a general judgment.” Braaten goes on to add the following:* Pannenberg is right! Prejudices do have to be cleared away before the historian will possess the frame of mind to treat the historical evidences of the resurrection as “evidence.” Perhaps historians have never acted

more unprofessionally than when dealing with the New Testament testimonies to the resurrection of Jesus Christ. This happens when inflexible assumptions about the nature of man and of the world prejudge

that such a thing could never have happened. The historical problem * Braaten, op. cit., pp. 37-42; but see, too, James M. Robinson’s lengthy, but detailed and informative, ‘Hermeneutic Since Barth,’ James M. Robinson and John B. Cobb, Jr. (eds.), The New Hermeneutic (New Frontiers in Theology, vol. I), (New York, Evanston and London, 1964). ** As reported in the Gospel accounts: Pannenberg’s concern is with “the presuppositions under which the historical question could be taken seriously” (Braaten, op. cit., p. 97). * Quoted in ibid., p. 98. © [bid.

202 TOWARD A LOGIC OF HISTORICAL CONSTITUTION

of the resurrection of Jesus thus has two sides, the historical testimonies themselves and the whole complex process which historians use in forming historical judgments about them.

If, for the most part, his theological colleagues are willing to leave the discipline of history to positivism or to naturalism, Braaten is not. What he prefers to see done is the reformulation of the criteria which govern what we take to be possible so that it becomes possible in principle to take seriously the factuality of Jesus’ resurrection as

an event like other events. Nothing less, in his view, is compatible with its having any significance at all for Christian belief. Thus, we see what Braaten’s problem is, and we see it in such a way, namely, in the context of the broader problem of the logic of historical constitution, that the problem can make sense to us. Understanding the

problem, however, does not make it easy to find a solution that Braaten could accept. For Braaten himself admits that he “cannot believe that a historian’s judgment would lean weightily in favor of the historicity of the resurrection unless, inter alia, he were motivated to appreciate the historical basis of his actual faith knowledge of the risen, present Christ.” And he goes on immediately to express his agreement with the following statement of Alan Richardson: “Apart from faith in the divine revelation through the biblical history, such as will enable us to declare with conviction that Christ is risen indeed, the judgment that the resurrection of Jesus is an historical event is unlikely to be made, since the rational motive for making it will be absent.”*” What Braaten wants is that theologians not yield up historical reality and research when talking about the resurrection and its significance. In his view, “Historical reality that is qualified by the resurrection of Jesus as its aim and meaning does not elude the methods of historical reason; yet reason needs faith as the dynamic of its vision.” He goes on to add, “What reason sees is seen by reason, but ¢hat reason sees what it sees is made factually possible by faith.”* What shall we say to all this? Clearly, what Braaten wants is not sumply that historiography be not so wedded to certain ways of limiting

possibility as to preclude the possibility of human resurrection. And that is because he doesn’t believe that historians will be open to that

Tbid., p. 102. *8 Tbid.; his italics.

TOWARD A LOGIC OF HISTORICAL CONSTITUTION 203

possibility—and thus be open to taking seriously and on their own terms texts which purport to contain the testimonies of witnesses who claim to have seen alive one whose death they had witnessed—unless they were already committed to the truth of still another claim that is logically independent of the first claim. And that is that one particular resurrection has certain extraordinary consequences beyond the mere return to life of some person who had died. In other words, while Braaten has formulated his problem in a way which easily fits into the framework of our own attention to the question of what the intellectual elements are which enter into historical constitution, the conclusion at which he arrives is that no one is likely to be open to accepting the historical factuality of the resurrection of Jesus who is not prepared to accept the Christian idea of its meaning. It is hard to doubt that he is right. But there is a point that 1s worth drawing from this conclusion. Historians may well agree on some matters of fact without agreeing on their significance. All may agree that on the established day and at the given place a certain action was carried out. Yet it is surely possible for there to be a number of opinions on what the effect of the action was, on how essential it was or was not to the events which followed it in time. I should think that this could

be the case with respect to any event or action historians might constitute and concern themselves about. Braaten, however, says with respect to the claim of Jesus’s resurrection that it is not an historical

event which could be constituted by any historian who was not antecedently of the Christian opinion with respect to its significance.

Thus, while he wants to rescue the concept of historical fact from the restrictions imposed upon it by positivist historicism so that the resurrection of Jesus can be accorded the same status qua historical fact as Caesar’s crossing of the Rubicon, in the end he emerges with a resurrection which continues to be espistemically idiosyncratic.

Vul I began this paper by stating that what I wanted to investigate was what it is that makes an historical reconstruction—or, as we have since learned to say, constitution—acceptable, and after all these thousands of words later, what can we say about that? One possible answer to the question in general, which I have not discussed earlier

in this paper but which I have discussed elsewhere in order to

204 TOWARD A LOGIC OF HISTORICAL CONSTITUTION

defend,* has to do with what might be expected from a piece of historical constitution, what, so to speak, its logical function is in the

context of historical inquiry. And that expectation is that it makes sense of the historical evidence. Given the character of the evidence we arrive at some conclusions with respect to what the historical past might have been like, and what in effect this means is that what the constituted historical event does is explain the evidence. And this seems to leave us with the problem of the logic of that sort of explanation. ‘The logic of explanation tends to involve deduction, and if it is the function of the constituted event to explain the historical evidence, the presumption may be that the evidence must be deducible

from the event as constituted. I have long thought that such deducibility is required by the logic of the problems, but I must confess that I have never understood how it could possibly be the case. ‘Thus,

to recur to an example discussed earlier, while I have never had any difficulty in appreciating the basis of Collingwood’s conclusions with respect to the Irish colony in fifth-century Silchester, it was obvious

from the start that given the assumption of such a colony no one would or could deduce therefrom the existence of a tombstone such as Collingwood described. We have already seen that the relationship between historical events and historical evidence is not deductive, and it would thus be exceedingly difficult to determine just how it might be argued that the logical function of the constituted event is to explain the evidence—though I must admit that I continue to believe that that 1s precisely what its logical function is.

For the present, however, we may drop that question entirely, because attention to the logical function of the event as constituted is not quite what we need here. Our problem, after all, is not with how the constituted event functions, but with how it is constituted, how the historian comes to know that such an event took place or arrives at the conclusion that 7 zs reasonable to beleve that some such thing happened. Far from talking about the relation between evidence

and events already constituted, our concern throughout this paper has been with the antecedent stage during which the event emerges from or is constituted in the course of historical inquiry. And so far as that is concerned, I fear that what must be said is that there are *° See my papers cited in n. 23 and ‘Collingwood on the Constitution of the Historical Past,’ in M. Krausz (ed.), Critical Essays on the Philosophy of R. G. Collingwood,

(Oxford, 1972), pp. 262ff (In this volume pp. 312-336).

TOWARD A LOGIC OF HISTORICAL CONSTITUTION 205

no recipes for historical constitution. ‘There are no rules or checklists with reference to which we can know that the job of constituting some swath of the historical past has been done correctly. This much, however, was already apparent when it was recognized that historical constitution was not a determinate species of inference— along deductive or inductive lines—but was rather something like abduction. That kind of thinking is not the sort that applies settled canons of inference to agreed-to starting points in order to arrive at conclusions certified as sound by the correct application of the canons.

It is, rather, a species of trial and error thinking, and if we are not able to discern in it a pattern of inference analogous to those of the better studied forms of thinking, we have, it seems to me, come to discover a number of its salient features. Its point of departure is the body of historical evidence. While we now see that the collecting and ordering of the evidence is not actually first in time inasmuch as what the evidence is which belongs together and is relevant to the solution of some problem of history is itself part of the attempt to deal with the problem—is itself affected by the direction in which the historian’s work is taking him—yet with respect to some for-themoment finished product of historical constitution its point of departure is a body of historical evidence. Obviously, if the work is suc-

cessful, it has made sense of, explained the evidence—again the problem of the preceding paragraph. Yet it is clear from our discus-

sion that we have to do with rather considerably more than the evidence itself, and the acceptability of an historical account to the community of scholars will depend not merely on whether or not it satisfies some conception of the logical relation which must exist between it and historical evidence. ‘The plausibility to be sought in an historical account is not simply that of logical possibility. What can be believed is determined by the state of relevant knowledge in the historian’s intellectual community, and this is no doubt one of the reasons which precludes any effort to see the problem as directly one of deriving the historical past from the historical evidence, for the derivation is mediated by the intellectual and theoretical conditions

of plausibility. The above discussion of the relevance of social and psychological sciences to historical constitution were intended to bring

this out, and so was that of Braaten’s claims for the resurrection of Jesus as properly within the domain of a history purged of its naturalistic inclinations. The two discussions in fact complement one another in the following way. The account of the role of the social

206 TOWARD A LOGIC OF HISTORICAL CONSTITUTION

and psychological sciences points to how our conceptions of what is believable about human affairs come to develop, and how sophisti-

cated common sense is affected by the outcome of systematic research in the relevant fields. If the conditions of plausibility are independent of the historical evidence and represent the introduction of elements seemingly extraneous to the historical past, at least we have the increasing assurance that those elements need not be relative to the subjective idiosyncracies of individual historians, but may, rather,

be rooted in the best available knowledge of the time. Our discussion of Braaten’s problem, however, led to the conclusion that the resurrection with which he emerges is epistemically idiosyncratic. In the context of this discussion, what that means is that without the intrusive factor of faith, Braaten cannot achieve the results he requires to have. And that, of course, undermines the claim he is making

on behalf of the resurrection to begin with. What he had not managed to do was present the sort of argument that would have enabled historians to take seriously that a resurrection—any resurrection—1is a plausible historical possibility and that the texts which contain what purport to be testimonies concerning the resurrection of Jesus could be sound evidence for what they claim. Thus Braaten does not really succeed in enlarging the realm of acceptable possibility for historians.°° The historian, then, must work within the limits imposed by the evidence and by prevailing conceptions of historical plausibility. Within

those limits, he must persuade his colleagues that the evidence is to

be ordered in the way he proposes as relevant to the problem he seeks to deal with, and that, in consequence, what he presents in his book or essay is precisely what it 1s reasonable to believe took place once upon a time.

°° | remember attending a session at one of the annual meetings of the American Philosophical Association, Eastern Division—I no longer remember the year—in which the consequences for philosophy of recent research in extra-sensory perception was being discussed by C. J. Ducasse and J. B. Rhine. The latter tried to insist that what was called for was a radical break in the scientific world-view along some sort of spiritualist lines. The former thought that it would be possible to accommodate the findings of the research in question by means of a modification of the prevailing general orientation in the direction of some sort of subtle scientific materialism. I no longer recall the details of either position, but the relevant point is that only if something along the lines of Ducasse’s proposal could be worked out could the scientific respectability that still seems to elude Rhine’s field be won in the present intellectual environment. Braaten’s problem is similar.

IMPEDIMENTS TO EPISTEMOLOGY IN ‘THE PHILOSOPHY OF HISTORY

I

It is clear from the title that I have chosen to give this paper that the more I have attempted to come to grips with the task of wnting about epistemology in the philosophy of history practiced in the Anglo-

Saxon world, the more I have come to doubt that there is all that much to write about. No one doubts that history—the discipline—is intended to produce knowledge, and even those who are skeptical

about the success that is to be credited to it, or even about the possibility that such a discipline could achieve any success, would agree, presumably, that it is certainly the intentions of those who engage in its practice to produce a body of knowledge. And thus, one would suppose, that the philosophy of a knowledge-producing or knowledge-intending enterprise must, willy-nilly, contain some degree of epistemology. ‘To write about the product of a knowledge-seeking enterprise is, surely, to deal with something having epistemological relevance and significance. ‘Thus, it may be difficult for readers to doubt—at least at this point—that I must have a somewhat eccentric view of what epistemology is.

It will soon become clear, if I may anticipate a bit, that what leads me to take this position is the stage of the historian’s work that

most philosophy of history takes as its point of departure. In my book, Historical Knowing,' | distinguish between what I call the infra-

structure and the superstructure of history.’ By the latter I mean those products of the historical enterprise which are typically consumed by those readers who are not themselves historians, the accounts produced of times past, mainly narratives but not only such, which are intended to give their readers some idea of what some part of the human past was like. In contrast, most readers of works of history never become acquainted with the infrastructure of history, that phase of historical research during which historians apply ' Historical Knowing (Austin, 1976).

2 Ibid., 140ff., 143f., 148, 200.

208 IMPEDIMENTS TO EPISTEMOLOGY the techniques and methods of their discipline, and by thinking historically—whatever that proves to mean—about historical evidence come

to some conclusion about what it is most reasonable to believe took place in the historical past. My claim—in Historical Knowing and else-

where—is that such reliability as an account of the human past may have depends on what is done at the infrastructural phase, and that no work of history can be any better than what is done during that stage of historical investigation. In other words, if history is to be taken seriously as a cognitive—not merely literary—discipline to which

considerations of truth or falsity are relevant, it is because of the progress made over the course of centuries in the sharpening of the methodology of the infrastructure of history. Philosophical writings on history tend almost always to take their point of departure from something in the superstructure. Even where the title of a work suggests otherwise, a cursory examination of its content makes it clear that this is the case. Thus, not too long ago there appeared a work called The Emergence of the Past, but the subject

of the book is how historians deal with a past already emerged.” There are, of course, interesting things to be done with a past already emerged. One could explain it; or one could interpret it. And one could contemplate it in the belief that it must surely contain lessons for us that may be put to use as we seek to confront our present and effect our future. It may be argued that in weaving together what comes to be known

by means of the application of the methods of the infrastructure so as to produce a narrative, one makes a new contribution to knowledge. Thus, the argument might continue, attention to the way in which historians select in order to produce coherent accounts of some

swath of the human past 1s precisely to pay attention to the way in which new knowledge is acquired. And that would make attention to the way in which narratives are formed—something which belongs to the superstructure of historical work—part of the epistemology of history, which seems to be contrary to what I have been saying. In due course, I shall try to suggest that historians’ selection is largely

a myth. To select means to make a choice from a wider body of whatever it is that the choice is made from. It 1s usually claimed that historians choose from a body of facts. Others, more attuned to the idiom of recent philosophy, only speak of selecting from a collection > Dale H. Porter, The Emergence of the Past (Chicago, 1981).

IMPEDIMENTS TO EPISTEMOLOGY 209 of sentences. But, increasingly, it seems to me that the role of selec-

tion has been exaggerated, and that not attending to the way in which the historical past actually emerges in the course of work at the level of the infrastructure, philosophical writers have tended to perpetuate a myth of selection.

I In this section of the paper, and the two that follow, I shall briefly consider epistemological claims put forward in a number of comparatively recent books, since it seems best not to consider writings that have been available for a long time and, in consequence, have been already much discussed. ‘Thus, I shall not consider here wellknown works by Morton White,* Arthur Danto,’ Patrick Gardiner,° and William Dray,’ among a great many others. All four of the writers

I have named are, in one way or another, philosophers of the analytical school. There are, in addition, well-known writers who are taken to be idealists, and with whose views I find myself rather more in sympathy, Michael Oakeshott® and R. G. Collingwood,’ but their views, too, have been very much discussed over the years. After dis-

cussing a small number of more recent writings, I shall devote a section to Jonathan Gorman’s recent book,'® which, to my mind, illustrates clearly what happens when an attempt is made to say how

the human past comes to be known without attending to anything other than accounts which report what historians have reason to believe without determining why they have reason to believe what they believe. In the final section of the paper, I shall return to the question of selection in order to indicate why I have come increasingly to doubt what is surely a truism of historical method and philosophy of history, namely that in the course of producing their accounts, historians make selections. I shall begin my survey of recent contributions to the epistemology of * Morton G. White, Foundations of Historical Knowledge (New York, 1965). ’ Arthur C. Danto, Analytical Philosophy of History (Cambridge, England, 1965). ° Patrick Gardiner, The Nature of Historical Explanation (Oxford, 1952), ’ William H. Dray, Laws and Explanation in History (Oxford, 1957). ® Michael Oakeshott, Experience and Its Modes (Cambridge, England, 1933). ’ R. G. Collingwood, The Idea of History (Oxford, 1946). ‘0 J. L. Gorman, The Expression of Historical Knowledge (Edinburgh, 1982).

210 IMPEDIMENTS TO EPISTEMOLOGY history by looking briefly at Maurice Mandelbaum’s The Anatomy of EMistorical Knowledge.'' Mandelbaum has been contributing to the philo-

sophical clarification of the problems of history longer perhaps than any contemporary who writes in English, yet the book before us is

comparatively recent and has not been much discussed. The book concerns itself with any number of issues, and devotes a considerable amount of space to Mandelbaum’s views of causation. For our purposes, however, we must overlook much of it. Rather, our attention, to begin with, must focus on Mandelbaum’s conception of the varieties of history, because out of that will come the epistemological questions of

interest to us here. Mandelbaum claims that there are two basically different kinds of history, and that these are distinguished in principle. He calls them general history and special history, and he makes basically different kinds of epistemological claims concerning them. A general history is one which deals with a society as a whole. Its subject is taken by Mandelbaum to have manifest continuity, a continuity provided by a people living on its land or territory over some considerable period of time.'* Notwithstanding the vicissitudes, transformations, and changes which characterize its history, the palpable

reality of such a unit of historical interest provides the basis and possibility for the production of objective history, history containing propositions or statements which may be said to be true, true in the only sense Mandelbaum allows to be relevant to history, the correspondence sense of truth.’ The special histories are something else again. Their possibility depends not upon a sociohistorical reality presumed to have an independence which is not affected by the interest of historians, but is entirely dependent on such interest. Histories which deal with aspects of culture—histories of art, music, philosophy, or literature—are such histories. No historian could possibly deal with all of the works which

might fall within its sphere. This or that history deals with some determinate subset of what is available, a subset chosen by the scholar who produced the work that deals with it. Such unity as that subject

matter has is given it by the interest of the historian, and, thus, it cannot fail to be affected—or infected—by subjectivity. “Unlike the '! Baltimore, 1977; see my review in Philosophy of the Social Sciences 10 (1980), 341-

a Mandelbaum, 19. 'S Mandelbaum has insisted upon this as long ago as his first book, The Problem of Historical Knowledge (New York, 1938), 185ff.

IMPEDIMENTS TO EPISTEMOLOGY 211 elements in the life of a society, they are not continuous, forming an indefinitely dense series which can be explored in ever increasing detail.”'* The social world which is the object of interest of general histories may well affect art or literature, philosophy or theology; thus the special histories cannot simply ignore the general histories. Yet what the special histories study is not so inextricably tied to the social world that it cannot be abstracted for purposes specific to the special histories, and such histories may well define areas of research

that cut across the natural boundaries of the general histories. To the extent that the subject of study is determined by the historians themselves, there is no compulsively given sociality which makes possible the objectivity of the general histories, nor does there exist the possibility of imputing truth to the claims of the special histories in the correspondence sense that, as we have observed, is the only sense of truth that Mandelbaum allows to history. ‘Thus, no matter how interesting special histories may be or whatever pleasure may be had from reading them, their standing as cognitive enterprises is considerably less, if indeed they have any such standing at all. I have questioned the way in which Mandelbaum draws the distinction between the two kinds of history and the consequences he derives from the distinction as far as their respective prospects for truth and objectivity are concerned: The distinction that is so central to Mandelbaum’s account does not come from the practice of history. If the societies in which we live present themselves to us in forceful ways that preclude our being unaware of them, no reconstructed society of the historical past was ever presented to its historian in such a way. The reconstruction follows upon the accepted canons of historical research, with historians thinking historically in light of the evidence and what it is reasonable to believe. The criteria of historical truth and falsity emerge from the practice of history... and they never make reference to the brooding presence of an antecedently real society."

What we know about the history of societies and what we know about the histories of the arts and sciences depend equally upon the practice of a discipline which develops its own criteria of truth and objectivity—as well as of factuality and reference. Why should Mandelbaum suppose that the distinction he makes '* Mandelbaum, Anatomy, 19.

') My review, 343.

212 IMPEDIMENTS TO EPISTEMOLOGY between general and special history has the consequence he thinks it

has? We have already seen that in his view general histories are founded in what I called the palpable presence of societies, whereas there is nothing that corresponds to that for the fields dealt with by the special histories. But that does not resolve the question. We need to ask why he thinks that this is the case. And the answer has to do with the nature and organization of the evidence. We have seen that so far as the special histories are concerned, they are taken to depend on what historians choose to put together as warranting study. One may deal with a period of art history, but one could not possibly deal with all of the works produced. Nor could one say that the art of the period would be appreciably other than what it was if a few works in fact not created were created, or if some works which exist did not. One may choose to deal comparatively with the art of different periods or with art of different people. Whatever one chooses

to do turns out to require an ordering of the relevant evidence in ways determined by the scholar. This, in Mandelbaum’s view, is in contrast to the way in which the evidence used for the general histories is ordered. In his view, that evidence has an inherent order of its own and is in no way dependent on the work of historians. The evidence itself provides a chronological framework, and while it cannot be gainsaid that the evidence is not always consistent—one document might well contradict another on some point or other, perhaps

even an important point—this in no way undermines the way in which the evidence hangs together. Its hanging together in that way is a Characteristic of its own. It does so entirely independent of historians’ work. And it is this which points to the palpable reality that the object of a general history has, and which no object of a special history ever can.'° This view of the relation of evidence to the general histories—and the contrast with the analogous relation in special histories—is simply wrong, but the interesting question is why anyone should think

that it is mght. It is not the case that first the evidence relative to some problem of history—or general history—is piled up, and then the historian sets about the task of making sense of it by producing an historical account. Yet it can appear to be so if one simply sets about to read works of general history, especially such works as deal '© Mandelbaum, Anatomy, 171f.; this position was already fully formed in the earlier work, Problem, 251ff.

IMPEDIMENTS TO EPISTEMOLOGY 213 with comparatively recent history. Of course, for comparatively recent history there is a good deal of historical evidence, and the knowl-

edge we have as to the way it is recorded and preserved makes it more reliable than that sort of thing for earlier periods. In Histoncal Knowing,'’ I tried to show that far from the evidence being given first and, thus, capable of providing a framework—chronological or any other—for the accounts historians produce, deciding what evidence

belongs together is part of the historical problem. In the course of working out the solution, historians determine that some evidence belongs together with other evidence, that writings once thought germane to the issue do not, in fact, have any relevance to it, and so on. [he examples I used were from problems and periods concerning which the evidence is sparse, so that the historian is constantly making explicit reference to it, and readers learn more about the ways in which historians use evidence in the course of determining what the historical past was like than they typically experience in reading books on the recent past. I might add that in this respect it

is like reading works of special history. An historian of art never leaves the reader far from the evidence, since it 1s not possible to produce works of that sort without explicit discussion—descriptive and critical—of specific paintings or sculptures, novels, poems, or musical scores. And far from determining first the works of art to be studied and then proceeding with the historical study, not infrequently there may be differences of opinion—or an individual scholar might have difficulty in deciding—whether, for example, a given poem is properly part of the history of romantic poetry. Those who read the finished accounts of general histories of recent times almost never have the opportunity to see how the writers

of such works actually use evidence and how they decide, at last, what the evidence actually is that bears upon their subject. By the time the account is written, the questions of evidence have been decided, and it is easy enough for those whose attention is limited to

the superstructure of history to think that first the evidence is laid out and then the historical work is done. Likewise, given that the evidence is all there, referred to in notes, it is easy to think that this evidence both gets its order from the palpably real society whose history is being written and gives the shape to the history that emerges.

7 Py. 52-59, 128-130.

214 IMPEDIMENTS TO EPISTEMOLOGY Ii]

I next turn to Leon Pompa and the paper he contributed to the Festschrift in honor of W. H. Walsh.'® Pompa’s intention in this paper is to defend the possibility of a realist epistemology for history against

constructionist and idealist criticism of that point of view, and, as one who has himself been unsympathetic to and critical of historical

realism, | must admit that Pompa has written an excellent paper, clearly written and cogently argued. But, in the end, it fails; at least it does not convince me. The reason, as I shall try to show, is that in order to make its point it looks not within the practice of history but rather imposes from outside an interpretation of the use of evidence and its significance for our idea of the past. It is my opinion that the ease with which philosophical writers do that—move outside the practice of the discipline in order to impose upon it—is another impediment to the development of a genuine epistemology for history.

I want first to sketch briefly the position to which Pompa is responding: since his point of departure for the presentation of that is for the most part things that 1 have written, I shall simply state the position without making references. My claim is that the activity of historians takes place entirely in the present. Evidence is confronted,

dealt with critically—though the details of this need not detain us here—and the historical past made to emerge as historians seek to make sense of the evidence. A careful scrutiny of what the activity in which historians are engaged is actually like makes it increasingly

clear that the real past—a past which was presumed once to have been but is no longer and cannot be something historians or any of their contemporaries may encounter—plays no role in the formulation or testing of historical hypotheses. Thus, I was led to make a distinction between this real past and the historical past. It is the latter which emerges from or is constituted by the work of historians, and it is only the latter which has a role to play in history. Accord-

ing to this view, the real past is simply irrelevant to the practice of history and has no role to play in the determination of historical truth or falsity.

'8 “Truth and Fact in History,” in Substance and Form in History, ed. L. Pompa and

W. H. Dray (Edinburgh, 1981), 171-186.

IMPEDIMENTS TO EPISTEMOLOGY 2195 It is commonly held by those who would reject antirealist positions such as the one just sketched that only realism gives an assurance that the historical account is indeed an historical account rather than simply a story that, to be sure, may be internally coherent, yet may have nothing to do with the actuality of the human past. My own stance with respect to such criticism is to ask that the critic show precisely where the real past of historical realism actually enters into the work of historians. All too often, it turns out that for the realist the real past is something that needs to be asserted and that without its assertion there are dire consequences to be faced. I think, in the end, that this will turn out to be Pompa’s position, but before we actually reach that point it will prove interesting to see what direction his defense of realism takes, because it takes shape in the form of an important question. And that is: How can the historian know that the evidence is evidence? There must, presumably,

be some reason for selecting among all of the things within our experiential field some particular subset, asserting that it is historical evidence or historical evidence of some determinate sort bearing on

some particular swath of the human past, and constituting on its basis an account of something alleged to be historical. I suppose that what Pompa wants to avoid is the possibility that anyone with sufhcient ingenuity could select—presumably in an arbitrary way—any

subset of what he finds in his field of experience and on its basis construct what purports to be an historical account. The mere givenness of what the field of experience contains does not yield knowledge, and what is wanted 1s something that transcends the given.'” This, of course, is not peculiar to history. All knowledge

transcends what is given. Should I look through the eyepiece of a microscope on the stage of which is a culture of some sort, I would see what I suppose would be a formless glob. Even if what I saw were not formless but determinate geometrical shapes, I would not see anything in particular. Someone else, however, having appropriate biological knowledge, might look through the eyepiece and see a bacillus of determinate strain. This view of the problem is not, how-

ever, the way Pompa sees it. Nor is it to be wondered at, for I suppose that a realist would say that what the biologist and I both see is the bacillus. I may not know enough to recognize it, but that is what it is. Its being a bacillus is what it really is, and it doesn’t 9 Thid., 185.

216 IMPEDIMENTS TO EPISTEMOLOGY depend for its status as such on the existence of a concept which it then instantiates. (As if to suggest that until the history of biology reached a certain point, there weren’t bacilli at all.”°) The solution that Pompa offers to the problem he poses has to do with use of causal inferences.?! There are two aspects to this. One has to do with the claim that the kind of reasoning involved is everyday reasoning, not peculiar to history. If he is nght about the value of causal inferences in solving the problem, there is nothing excep-

tionable about this. Surely, no one ever claimed that the sort of thinking that is done in history is anything but ordinary human think-

ing, though there was—I hope is no longer—an unfortunate tradition that presumed to see in Collingwood’s claim that historians know

what happened by rethinking past thoughts an appeal to intuitionist methods requiring strange entities for objects. If historians were required to have special kinds of thinking and inferring, I cannot imagine

where they would have found them and how the discipline would have gotten started to begin with. I think that Pompa stresses the point because he must assume that a position of the sort he wishes to impugn, namely, one that insists that the historical past is constituted by means of historical research and is not tested against something “real” that is brought in from outside the framework provided by the discipline, must intend to call upon ways of inferring that are unique to history. But this is not the case, and I cannot imagine what such ways would be lke. To focus attention on causal inferences 1s to focus too narrowly. There are more than causal inferences involved in the effort of historians to determine what the historical past was like. Historical constitution is an activity rather like what Peirce called abduction, and is constrained not only by the evidence, but by the informed, sophisticated common sense of the historians’ time and place. It 1s this that determines for the community of scholars what it is plausible to believe, and in our time this 1s effected to no small degree by natural, behavioral, and social science. It would, however, take us

away from our present discussion to enter into the details of this, and so it seems best to say no more about it.” 20 Which, I must admit, is how I incline to think about it. 7! Pompa, 181. “2 But see my “Toward a Logic of Historical Constitution,” in Epzstemology, Methodology and the Social Sciences, ed. R. S. Gohen and M. W. Wartofsky, Boston Study in the Philosophy of Scuence 71 (Dordrecht, 1983), 19-52. (In this volume pp. 171-206.)

IMPEDIMENTS TO EPISTEMOLOGY 217 I observed above that there are two aspects to the way in which Pompa uses the idea of causal inference in his attempt to determine what in the field of experience is historical evidence and what is not. Pompa tells us “that something can be historical evidence only if it is public and is a consequence of causal processes which give it a character which transcends that which it has in virtue of being available to present experience.””’ That the evidence is public is not a matter of dispute. The question to be dealt with 1s how we are to take his idea of causal process which transcends the given. If I understand the drift of his discussion, Pompa’s position is that we admit as historical evidence what we admit only because there is a causal

chain which connects its present givenness to the event in the past concerning which it is historical evidence. Against this view, there are two lines of criticism that could be developed; I shall try to indicate what they are and then point to an alternative solution to the problem raised by Pompa, more in accord with the antirrealism Pompa

seeks to impugn. The first line of criticism would take its point of departure from a proposition that Pompa would dare not reject: that everything in the

present field of experience of anyone—historian or other—is the outcome of a causal chain. And thus we may wonder how Pompa picks out from all those outcomes of causal chains he experiences in a world he experiences as a causal nexus those which are historical evidence. That everything presently given to experience 1s the result

of a causal chain is a point of ontology which does not provide a principle for making the distinction that Pompa requires. Furthermore, to point to the second line of criticism, that certain determinate givens are historical evidence is surely an hypothesis which is made in the course of some particular historical investigation. They may prove to be evidence relative to the solution of the problem or they may not. That is certainly not known in advance, and it is not something the evidence zs, independent of the problem, the research,

and the proposed solution. The point of this line of criticism, as I

am sure must be clear, is that Pompa has not led us out of the woods of idealism and the swampland of constructionism to the solid ground of realism. His realism is an assumption with respect to the evidence, and one that cannot be made without begging the question. Yet Pompa’s problem is a genuine one. Not everything presented *3 Pompa, 185.

218 IMPEDIMENTS TO EPISTEMOLOGY to our field of experience is historical evidence. What direction must we take in order to find a solution? I would like to make a distinction between two epistemic attitudes which I call sensory and hermeneutic.** The former reveals what is available to be experienced by any-

one possessed of the usual sensory capacities of human beings, the latter requires antecedent preparation and is not the common heritage of mankind as such. It is clear that not everyone is capable of picking out historical evidence simply upon inspection of what is given to experience. Anyone having ordinary human sensory capacities can

experience a blue patch or a pungent smell. To identify historical evidence requires something else, and it is not too much to say that no one not possessed of an historical sense could be open to discovering

such a thing as historical evidence, nor could there be historical evidence without the discipline of history. And with the emergence of new questions and the development of new techniques, things become historical evidence which had not been so before.” All this is reminiscent of the microscope-bacillus example referred to above. That

it cannot be made compatible with historical realism seems obvious enough, and could not, therefore, serve Pompa’s purpose. IV

What I want to discuss in this section comes from the introductory chapter of Behan McCullagh’s recently published Justifying Historical Descriptions.”® It is a book that surely requires serious attention, as it deals with all manner of cognitive procedures which, in its author’s

opinion if not always mine, are used by historians as they seek to accomplish their tasks.

While the book is supposed to be about how “historical descrip** See my “Epistemic Attitudes and History,” Philosophy and Phenomological Research

37 (1976), 181-192. (In this volume pp. 130-142.) * And this change of status in no way affects anyone’s inclination to say that what the thing is is the outcome of a causal chain. The ontology of causal chains

has simply not the power to accomplish here what Pompa hopes for. Donald R. Kelley gives an example of how things can become evidence in his extraordinarily fine book, Foundations of Modern Histoncal Scholarship (New York and London, 1970).

The accretions to Roman law which early historians in the Italian and French renaissance sought to remove in their attempt to restore the original become transformed into evidence for the institutional history of the periods during which the accretions were made. °C. Behan McCullagh, Justifying Histoncal Descriptions (Cambridge, England, 1984).

IMPEDIMENTS TO EPISTEMOLOGY 219 tions” are justified, all too frequently the examples presented are about the treatment and evaluation of evidence. ‘These are two distinct areas

of activity, and while they may be part of one and the same overall course of research they have their own special characteristics.”’ No account of the discipline of history can hmit itself to historical evidence without attempting to make sense of how we came to know the historical past. ‘That Behan McCullagh is an historical realist 1s clear enough, not only from his book but from any number of his earlier papers as well. Perhaps it is that which makes it easier for him to deal with the ways in which historians treat historical evidence than with the more imaginative ways in which they call the historical past into being. In his introductory discussion, McCullagh claims that only if four assumptions are made are historians justified in believing in the truth of their conclusions. It may be noted that what he says in the book makes it entirely evident that, for him, truth sempliciter is correspondence, and that truth without qualification is contrasted with coherence.*® The assumptions which every historian—as, indeed, every seeker of empirical truth of whatever sort—must make are four in number, as follows:

(1) An assumption that the world exists and has existed independent of any beliefs about it. (2) An assumption that perceptions, under certain conditions, provide an accurate impression of reality. (3) An assumption that reality is structured according to most of the concepts by which we describe it. (4) An assumption that our rules of inference are reliable means of arriving at new truths about reality.”

This certainly does seem to be a reasonable set of assumptions, and it is hard to see how anyone could flatly declare them to be false. “7 Tt is interesting to observe that when Collingwood began to overcome the historical skepticism of his Speculum Mentis—induced, I have no doubt, by the realism to which that early work was committed—he began to think of history as the science of historical evidence. I don’t say that he stated that explicitly, but I do think

that one can find this idea contained implicitly in the articles he published in the years following Speculum Mentis, and collected conveniently by William Debbin in R. G. Collingwood, Essays in the Philosophy of History (Austin, 1965).

8 McCullagh, 2. 29 Ibid, 1.

220 IMPEDIMENTS TO EPISTEMOLOGY Yet there is no way that they could have been derived by reflection

on the discipline of history and its practice, and I suggest that McCullagh’s book is just one more attempt to foist upon history something brought in from outside. I have tried to suggest at various places in my discussion that the basic impediment to epistemology in philosophy of history is precisely the tendency not to look at the way in which the historical past is actually derived. A series of brief comments on McCullagh’s assumption will, I believe, indicate how irrelevant they are to our understanding of what history actually is. I have no doubt that if asked, historians might well acknowledge that they believe in the truth of the four statements, and some would not know how otherwise to justify their decision to become histori-

ans. Yet for all that, these are not assumptions that underlie the practice of history, nor can appeal to them serve to justify the conclusions that historians reach. Assumption (1) is the central tenet of McCullagh’s realism, yet it is entirely vacuous. No claim to knowledge in history or any other empirical discipline can be justified by asserting that a world exists independent of our beliefs about it. To say that such a world exists is to say nothing in particular. To attempt to acquire beliefs about anything is to be engaged in a disciplined endeavor, and if one believes

that that discipline is capable of achieving truth—which need not be restricted to correspondence—one believes, presumably, that a case can be made on behalf of a way of knowing such as that. Assumption (1) plays no role at all in this. Whatever one might wish to say about the relevance of assumption (2) to the justification of other disciplines, it clearly has no rel-

evance to history. It need hardly be argued that historians do not perceive the historical past. And the microscope-bacillus example would

tend to suggest that even genuine perception is not quite what McCullagh seems to be suggesting, for it is clearly infected with a conceptual component, and the conceptual component surely changes

in the course of the history of science and the history of common sense.” We have noticed that McCullagh frequently discusses historians’ treatment of evidence, and perhaps he would wish to urge that evidence—which exists, after all, in the historian’s present—is perceived. On this I would make two comments: 1) That something historical is evidence is not perceived in the sensory way even if that 0 See C. I. Lewis, Mind and the World-Order (New York, 1929).

IMPEDIMENTS TO EPISTEMOLOGY 22] something is itself so perceived.*' 2) And whatever is done with the evidence, the point, in the end, of historical research is to constitute the historical past, to provide what McCullagh—and a great many others—think of as historical descriptions. I do not know how anyone can make assumption (3) after what

we have learned during the course of the last quarter of a century about the history of science with its changing paradigms” or ideas of natural order.*’ If we know that our conceptualization is subject to change, we surely cannot assume that reality—as a realist would take

that to be—is studied in the way McCullagh says. And this is the case for the sort of conceptualizing that interests historians as well. We think of man in society in all manner of ways which were not available to our predecessors, and none of us doubts that in the course of time newer ways of thinking about them will develop. We might criticize a work of history because is rests on outmoded assumptions,

but we could never justify it—or even the entire historical enterprise—by reference to assumption (3).

The meaning of assumption (4) is not entirely clear. In a letter,” McCullagh distinguishes between strict and informal inferences, but this is not completely explicit in his book; and the language of the assumption speaks of “our rules of inference,” with “rules” sounding

more strict than informal. There are no known rules of inference that historians apply to evidence in order to get historical descriptions. I rather think that what Peirce says about the acquisition of knowledge generally applies to the constitution of history as well. He calls it “abduction,” and he contrasts it with known forms of inference as follows: “Deduction proves that something must be; Induction shows that something actually 1s operative; Abduction merely suggests that something may be” (Collected Papers, 5.171; italics in origi-

nal).°° He says “every single item of scientific theory which stands

established today has been due to Abduction,” (5.172); and that abduction “consists in examining a mass of facts and in allowing these facts to suggest a theory. In this way we gain new ideas; but *! See my “Epistemic Attitudes and History.” * ‘Thomas S. Kuhn, The Structure of Scientific Revolutions (Chicago, 1962). * Stephen Toulmin, Foresight and Understanding (Bloomington, Ind., 1981).

** Letter to me from Behan McCullagh dated January 6, 1986. * References in the text are to Charles S. Peirce, Collected Papers, 8 vols. (Cambridge, Mass., 1931~58); numbers preceding the dot refer to volumes and those following it to paragraphs.

222 IMPEDIMENTS TO EPISTEMOLOGY there is no force in the reasoning” (8.209). Finally, “Anything which

gives a rule to Abduction... puts a limit upon admissible hypotheses” (5.196). Rules of inference are not applied; there is thinking about the evidence in terms of the possible and the plausible, and new knowledge most definitely emerges. It may be that all this is acceptable to McCullagh, and is in accord with his sense of informal inference and that “rules” in “rules of inference” was an inadvertent

slip. But I must say that the stricter construction seems to me to accord more with the spint of his realism.

Vv

There are at least two points developed in Jonathan Gorman’s The Expression of Historical Knowledge that could not be accepted by anyone

who actually attended to the way in which the historical past comes to be known: that historical accounts are made up of atomic statements and that each such statement 1s the outcome of a deductive argument. ‘hese, while different, are not entirely unrelated, as we shall see; but let me say at once that each of the claims is false, and it will be interesting to determine why they are made. J have already observed that by and large philosophy of history

takes its point of departure not from the past as it emerges in the course of historical research, but rather from a past already emerged,

courses of events as they are reported to have occurred in the accounts historians write. Gorman’s book is about written accounts, not about accounts being written. The written account is made up of sentences—or sentences expressing statements—and the sentences

are all already there. Since they are already there, reading the account does not enable us to see how the sentences were generated, and it is easy to come to the conclusion that a work of history is made up of atomic sentences. To be sure, some of what satisfy our grammatical criteria for sentences may be complex and express more

than one statement, but it is no problem to deal with that. The point is that the account may be analyzed into the atomic sentences—

or atomic sentences expressing statements—which make it up, and we. may then presume to treat them in the manner prescribed by modern logic.*° Gorman’s atomism comes to color his entire concep°° Gorman, 81f.

IMPEDIMENTS TO EPISTEMOLOGY 223 tion of the statements which make up an historical account. Apparently, he takes it that atomic statements make discrete claims to truth, for he suggests that a single film shot may be likened “to an atomic truth-claim in historical writing.”’’ And he seems to think that before historians produce their accounts they select the atomic sentence they require, presumably from a supply of such sentences.”

He does say explicitly that “From within our total infinite set of possible historical statements, we may draw any number of internally consistent sub-sets, consistent or inconsistent with other sets.”%?

So far as historians at work are concerned, there is no such infinite set of possible historical statements, nor, for that matter, a finite set. Historians reconstructing some swath of the human past have no set to choose from at all. Nor are the statements which make up the written account atomic in the sense that they exist m logical independence one from the other, candidates for possible selection if only they be consistent with those chosen already. The statements which

make up the account are not atomic in any way that is the least bit interesting from the standpoint of epistemology. ‘They are brought into being altogether in the same activity of historical constitution, and when the work is finished there are no leftover statements which turn out not to have been selected. What historians write they write because the force of the evidence and the ways in which historians

think about evidence lead them to the conclusion at which they arrive—not atomic sentence after atomic sentence, but one total sense

of what some part of the human past was like. ‘To be sure, this is written sentence by sentence, but the sentences are not atomic, nor are they chosen one at a time. The one is created by the historian because it must follow—given his or her sense of what transpired— from what preceded. So the actual character of the constitution of the historical past 1s lost sight of when one’s point of departure is the finished account

which may be analyzed in the way Gorman does, as a collection of atomic statements. What is lost sight of is precisely that part of historians’ work in which knowledge is brought into being, where considerations of truth and falsity are at issue, and when issues of epistemological interest are at stake. And what I say is independent °? Tbid., 90.

°° Tbid., lf. 9 Tbid., 82.

224 IMPEDIMENTS TO EPISTEMOLOGY of my inclination to reject realism in epistemology—and particularly in history—because whether or not one is a realist, epistemological questions come into play only where knowledge is being acquired, not reported.

That the issues are distinct will be made clearer by attention to another passage in Gorman’s book. At one point, Gorman says: “An atomic statement is relevant to a description of k if and only if that

statement has an extension j which is part of the essence of k.”* Gorman’s k would appear to be one of those independent reals of McCullagh’s assumption (1), and is, thus, independent of the statements which deal with it. In the actuality of historical procedure, such reals play no role. Consider briefly the historians’ debate over Bacon’s Rebellion, an uprising against the governor that took place in Colonial Virginia in the seventeenth century, as that is presented in Murray Murphey’s Our Knowledge of the Historical Past.*' Bacon’s

Rebellion would be a & of the sort that Gorman has in mind, and he would want to claim that it is independent of the statements historians make about it. Yet careful attention to the character of the dispute as reported by Murphey makes it clear that there is no realistically independent Bacon’s Rebellion concerning which the historians are in dispute, because the different accounts of what Bacon’s

Rebellion consists of attribute to it mutually incompauble properties.*7 The issue which must ultimately be resolved is which Bacon’s

Rebellion will in due course prove to be the one we have most reason to accept. The realistic assumption of k contributes nothing to our

understanding of how the historical past comes to be known, but I suppose that anyone who was inclined to be a realist would make that assumption even if he or she were inclined to examine the infrastructure of history and not begin with accounts which he or she then analyzes into atomic sentences. Such an analysis is possible only for one who avoids the infrastructure entirely. I noted earlier that in Gorman’s opinion each of the atomic state-

ments which make up an historical account is the outcome of a deductive argument.** Given the views Gorman holds, that does not

© Tbid., 95.

*' Undianapolis, 1973), 102-111. * See my “A Note on Historical Interpretation,” Philosophy of Science 42 (1975), 312-319. (In this volume pp. 119-129.) *8 Gorman, 83ff.

IMPEDIMENTS TO EPISTEMOLOGY 225 seem to be entirely unreasonable. How else do you justify an atomic statement? Surely the best way is to derive it deductively from other

statements known to be—or taken to be—true. I suggest that had Gorman not subscribed to the views he holds about historical accounts, it would never have occurred to him that historians use deduction in the way he says. I have tried above to show why historical accounts are just not to be analyzed into atomic statements, and earlier in the paper I have indicated that Peircean abduction, rather than deduction or induction, is the way in which the historical past comes to be known. Thus, there 1s no special reason to deal with the advocacy of deduction that we find in Gorman’s book. No doubt, deduction has its role to play in our cognitive economy, but no role at all on the historian’s road to historical truth. But that is not a road than can be discerned in finished accounts of historians’ results.

Vi

I want to bring this paper to a close by considering the question of selection in history. My considering the question is more a reflection on the issue rather than an offering of a worked-out conception. I do not wish to say that selection never takes place in the writing of history. There are well-worked fields or periods, for the most part recent, for which it is not possible for an historian to include in a reasonably-sized book everything that might be known. In addition, a writer trying to present a social history of England in the nineteenth century is explicitly committed to omitting many things that he knows about the period. Such an history might have comparatively little need for the inclusion of all manner of economic and political facts,** yet the criteria of “social,” “economic,” and “political” with which the writer works may well lack the sort of precision that would enable him or her to determine unambiguously how this or that fact is to be classified. And so the writer may have to determine both how to classify it and whether or not to include it. There is an established tendency to think of selection in terms of the activity of the individual scholar. It 1s no news to readers of this * T am using “fact” or “facts” here to stand for the kinds of occurrences or entities that might engage the attention of an historian.

226 IMPEDIMENTS TO EPISTEMOLOGY journal that no small part of the discussion as to whether or not objectivity is possible in history takes its departure from the fact that the individual historian makes selections and that gua individual historian he or she 1s encumbered by idiosyncrasies, values, ideological

commitments, and the outlook of his or her time and place, all of which stand between him or her and the possibility of making the sort of dispassionate judgments that objectivity would require. History has been seen to be most burdened by this disability, though the social and behavioral sciences have not been entirely immune. That physics 1s carried on by physicists, geology by geologists, and so on for the remaining physical and biological sciences, seems not to result in such dire consequences. I suppose that some would say that so far as these last are concerned, their practitioners come to their task with criteria established by the theoretical and methodological states of their disciplines which serve to keep them from falling into the abyss of subjectivism or relativism. I do not wish to pursue this matter any further, beyond saying that one reads increasingly that the emperor of the old myth or ideology of natural science seems not to be as well attired as once we thought. Some defenders of the possibility of historical objectivity-notwithstandingthe-fact-of-selection have attempted to show that there are occasions when selections are made in terms of criteria which are—or could be—explicit and, thus, need not be subjective. And one well-known

orientation in the study of history and the social world, namely Marxism, takes it that its own scientific nature protects it from the subjectivity and relativism into which all other attempts to study history must inevitably fall.

But while reference to selection tends to be confined to the range of issues indicated by the preceding paragraph, that is not what will concern us here. ‘To speak about the selection that takes place when an historian decides to write up what is known about English society in the nineteenth century is to speak about a stage of historical work later than that stage that I find to be interesting from the standpoint of epistemology. More frequently than not in this paper I have spoken of “historians’”—in the plural—where one usually expects to find “historian’s,” singular. ‘That is because I do not think of idiosyncratic historians carrying on their individual activities from within the depths of their own subjectivity, but rather of a systematic discipline to which trained scholars are collectively committed. Scholarship creates historians’ traditions, both as to what the human past

IMPEDIMENTS TO EPISTEMOLOGY 22/ was like and how the human past is to be known, and in the course of working within such traditions historians come to alter them. I do not want to suggest that historians work and write anonymously. Nor could I deny what is obvious: that there are controversies and that controversies are identified with individuals. That is certainly the case

for a controversy that I shall be discussing a bit later on, that over Bacon’s Rebellion, to which I referred in the previous section. But in serious, not merely tendentious, cases the controversies are not to be

discussed as the confrontation of subjectivities. Nor are they controversies over what histories contain the best selection of facts or statements, selections presumed to be made from an existing stock of such things.* Where does the idea come from that historians select from a pre-

given supply of facts or statements? In one sense it does not seem difficult to conjure up an answer. Any work of history seems to be a collection of facts, and any of us could open such a work at random and start listing the facts it contains: it was the case that a, it was the case that 5, and so on until the end. And if the book contains these facts, the author of the book quite plainly put them there, that is, he selected them for inclusion in his book. That is how Gorman gets the atomic statements he thinks make up a work of history, and

it is hard to doubt that others have done the same thing. There is, however, a well-known handbook of historical method which offers an account of such matters, and, from its reputation, it may well be the source, directly or indirectly, for the views of many writers on the subject. This is Langlois and Seignobos, Introduction to the Study of History.*© In some ways it is hard to understand why it

should have had the reputation it had, since practicing historians should have discerned right off that its conceptions both of historical evidence and historical method were entirely too limiting, given the way historians actually work. Indeed, the moment it came from the

press it was not less than three centuries out of date.*’ The book does reflect a pervasive, tough-minded positivism, and perhaps that suited the mood of historians who were concerned about the status of their discipline as a science. * That seems to be the view of Morton G. White; see his Foundations, 225ff. * C. V. Langlois and C. Seignobos, Introduction to the Study of History (New York, 1966) first published in 1898 and immediately published in English translation. *’ George Huppert, The Idea of Perfect History (Urbana, Ill., 1970), and Kelley, Foundations of Modern Historical Scholarship.

228 IMPEDIMENTS TO EPISTEMOLOGY Langlois and Seignobos divide the work of historians into two parts, analytical operations and synthetic operations. The former deals with

evidence, but the only evidence discussed in the book is written documents. Without getting into the details of the analytical operations, the internal and external criticism into which they are divided, and the various activities which make these up, we are left when the analytical work is done with critically-refined evidence, each document of

which consists of simple declarative sentences, with each such sen-

tence containing a fact. It would appear, then, that the authors of the handbook read the documentary evidence the way I have been suggesting that some philosophical writers read historians’ accounts. In any event, having completed the analytical operations, the hand-

book goes on to tell us what is to be done with the fact-containing sentences which are their result. This leads us to the synthetic operations. At this stage historians produce accounts, and in order to do this they select from the facts they have amassed those they want to include. It is clear from this view of the matter that the facts are not only atomic, but are entirely independent of the narratives in which they may appear. Having gone successfully through the crucible of analysis, their truth is no longer in question. It is no wonder that so many writers on relativism in history suspect that most problems arise at the stage of selection.

I suggest that the account we have been considering is a myth. The two stages do not exist. There is no selection from an antecedently established stock of fact-containing statements. The facts and the account are constructed in the course of the same intellectual endeavor, and where an historian is breaking really new ground, not putting together an account of English society in the nineteenth century on the basis of ground already broken and largely accepted by the community of scholars, what we find in Langlois and Seignobos*®

is not to be found at all. ‘This will be more apparent after we have considered briefly an example from actual history, how historians have treated Bacon’s Rebellion.* The situation seems to be as follows: An Englishman named Nathaniel Bacon arrived in Virginia in 1674 and established a plantation there. For whatever reason, he became leader of a revolutionary movement which succeeded in *8 Cf. Morton White, Foundations, the chapter entitled “Historical Narration.” * A brief, but entirely serviceable, account of the matter may be found in Murphey, 102-112.

IMPEDIMENTS TO EPISTEMOLOGY 229 driving off the governor, Sir William Berkeley, who nevertheless managed to keep control of his military and naval forces. Before the matter could be settled by force of arms, Bacon died suddenly, leaving his followers demoralized and easily defeated by Berkeley’s forces.

Murphey, in the book to which reference was just made, presents the views of three historians on Bacon’s Rebellion, but it will suffice

for our purpose to deal with two, T. J. Wertenbaker and Wilcomb Washburn. On Wertenbaker’s view the rebellion was occasioned by an Indian War, which presented Bacon and his followers the opportunity to react against Berkeley’s persistent violation of their nghts as free men. Bacon is the hero, Berkeley the villain. The real causes of the rebellion were taken to be economic and political oppression of the Virginians by the representative of the English crown, and the

reformist intentions of the rebels are reflected in the enactment of Bacon’s Laws during the brief period that the rebels controlled the legislature. On Washburn’s view, the story is quite different. Berkeley is the hero; Bacon is the villain. ‘The latter is depicted as ruthless

and intent upon despoiling the Indians so as to take possession of their land. Berkeley would seem to be the upholder of legitimacy. There is no possibility of seeing this dispute as the result of different selections that two historians have made from a pre-given set of facts or fact-containing sentences. According to Morton White, “One historian who examines another’s history of the same subject may agree that each singular causal statement made by the other is true and yet feel that the history as a whole is inferior.”°° I rather doubt

that such a situation has ever occurred, and it is certainly not the case here. The statements in both accounts cannot all be true. They are not mutually compatible or consistent. Bacon’s Rebellion on Wertenbaker’s account of it is not compossible in the historical world that contains Washburn’s account. ‘There was clearly no common stock of statements from which the two historians made their differ-

ent selections. ,

The accounts are put together without making selections. In the course of thinking about the evidence, wondering just what evidence goes together with other evidence, and what it seems reasonable to believe, historians constitute the facts which characterize the swath of the human past they are concerned with. They do not put together atomic facts. It is because they think fact a that they are inclined °° White, 225.

230 IMPEDIMENTS TO EPISTEMOLOGY toward fact 5, and if fact 5 proves to be untenable, they may have to do something about fact a. The account hangs together because its statements are not atomic, logically independent in the manner favored by logicians, but intimately intertwined in their genesis and function.”

I said near the beginning of the present section that I think not of idiosyncratic historians making selections but rather of historians working collectively within the framework of an historians’ tradition that 1s shaped by their work. Yet even if I have impugned the idea of selection, I seem to be left with two idiosyncratic historians confronting each other over an unresolved issue. I, for one, cannot resolve the issue. | cannot even predict that some day it will be, for I cannot know whether evidence and method adequate to that task will ever be available. But it may be resolved. Agreement is often— might I say usually?—-reached in due course when the community of scholars comes to recognize that some particular view of a matter seems to be on the whole, correct.°* And when that happy time arrives,

the general agreement will be about what Bacon’s Rebellion was, and the sense of what it was will not be something constructed out of antecedent facts.

‘! This is obviously rather sketchy; I have discussed such matters in greater length and detail in work already cited: Histoncal Knowing and “Toward a Logic of Historical Constitution.” °2 Cf. Historical Knowing, 102-124.

HISTORICAL BEING Is it possible not to have a sense of the historical? I remember how surprised I was when I first saw the Church of the Holy Sepulchre

in Jerusalem, and how I had no doubt that the people who lived and worked around it surely must not have such a sense. I now suppose that I could be mistaken about that, and that perhaps what I saw was owing to their not drawing the distinction between the holy and the profane the way we do. Nevertheless, I shall proceed in the direction of my original thought in order to move my discussion

along the lines I wish it to take. What I saw when I first saw that church was that it was not set off from the everyday life and work of the people who lived and worked in its vicinity. Shopping stalls were virtually up against its walls, and while it was a place of great importance—both with respect to its holiness and its history—it was in no way separated from the mundane existence which was all around

it. Surely, I thought, if the accounts recorded in the Gospels took place not in the Middle East but in what is now the western world—

Paris, London or New York—the almost complete interaction of historical site and everyday life would not be what would be witnessed by someone coming upon the Church in its western location. One supposes that it would be separated from the world of everyday——no doubt by a fence of some sort—and rather than be open to anyone who wishes to enter, it is hard to doubt that there would be a booth at the entrance and a charge for admission. That the Church

of the Holy Sepulchre, built on the reported site of Jesus’ burial, could be so much a part of the everyday world can only be, | thought to myself, because the people who live and work around it simply lack a sense of history. There are writers who would seem to deny that human beings can lack a sense of history. Heidegger’s Being and Time, for example,

makes much of historicity as a central feature of human being. I do not wish to deal with Heidegger in any detail, and it would certainly be digressive for me to do so, but I have long believed that when Heidegger speaks of historicity, he is actually talking about temporality. It is certainly plausible that human beings live and experience in ways to which temporality is central, indeed, necessary, even if it

232 HISTORICAL BEING should not be the case that the historical sense is essential to them. Likewise, Arthur Danto is surely mght, in the fifth chapter of his Analytical Philosophy of History,’ in pointing to the way in which our

language is temporal containing, as it does, terms which are pastreferring and which cannot be used without implicating events in the past. Thus, to use his own examples, if you call the mark on my arm a “scar,” you imply that there was some past event in which my

arm was cut, and if that man 1s called a “father,” then there must have been some past course of events which led to his fatherhood. It may, of course, be the case that the referents of these remarks could be mentioned in a purely phenomenal language, but that would not be our language. Our language is unambiguously past-referring. But the temporality

of life and language is not the same as the historicity of it. It may

well be the case that ordinary language permits us to call what Heidegger and Danto deal with “historical,” and I do not presume to say what may or may not be said in ordinary language. But that

is not what I want to deal with as a sense of history. For this we need to be interested in a past which is the course of human events, and neither Heidegger’s “historicity” or Danto’s past-referentiality necessitates that. It does seem that “scar” and “father” as construed

by Danto do involve some past events, but they are not real such events, only analytical dummies. If that 1s a scar there must have been a cut, but nothing is particularly known about the specific cut of the arm and how it was brought about. Could Kant gua author of The Critique of Pure Reason have written

an essay on “natural being”? I mean, obviously enough, an essay analogous of the one I am trying to write now under the ttle “Historical

Being.” If in referring to the demg of anything one must transcend the limitations of knowing and refer to a reality which is independent of those limitations, I dare say that any number of people would conclude that such an essay was beyond the limits Kant set for knowledge: if there are things-in-themselves they cannot be known. Kant

talks of nature as appearance, and inasmuch as what transcends appearance cannot be schematized it cannot be known, cannot be the subject matter of discourse. Hegel thought that Kant’s position was easily dealt with. If there is appearance there must be something ' Arthur C. Danto, Analytical Philosophy of History (Cambridge: Cambridge University Press, 1965).

HISTORICAL BEING 233 that appears, that we cannot refer to anything as appearance without knowing that of which it is an appearance. To my mind, Hegel’s

refutation of Kant on this point is only a verbal trick, not to be taken seriously. In the end, all that Kant wants to claim is that there is no way of knowing—hence talking about-—anything except as it 1s

known to us. To reject Kant’s position would require that we be able to know about things independent of the way we actually know about them. The permanent contribution, pace the realists, of Kant’s analysis is that the things that are for us are ineluctably tied to the ways we come to know them. I must digress briefly in order to draw back from what may seem to be a rather more complete espousal of Kant’s position than I care to make. I have no difficulty in accepting Kant’s tying the objects

which make up our world to the ways in which we come to know them——come to constitute them—but I think that in its specific details Kant’s own conception is entirely too ngid. He ties the categories, which are the conceptual tools by means of which we schematize

the givens of sense so as to produce an organized world, to our shared rationality, thus making them our permanent possession devoid of all historicity. If Kant were nght about this, one consequence

would be that we humans would always have lived in the same unchanging world, but increasingly research in the history of sci-

ence, inspired by the vision of Thomas Kuhn? and Stephen Toulmin*’—who were themselves anticipated by R. G. Collingwood*—

makes that a dubious belief. And if the world does seem to change in the way indicated, then, to speak in Kantian terms, the categories by means of which we schematize the givens of sense must themselves change in the course of time. That our world is constituted by us in the course of being known I take to be the permanent heritage of the Ist Critique, but since its appearance any number of thinkers have attempted to show how we structure the world in the course of knowing it without accepting the view that the conceptual apparatus in terms of which this is done is permanent and unchanging.” * Thomas S. Kuhn, The Structure of Scientific Revolutions (Chicago: University of Chicago Press, 1952). > Stephen Toulmin, Foresight and Understanding (New York: Harper & Row, 1963). * R. G. Collingwood, An Essay on Metaphysics (Oxford: Oxford University Press, 1940).

° "A valuable historical overview of this from Kant into the 1940s may be found in Thelma A. Lavine’s “Knowledge as Interpretation: An Historical Survey,” Philosophy and Phenomenological Research, vol. 10, no. 4, June 1949, pp. 522-40, vol. 11,

234 HISTORICAL BEING But what shall we say in answer to the question as to whether or not Kant gua the author of the Ist Crtzque could have written an essay on natural being. Clearly, there are those who would insist that he could not, for, they would say, from the conception of knowing that that Critique presents there could not be any sort of human access to natural being. We could have only the appearance of being that could never itself be known. On such a view, if Kant were right there could not be any use for the expression “natural being.” This response, however, makes sense only if there is something to be said on behalf of a nature which is presumed to exist notwithstanding that there seems to be no way for us to know about it. It is surely a distortion to say that for Kant there cannot be natural being. ‘There most assuredly is such a being for him, but it is tied completely to the way we come to know it. It is nature-as-known, and it is too much to think of characterizing a nature as it is independent of being known. It would not be quite correct to say, for example, that for Kant appearance, the phenomenal world, is Euclidian

and Newtonian, but nature is something else again. For him, it is precisely that nature is Euchdian and Newtonian: to talk about a “some-

thing else again” is to talk about what cannot be talked about because it would be to talk about what cannot be known. There zs nothing to talk about apart from what we are capable of knowing, and while realists would like to fall back on an independently real as

it 1s apart from being known, there is no such thing. In our own time, natural being is not Euclidian and Newtonian because nature as we know it cannot be known in that way. In that he took it that natural being as it was then known was the only way it could be known, a reflection, as I have ahead indicated, of his unhistorical conception of the categories, I should want to say that Kant was mistaken, but in taking the view that natural being is nature as known, he surely was not.

Natural being is not the subject of this paper, and it is time to drop the matter. I introduced it only as a pointer to the position I shall be taking with respect to historical being. I began by asking if it is possible not to have a sense of the historical, and I am not sure no. 1, September 1949, pp. 88-104. More recently, this material has been incorpo-

rated in the same author’s attempt to bring her account up to date called “The Interpretive Turn from Kant to Derrida: A Critique,” in T. Z. Lavine and V. Tejera, eds., History and Anti-Aistory in Philosophy (Dordrecht, Boston, London: Kluwer Aca-

demic Publishers, 1989), pp. 32-121.

HISTORICAL BEING 239 that I know how to answer. I cannot believe that an answer is possible aprion, and that suggests that an affirmative answer would require the actual discovery of people who failed to have such a sense. As I indicated above, I once thought that the behavior of the people who live and work near the Church of the Holy Sepulchre proved them to be without an historical sense, but I do believe that that judgment is premature. Likewise, when I heard it said that the monks at Santa Katarina—an Orthodox monastery atop what is taken

to be Mount Sinai—would use old manuscript to start fires, I had no doubt that they lacked a sense of the historical, but that judgment, too, is premature, for one would have to know considerably more about people than just one kind of behavior to know what sorts of beliefs they held or the ultimate presuppositions to which they adhered. Most-——perhaps all—people have a sense of their own past. There are traditions with respect to the people’s origin, myths concerning how institutions and practices came into being, accounts of the beginning of one thing or another. This sense of their own past, however, is surely not an historical sense. Few of us would agree that the genres in which these myths and traditions are expressed are history.

So it is clear that merely having a sense of the past of one’s own people or community is not the same thing as a sense of history. Certain books of the Hebrew Bible—say, Joshua, Samuel, and Kings—

are called historical books, and inasmuch as there are many who think of history in terms of the narrative mode of expression,° there would seem to be no difficulty in speaking of the historical sense of biblical Israel. But if one does not incline to think of history in terms of a particular literary mode, one could just as easily think of those biblical books as records of the traditions of ancient Israel recorded

and preserved in a certain way. Why call them “histories”? One doesn’t call War and Peace a history; we certainly distinguish between

a work of history and an historical novel. And one doesn’t think of the /had as history—though it may function as an historical source or piece of evidence—even though it, too, preserves a people’s tradition concerning its past. It may be possible to say that there is something historical about War and Peace and about the book of Samuel, yet for all that, I should ° Among other well-known philosophers of history, one thinks of F. R. Ankersmit,

W. H. Dray, W. B. Gallie, L. O. Mink, M. G. White and H. V. White.

236 HISTORICAL BEING want to say that they are not historical. And that is because they are

not produced in the historical way, they are not the result of the kind of research that the discipline of history requires.’ I shall not attempt to characterize the nature of that discipline, because it itself

has had a history, the way it 1s carried on has undergone change over the centuries and, indeed, millennia, and the point I am trying to develop in this essay doesn’t particularly depend on the character

of its practice at any particular time. But it is the practice that 1s central to what I want to say, for I want to tie historical being to what can be known in the historical way, precisely as I tried to suggest

above—in the broad spirit of Kant—that natural being had to be tied to the ways in which nature is known, that there could be no transcending of natural knowing in order to encounter a real naturein-itself.

It is obvious enough that what I have said so far opens me to the accusation of scepticism, and I should like to say a few words to counter that. I have long believed that there is no such being as “the sceptic” although one certainly does read a great deal about him in the literature. ‘There are two ways of being open to the charge of being a sceptic, one owing to the views that you yourself hold and the other owing to the views of another, namely, of the one who charges you with scepticism. With respect to the first, I really cannot believe that Humeans are subjectively sceptical about or in any way actually doubt the existence of other selves, that they take seriously the possibility that they—each of them—may be alone in the world. But having committed themselves to the epistemic primacy of sense impressions, how to account for our knowledge of the other becomes problematic. I do not believe that sense-data empiricism has the conceptual resources with which to resolve the problem,® but that

need not detain us here. But that they try so vigorously to solve what they call the problem of other minds—undeterred by successive failure—is clear evidence that they have no doubt about the otherness of the other or about the possibility of resolving the problem. ” See my Histoncal Knowing (Austin, TX: University of Texas Press, 1976); “History and the Primacy of Knowing,” History and Theory, Beheft 16, vol. 16, no. 4, 1977, pp. 29-52 (in this volume pp. 143-170) and “Toward a Logic of Historical Constitution,” in R. 8S. Cohen and M. W. Wartofsky, eds., Epistemology, Methodology, and the Soctal Sciences (Boston Studies in Philosophy of Science, vol. 71), (Dordrecht,

Boston, London: D. Reidel, 1983), pp. 19-52 Gn this volume pp. 171-206). ® See my “Why the Problem of Other Minds?” The Philosophical Forum, vol. 2, n.s., winter 1971, pp. 271-77.

HISTORICAL BEING 23/ In the end, what their alleged scepticism comes down to 1s that making

a philosophical commitment of a certain sort simply opens them to difficulties of another sort. Now, to the second basis for the accusation of scepticism, a view of the accuser himself. In Maurice Mandelbaum’s review of Collingwood’s The Idea of History, we find the following words: “Collingwood’s

thesis that historical knowledge consists in the re-enactment in his own mind of the thought which underlay past actions must, I be-

lieve, inevitably lead to skepticism.” It could be suggested that Mandelbaum’s use of “I think” implicitly anticipates the point I want to make,'? namely, that there is nothing sceptical about Collingwood’s position per se: It is sceptical only if one makes an assumption that Mandelbaum makes, namely, that a true historical account is true in virtue of describing a past event as it really was when it was present, that history satisfies the correspondance theory of truth. If you think,

as Collingwood did, that historical truth is what the weight of the evidence makes it reasonable to believe, then Collingwood’s notion of re-enactment is not the least bit sceptical. If you read the part of The Idea of History in which it appears, without the blinders of philosophical presuppositions, it is hard to avoid seeing that far from leading

to scepticism, it is a central notion in Collingwood’s attempt to show how the historical past comes to be known. That Mandelbaum takes Collingwood’s view to lead to scepticism is entirely owing to the fact that he can see no way for the real past

as it was when it was present to emerge into the light of day if Collingwood is right. History, for Mandelbaum, who has always in-

sisted upon historical realism, is about the real past, and any doctrine which states or implies that the real past is not to be had, that all we can have is what there is reason to believe, is, perforce, sceptical. For a sceptic, after all, is one who denies that knowledge, or

knowledge of a certain sort, is to be had. Even if one could show that no one could have knowledge of the sort in question, one would still be sceptical with respect to it; in the matter before us the claim will be that historical knowledge is impossible. And that suggests that there is something defective in our capacity to acquire knowledge.

> Fournal of Philosophy, vol. 44, 1947, p. 187. '0 And made more than twenty years ago in “Collingwood’s ‘Theory of Historical Knowing,” History and Theory, vol. 9, no. 1, 1970, pp. 3-31; p. 3. (In this volume pp. 273-311; p. 273.)

238 HISTORICAL BEING No one would take seriously the claim that the mathematical skills of so-and-so are defective in view of his incapacity to determine the highest prime number, because there is no possibility of discovering a more competent mathematician who can. But Mandelbaum wants to claim that the scepticism he discerns in Collingwood’s account is a peculiarity of that account. It precludes taking seriously the possi-

bility of determining what the real past was like, whereas in Mandelbaum’s view that is precisely the purpose of historical research. On Mandelbaum’s view, history makes available to us an account of

historical being which is not simply an account of the human past— or some part of 1t—which 1s only what it is reasonable to believe. We seem to have a situation analogous to what we find in Kant. Kant speaks of a nature which is known to us because we constitute it by means of the categories of the understanding which we apply to the manifold of sense we apprehend through the forms of intuition; things-in-themselves are unknown to us and are of which we cannot speak. Collingwood seems to speak of an historical world, which is known to us because it is the outcome of our constituting"! research activity, and an historical-past-in-itself concerning which there

is nothing to be said.’? It is this distinction that Mandelbaum wants to get around. There is a tradition with respect to the way in which historians

produce their accounts which may seem to open the way to the achievement of Mandelbaum’s goal. We are often told that the histonians proceed by selecting from an apparently pre-existent supply

of facts those that they need in order to produce the accounts— usually presumed to be a narrative though, in fact, this is not always the case'*—they wish to produce. This formulation seems to suggest

that a distinction has to be made between the facts of history and historians’ accounts, and, clearly those facts are known to historians. How else could they choose among them? And, surely, those facts which are not selected do not just disappear. Rather, they continue to exist independently of historians’ accounts, waiting, perhaps, to be chosen on some future occasion. '' The term used that way is mine, not Collingwood’s; see my Historical Knownng, op. cit., p. xx and ch. 3; “History and the Primacy of Knowing,” op. cit.; and “Toward a Logic of Historical Constitution,” op. cit. '2 Nothing to be said by us. On some views, God could be said to know the real past as it was when it was present, but then, I suppose, he would have no need for an historical past as constituted in historical research. '5 See my AMistorical Knowing, ch. 5.

HISTORICAL BEING 239 The idea that historians select in the manner indicated has been with us so long that it is simply taken for granted and presumed to be true. For all that, there is nothing to it: it is simply a myth.’* It is not the case that the historian selects what he needs for the particular account he wants to write leaving what he doesn’t need without in any way impugning the integrity of its existence. It 1s not that I want to say that there is never any selection in the production of writings on the historical past. A great deal is known about the history of England in the nineteenth century. Someone may choose to write a social history of England during the reign of Queen Victoria, either because he and his publisher believe that there is a market of people interested in that subject or because it is felt that the textbooks available to introduce the subject to students are defective in various ways and a new one lacking those defects is needed. The writer would then select social historical facts and omit economichistorical facts. And since what is known about the social history of

the time may be far too much to be included entirely in such a book, a further selection may be needed. But none of this has to do with the creative constitution of the historical past in the course of genuine historical research. It is, rather, the sort of thing that follows upon genuinely original historical writing and may be thought of as parasitic upon it. Such work ought not to be denigrated. It is, surely, appropriate to make available to larger audiences the outcome of academic scholarship, and it requires special skills, particularly in

writing, to do it well. But it is simply not the same thing as the genuine advancement of historical scholarship, that stage of historians’ work on which the historical past and the facts which make it up are brought into being. When one looks carefully at the character of disagreement among historians working at the genuinely creative level of their discipline, one does not find, pace Morton White,’® that each of them has made a different selection from a pre-existent collection of historical facts or, as White prefers to put it, from a collection of statements each of which is known to be true and each of which embodies an historical '* See my “A Note on Historical Interpretation,” Philosophy of Science, vol. 42, no.

3, Sept. 1975, pp. 312-19 (in this volume pp. 119-129) and “Impediments to Epistemology in the Philosophy of History,” Anowing & Telling History, History and Theory,

Betheft 25, vol. 25, no. 4, 1986, pp. 82-100; pp. 96-100. (In this volume pp. 207230; pp. 225-230.) ' Morton G. White, Foundations of Historical Knowledge (New York and London: Harper & Row, 1965), p. 227.

240 HISTORICAL BEING fact. Inasmuch as I have discussed an example in appropriate detail

in each of the two papers cited in note 14, I shall not repeat that discussion here, but I do want to mention what the outcome of that

discussion was, because it is that outcome that is relevant to the present attempt to characterize historical being. Our point of departure was the examination of an account of two different historians’ views of the nature of an historical event called Bacon’s Rebellion,

which took place in the colony of Virginia in the 1670s. The two accounts clearly diverge, and that, on the selection-from-antecedent-

facts pomt of view, ought to mean that each of the two historians has chosen to select different facts in order to emphasize the side of the rebellion he chooses to emphasize. Uhat, however, is clearly false. The facts which make up one account are simply not compatible with those which make up the other. According to one account the colonial Governor, Sir William Berkeley, was the oppressor, and that Nathaniel Bacon led a rebellion against him in order to defend the rights of the settlers, rights they possessed as Englishmen. ‘The other takes it that Bacon was motivated by greed, and that the governor was the moderate leader of the colony who tried to keep in check the consequences of that greed. This, | think, will do for our present

purpose. I cannot believe that advocates of the selection point of view could be open to the possibility that the pre-existent facts from which the choices are made could, at least in some cases, be logically incompatible. Yet the characterizations of these leading figures of Bacon’s Rebellion, Bacon and the governor, are in fact not mutually compatible, they are not compossible in the same universe. And so they cannot be pre-existent, just waiting to be selected. ‘Thus, we may ask, where do they come from?

I could say that historians make them up; but I do not want to suggest that there is anything arbitrary about the way they come into being. As it is, there 1s nothing I could say along the lines | am pursuing here that would dissuade realists from thinking that my view is arbitrary given the realistic tendency to take its stand with reals which are independent of the ways in which knowledge—of them— is acquired. Nevertheless, I should want to say that historical facts come into being in the course of historians’ work. ‘This involves the application of historical method and historical thinking to the data— evidence—historians use. One might say that the evidence is, in fact, the point of departure for the historians’ work, but that suggests that

first the evidence is piled up and then the task begins. The fact,

HISTORICAL BEING 241 however, is that how the evidence is piled up, what evidence belongs with what other evidence, is itself the outcome of historical research,’®

so I shall not put it in that way. It is easy to think that an historical account is simply a collection of discrete historical facts. If one opens a work of history and proceeds to read, that is what one seems to find. Sentence after sentence reporting fact after fact, each put in its place by the historian

who wrote the book. The book, after all, is made up of discrete sentences, and each sentence has something to tell us. The historian

who produced the book decided what he would tell us and—one presumes—what he would not, and where, one wonders, did the facts he chooses to report come from? Clearly, one concludes, they are elements in an independently existent past, elements from which he chooses. Otherwise, how can we say that his is an account of a swath of the human past, not something he simply invents.

That, as I say, is what might seem to be the case to one who simply reads what historians write. It 1s a conclusion with respect to written accounts, but it is not a conclusion with respect to accounts being written. If we attend to what goes on while the historical work— the genuinely creative historical work as referred to above—is being done, one sees something else entirely. It 1s because what he does

leads the historian to say something in particular, that he is led to say something subsequent. Should it turn out that that something subsequent cannot be said, the evidence makes it difficult to say, then the historian may have to re-think the plausibility of the first thing. All I mean to say is that the revision or even rejection of the second thing he wanted to say may prove to be just incompatible with what he started with, that the two are not compossible in the same historical universe. The entire account emerges from the same intellectual effort, with the historian changing, revising, amending what

he writes, until a logically coherent account is produced which is reasonable to believe, is consistent with the evidence, and makes sense of the evidence.

The facts, then, far from being independent of the historian’s account are created in the course of the account’s being created. That we may consider a possibility, then revise it, reshape it or deny it entirely, cannot be made consistent with the independent pre-existence of historical facts from which selections are made. Thus, I '© See Historical Knowing, pp. 47-49 and 57-59.

242 HISTORICAL BEING would say, the facts are constituted in the course of the historians’ creative work. Their being is dependent on the practice of the kind of discipline history is. In the example of Bacon’s Rebellion alluded to above, it is clear that different historians have produced conflicting accounts each of which constitutes incompatible facts. This is hardly an acceptable situation, and we cannot allow that Nathaniel Bacon, so far as the rebellion 1s concerned, is both hero and villain. But the state of historical knowledge does not remain eternally what it was at the moment that the second of the two accounts appeared, and one may hope that in due course the community of scholars will find a way to emerge with a generally consistent account.'’

Thus, we seem to come back again to Kant. For him, nature emerges from the constituting activity of human reason when the categories of the understanding are applied to the manifold of sense.

I do not want to say that there are permanent categories of the historical understanding grounded in the common rationality of historical practice. On the contrary, there is a history of the discipline of history, and in the course of time the discipline changes as new methods of historical work are brought into being and tested as to their usefulness. But, that, pace Kant, is what we have seen with respect

to nature, Kant’s own belief that the categories are fixed and final and are grounded in our common rationality being no longer taken seriously. I should want to say—in history as in science—that our world is the result of our ways of knowing. I can sympathize with the anxieties that lend realists to want to believe in a world independ-

ent of us and our ways of knowing, but I do not see how we can talk about what we cannot or do not know.

'7 For all I know, such may already be the case; I certainly have no expert knowledge of the historiography of colonial America.

THE PAST OF OUR PRESENT The natural assumption of common sense is that the past is over and done with, not subject to change or being changed. I should suppose—though I don’t really know—that not even theologians obsessed with the omnipotence of God could believe that God has the power to change the past, certainly not the past understood in - the realistic way. We may not know all that much about the past, yet it is all there, finished, unaffectable, waiting, as it were, to be discovered by historians—or by archaeologists, geologists, paleontologists, or whatever.

It is that past which Maurice Mandelbaum called upon to rescue history from the pitfalls of relativism. The relativism he sought to overcome in his surely classic work, The Problem of Historical Knowledge,' is an axiological—not epistemological—relativism, the view that

historians are, in the end, prisoners of their own values, or of the values of the class or community of which they are a part, that these values must always affect the work they produce. In consequence, the ideal of an historical truth which is recognizable and acceptable by all who are competent to judge is simply a chimera.

Mandelbaum’s attempt to rescue history from the situation just described involves making a distinction between fact and value, judgments of fact and judgments of value.” There is, in his view, a fac-

tuality which is inherently independent; independent, of course, of judgments of value, since the relativism he seeks to refute is, as already indicated, an axiological relativism. ‘he historically factual 1s grounded

in a reality which is independent of human intrusion, and the view of historical knowledge that Mandelbaum seeks to justify in his books “leads us to assume the philosophic correspondence theory of truth.”’

This can only mean that an historical statement—a statement produced by an historian as the outcome of historical research, the only sense of “historical” that is relevant to this essay—is true if and only

' Maurice Mandelbaum, The Problem of Historical Knowledge (New York: Livenght Publishing Corporation, 1938). ? Tbid., Chap. 6. > Ibid., p. 185.

244 THE PAST OF OUR PRESENT if it corresponds to the state of affairs that existed when that swath of the past to which it refers was present. Even if we allow that Mandelbaum has somehow freed us from

the scourge of axiological relativism—that he has or not is not, in the end, all that central to what happens—-what are we to say about a past the existence of which is independent of us and our cognitive capacities? What can we know about a past which is impervious to our knowing? Can there be such a thing as a past in itself analogous to the thing-in-itself to which reference is made in Kant’s Critique of Pure Reason? Kant, too, was intent upon overcoming relativism, though

in his case it was an epistemological relativism. Hume taught that all

knowledge of fact was grounded in the impressions of sense, and that led to scepticism in two ways. First of all, knowledge was said to

begin with impressions, not with external objects which might be thought to cause us to have the impressions, and so the existence of the world outside our sensory apparatus became a problem.* And, second, since each of us experiences his or her own impressions and

cannot compare them with those of others, we seem left with an epistemological relativism, the claims to knowledge of each of us being

relative to the impressions of each of us. Kant presumes to solve the

second problem by producing a table of categories of the understanding which is purported to be grounded in our common—hence

shared—human rationality. Thus, while each of us apprehends a private manifold of sense, we share a common world because it 1s structured by the categories we all have in common.” But if that eliminates knowledge as relative to the knowing subject, it does nothing to alleviate the difficulty that some—realists— would feel about the seeming inaccessibility of external things. Kant himself seems not to have been bothered about that. He was satisfied to have overcome the problem of subjectivism and was, apparently, willing to give up a realistic universe with respect to which the correspondence theory of truth obtains. He was satisfied that we could know a phenomenal nature constituted by the categories of the un* To be sure, it became a problem as early as Descartes, but an historical sketch of the problem of the external world would not serve our present purpose. 5 For an excellent though brief historical account of the problem of categories from Kant onward, see Thelma Z. Lavine, “The Interpretive Turn from Kant to Derrida: A Critique,” in Thelma Z. Lavine and Victorino Tejera, eds., History and Anti-History in Philosophy (Dordrecht, Boston, London: Kluwer Academic, 1989), pp. 32-121.

THE PAST OF OUR PRESENT 245 derstanding, and was thus willing to give up on things-in-themselves. What is implicit in Kant’s account 1s that to think that we can know

about and discourse about things-in-themselves is to think that we can know about things as they are, independent of our knowing about them. How can that possibly be?

It seems to me that Mandelbaum’s conception of the solution to the problem of relativism in history comes down to the same thing, a belief in the possibility of our knowing of a past-in-itself as it is, independent of our knowing of it. It may be that he failed to see that this is the case because his attention was focused on a relativism of the axiological kind, and if we could make out successfully his distinction between judgments of fact and judgments of value, that sort of relativism seems to be defeated. But that distinction has nothing to say about another kind of relativism, an epistemological, perspectival kind of relativism which takes its stand on the claim that all knowledge is relative to the ways in which it becomes known. If one

can take that view seriously then it becomes increasingly doubtful that historical realism and the correspondence theory of truth are what make it possible to overcome relativism. It is that possibility that I wish to explore in what follows, and I should like to take as my point of departure a passage from the writings of George Herbert Mead for which I have always had a special fondness, and which unambiguously embodies a perspective which is in marked contrast to that of historical realism: The outcome of what I have said is that the estimate and import of all histories lies in the interpretation and control of the present; that as ideational structures they always arise from change, which is as essential a part of reality as the permanent, and the problems which change entails; and that the metaphysical demand for a set of events which 1s unalterably there in an irrevocable past, to which these histories seek a constantly approaching agreement, comes back to motives other than those at work in the most exact scientific research.°

For our purposes, of course, “the most exact scientific research” includes the discipline of history, the only path we have to the historical past.

© George Herbert Mead, The Philosophy of the Present, ed. Arthur W. Murphy (Chicago and London: Open Court Publishing Company, 1932), p. 28. The entire chapter from which this is taken, entitled “The Present as the Locus of Reality,” is worth careful reading.

246 THE PAST OF OUR PRESENT The quotation from Mead makes its point with respect to the past,

but its point is quite general. Any attempt to take one’s stand on reals which are alleged to be independent of inquiry is motivated by

commitments which are independent of the systematic quest for knowledge. As I read the recent history of recent philosophy of science, with its dependence on the history of science to a degree which was simply unheard of to the philosophy-of-science orthodoxy of my student days (the early 1950’s), realism with respect to the objects of scientific knowledge has become increasingly less tenable. If such writers as Thomas S. Kuhn’ and Stephen Toulmin® are correct, what we find in the course of the history of science is that ways of construing nature themselves undergo change over the course of time, and, in important ways, they are not mutually compatible, they are not compossible in the same universe. Thus, unless we wanted to suppose that from time to time the universe itself undergoes funda-

mental changes in such a way that at one time the account of nature we find in Aristotle was indeed a realistic account of nature while at another that of Newton was, it is hard to see how realism can be a sound account of the scientific enterprise. Why in recent times there has been such a plethora of defenses of scientific realism is, perhaps, an interesting question, and I rather suspect that its answer

lies in the realm of psychology, but it is not a matter to be further discussed here.

The focus of our attention here is on the past, more particularly the human past as that is constituted in the course of historical research. I should want to say that the past for us is only a past that is constituted in this way, and to think of a human past that is really there, finished and unchanging, is to speak of something that can have nothing to do with us, of something we cannot describe. I am speaking of the human past in a cognitive, epistemologically responsible sense, a sense of the past as that would emerge from “the most exact scientific research,” to use words we have found in the quotation from Mead. In the course of this discussion I shall try to make it seem reasonable that the historical past for us is a past that emerges

in the course of historical inquiry, but I am certainly aware of the ’ Thomas S. Kuhn, The Structure of Scientific Revolutions (Chicago: University of Chicago Press, 1962). ° Stephen Toulmin, Foresight and Understanding (New York and Evanston, IL: Harper

& Row, 1963).

THE PAST OF OUR PRESENT 247 fact that we live in a variety of pasts. But I shall want to say that regardless of possible popular belief to the contrary, those pasts are not of the sort to which truth and falsity are relevant. hus we live in the past of our tradition, what I should call the mythic past of our

social community. As a people we are heir to a sense of our past, and we might well feel threatened in our sense of who we are if questions of truth be raised. Arthur Danto tells of a place in France in which is located a cannon which the natives take to be “a cannon placed here by Francis the First after the Battle of Cérisoles in 1544,” but while this may be an item in the community’s sense of its past, it is not a statement established in the historical way. David Carr, influenced by the philosophical method of Husserl

and Heidegger, attempts to show that narrativity and temporality are not, as he takes it that some philosophers of history claim, imposed by the manner in which historians produce their accounts, but are given pre-thematically.? By that he means that temporality and narrativity are already features—I presume ontological features—of human actions, and so must of necessity be elements of the accounts which historians produce. He begins with individual actors and know-

ing subjects, as you expect given the Husserlian foundation of his approach, and tries to move from the individual to the group, from the I to the we, which has a history. What Carr wants to show is that those aforementioned ontological features of individual actions are carried on over into events which concern groups. If this could be made out, he seems to think, then he would have shown that temporality and narrativity are inevitably elements in the works of historians. Whatever one may wish to say about Carr’s move from I to we, the case he wants to make about historical writing does not come off. He never moves to the discipline of history. In every case, a careful reading of his text makes it clear that he never moves beyond the “history” which is a people’s myth of its past. Thus, he says: “Ihe group looks ‘backwards’... to its own origins in the individuality and cross-purpose of its members, which has been surmounted by their mutual recognition and reconciliation.”!® That he can go on immediately to add that “it [the group] looks forward to * David Carr, Time, Narrative, and History (Bloomington and Indianapolis: Indiana University Press, 1986).

Tbid., p. 148. The entire chapter from which this is taken has any number of other statements of the same sort. He talks about what the group takes its past to be, not about what historians have come to know about its past.

248 THE PAST OF OUR PRESENT the carrying-out of its common tasks and it projects its own continued existence as the condition for this continued activity” makes it

clear that his focus is on a group’s sense of itself and not on the outcome of historical inquiry.

What we are concerned with, then, 1s a human past as it comes to be known, not as it was in itself when it was present. There are, of course, risks involved in such a conception of the historical past, for if historical truth is not determined by correspondence to the real

past how can we ever be sure that we have achieved it? And how can we be sure that historical truth is not only a mask for special interest and propaganda? I write these words less than two weeks before the fiftieth anniversary of the Japanese attack on Pearl Harbor which brought the United States into World War II and was greeted with great anger, bitterness and a sense of betrayal by the American people. Richard Reeves, the syndicated columnist, begins a column devoted to the subject with the following words: “Remem-

ber Pearl Harbor! The American battle cry of World War II has become a question in the 1990’s: Remember Pearl Harbor? And the correct answer is: Not me.”'' Given the present relationship between the United States and Japan, particularly, though not entirely, in the

economic sphere, Reeves reports that there is a tendency to softpedal the observance of the anniversary; and that suggests that the historical past may be re-negotiated, as it were, in order to satisfy other needs and interests. ‘That, surely, is the point of the closing words of Reeves’ column: “Why should history and other unpleasantness disturb commercial relationships? In these times, it is only appropriate that history, too, should meet the tests of free markets. If history is not good for business, then we should forget about it— until it is time to repeat it. Forget Pearl Harbor!” It is not reasonable to believe that the needs of Japanese-Ameri-

can relations could lead to a thoroughly re-negotiated past of the past-for-us, that the idea contained in Reeves’ closing phrase could actually be realized. Yet it does seem to suggest that the idea of a past which is not realistically there and not subject to change or being changed clearly opens the way to abuse. I suppose that one can slip away from such consequences in the example just discussed by pointing '' Binghamton Press & Sun Bulletin, 29 October 1991, p. 5A. This, of course, is the local paper in which I read the column, but inasmuch as Reeves is nationally syndicated it may be found in all manner of other newspapers.

THE PAST OF OUR PRESENT 249 out that it does not deal with a disinterested past which may emerge from the course of historical inquiry, but is rather clearly grounded

in special interest, in commercial and financial interests, and that what he describes is nothing less than the blatant use of propaganda to obscure what is unpleasant and harmful. Even if we had access to the real past as it was when it was present, who is to say that there wouldn’t be interested parties attempting to hide its true nature in order to serve purposes of their own? What we are required to do 1s to eschew the realm of propaganda altogether, attend to the ways in

which scholars seek to learn about the human past, and on that basis discover whether or not a realistic sense of a past independent of our knowing it is a plausible object of knowledge. Martin Bernal is engaged in writing a multi-volume work which he calls Black Athena, the subtitle of which is “The Afroasiatic Roots of Classical Civilization.” As of this writing, two of its volumes have

been published, of which I have seen only the first, called “The Fabrication of Ancient Greece 1785-1985.”'* The Greeks, Bernal tells

us, had traditions to the effect that their culture was influenced to some determinate extent by the cultures of Egypt and Phoenicia, but in the course of the two centuries specified above European scholar-

ship came to reject that view of themselves that the Greeks had, coming to the conclusion that no such influences ever existed. On the face of it, this seems nothing unusual. There are all manner of opinions with respect to themselves that the Greeks entertained which

may be dismissed by us owing to the fact that our techniques for historical study are superior to their own. If, as Collingwood tells us, the only method available to Thucydides was the cross-questioning

of witnesses,’ he could not write the history of his grandfathers’ generation, but our classical historians can surely do that. What Bernal

tries to show, however, is that the new view concerning the influences on ancient Greek culture is not simply the result of the application of new and more effective methods of historical knowing, but,

rather, the tacit insinuation into the scholarly process of racist attitudes which could not accept that oriental peoples could influence the sort of culture that emerged in Greece. One might want to say '? Martin Bernal, Black Athena, vol. 1 (New Brunswick, NJ: Rutgers University oe a 1987); see my review in International Studies in Philosophy 21, no. 3 (1989), pp.

'’ Robin G. Collingwood, The Idea of History (Oxford: Oxford University Press, 1946), pp. 25ff

250 THE PAST OF OUR PRESENT that we must distinguish between the outcome of genuine historical scholarship and the insinuations of racism, but it certainly is not easy to do so in practice, and it is interesting to note that the discovery of such insinuation was made by a scholar who looked in from the outside, as it were, since prior to becoming interested in the ancient world, Bernal’s field of research was southeast Asia. One wonders if those trained to be classical historians so completely appropriated the paradigm that dominated their field they simply took it for granted without questioning it. If Bernal is nght, then the idea that we might hope to discover, in a disinterested way, an historical past which is not realistic yet is nevertheless intersubjective and open to all, would certainly seem problematic. And yet, the notion that there is an historical past knowable by us as it really was independent of our knowing it continues, at least for me, to be a paradox. Perhaps, unless we can find some way to justify Mandelbaum’s point of view, we really have no choice but to give up on history and the historical past altogether. That is not a course that I am prepared to advocate, but before I attempt to push my position to its conclusion, there is one more example I should like to consider. I cannot claim to speak with authority about the Historikersteit, and what little I know about comes from a recent reading of only three books.'* The WHistortkerstett—the historians’ debate—concerns how

German history is to be construed given the Third Reich and the Holocaust, and it seems to involve at least two aspects. One would appear to be straightforwardly disciplinary, namely, how the relevant swath of the German past is to be constituted. The other may seem

to be less so, yet is not entirely not so, and it concerns the emergence of a past which is compatible with the self-respect of the German

people all these decades after the war came to an end. And this seems to suggest that, quite apart from questions of scholarship, there

may be negotiated. :

is a social process by means of which a suitable or acceptable past

It is, of course, not the case that historical questions may be fi-

nally settled so that no new issues may arise or proposals for reinter'$ Charles S. Maier, The Unmasterable Past (Gambridge and London: Harvard University Press, 1988); Peter Baldwin, ed., Reworking the Past (Boston: Beacon Press, 1990); and Gina Thomas, ed., The Unresolved Past (London: Weidenfeld and Nicolson, 1990).

THE PAST OF OUR PRESENT 251 pretation be made. Think of how Lee Benson’s development of new techniques for the questioning of evidence led to his fundamental reconstrual of the character of Jacksonian Democracy in ways quite unlike what the historians’ traditions had come to accept,'? and the effect of the application of the cllometricians’ methodology on proposals with respect to conceptions of slavery in the antebellum South.'®

It does not follow that every innovation with respect to our sense of some swath of the human past is accepted and, thus, replaces what was generally accepted before.'’ However, the fact that such proposals can be made and taken seriously shows how open to reconstrual our notions of the nature of the human past, or some parts of it, are. Thus, on the face of things, it should occasion no surprise that something so unusual as the Third Reich, something as extraordinarily evil as the Third Reich, should be the subject of re-examination as historians—and others—try to make sense of it. ‘There are a number of issues involved in the historians’ debate, and I have nowhere near enough space to deal with them as they deserve even if I were qualified to do so. I want only to indicate enough about them so as to make that debate—perhaps a congeries of debates—relevant

to our present interests.'* It is not unreasonable that the idea that Germany experienced a Sonderweg, an eccentric course of history resulting in the Third Reich as an aberration, should he subject to reexamination, particularly inasmuch as a Sonderweg would seem to defy

being rendered intelligible. We need not even bother with so extreme a view as that associated with the name of Ernst Nolte,!? to the effect that the Holocaust was a consequence of Stalin’s murderous treatment of the Kulaks, and the claim of Andreas Hillgruber that the behavior towards Jews and other civilians by the retreating German forces on the Russian front was justified by the need to '" Lee Benson, The Concept of Jacksonian Democracy (New York: Atheneum, 1964; first published Princeton: Princeton University Press, 1961). '© Robert W. Fogel and Stanley L. Engerman, Time on the Cross, 2 vols. (Boston and Toronto: Little, Brown, 1974). '’ Thus the Fogel/Engerman thesis as to the nature of slavery before the Civil

War was subject to much critical scrutiny and opposition; see, e.g., Herbert G. Gutman, Slavery and the Numbers Game (Urbana, Chicago, London: University of Illinois Press, 1975).

'® I can only urge readers to dip into the literature themselves: the titles mentioned in note 14 and much else besides. ‘9 Nolte’s views are discussed in all three of the books mentioned in note 14, and interested readers may find references to his relevant writings in them.

252 THE PAST OF OUR PRESENT defend the eastern regions of the German homeland. German historians who are trying to deal with the Third Reich rarely feel called upon to justify it or rationalize its evil in any way. The problem of history as disciplined inquiry is to see what sense may be made of the relevant swath of German history. For example, one way to move attention away from the aberration of the Third Reich and attempt

to discover normal continuities is to attend to everyday life, as distinct from the realm of politics and ideology. Or, one may attempt to discover how the German case is just a special case of the western European response to the conditions of the modern world.

Either of these could be handled in terms of the methods of the discipline of history, and either could be understood as an attempt to discover what actually was going on in Germany during 19331945, as well as an attempt to understand it. But there are those— historians and others—who are motivated by other considerations entirely: what they seek is a German past with which they can live, a German past which does not clash with a decent sense of German being or identity. It 1s, of course, this last motive that creates a problem, and one can easily understand that historical realists, ike Mandelbaum, would seek to argue that only their realistic doctrine makes it possible to cope with it. T’o the extent that the first motive is strictly an epistemic

one, the realists could hope that in due course it would lead to knowledge of a past as it really was when it was present, and thus undermine the non-epistemic interests of the second motive. The only

way to prevent the abuse of the human past for whatever purposes or interests 1s to produce the real past in order to confront and defeat the abusers. However worthy such a goal, in the end it cannot be realized. Kant’s intuition with respect to our knowledge of nature and the

world around us is one that cannot be gainsaid. He was clearly mistaken in thinking that the categories in terms of which we constitute our world are unchanging and unhistorical, but clearly nght in

his view that the world for us is the world as we come to know it. The things-in-themselves are not accessible and cannot serve as the touchstone against which to test the truth or falsity of our claims to knowledge. ‘The same holds with respect to the historical past. The only past that we can talk about is the past as it is known to us. But this is not a counsel of despair. In spite of the seemingly unpromising epistemological conditions under which history is practiced, there

THE PAST OF OUR PRESENT 293 is a good deal of agreement among historians both as to what the human past was like and what the techniques and methods are in terms of which that past may be known. There are always unresolved problems, but we may hope—even expect—that in due course resolutions will appear. What cannot be overlooked is that historical inquiry now going on is our only source of knowledge of the human

past, that there is no transcending that inquiry in order to confront the real past, that the past for us is the past of our present.

THE SOCIOLOGICAL HISTORIOGRAPHY OF CHARLES TILLY I

A dialectical tension pervades B. G. Mogilnitsky’s essay, “Some Tendencies in the Development of Contemporary Bourgeois Historical Thought,”' in that he seems to be saying two different kinds of things

about bourgeois historiography at odds with each other. I do not believe that statements seemingly at odds are immediately self-canceling. On the contrary, statements that express an inherent tension

sometimes serve us better than those that do not conflict. Thus, social structures are both stable and changing at the same time; the Old ‘Testament covenant of God and Israel is both an agreement

between two parties and a gift from one party to the other; and societies are both individual and non-individual. What we need are

concepts that are able to encapsulate the inherent tension in each such pairing.’ But, on the other hand, not every tension need be admissible. And, I must confess, I have my doubts about the one pervasive in Mogilnitsky’s essay, namely, that bourgeois historiogra-

phy is both the causal result of certain social factors and is, at the same time, an ideological instrument of bourgeois class interests and the imperialism of the western states.

On the one hand, bourgeois scholarship is caught in the grip of social forces which shape the course of its growth and character. But, in Mogilnitsky’s view, bourgeois scholarship is not itself capable of the degree of self-reflection that would be necessary for it to dis-

cover for itself what its nature 1s. To accomplish such a thing, it ' Mogilnitsky’s essay originally was published in Voprosy istorta no. 2 (1987) and, in a translation by Sidney Monas, appears in Western and Russian Historiography, ed. Henry

Kozicki, intro. Sidney Monas (London, New York, 1992), 45-70. ? See Leon J. Goldstein, “Reflections on Conceptual Openness and Conceptual Tension,” in Freedom and Rationality: Essays in Honor of John Watkins, eds. Fred D’ Agostino

and Ian Jarvie (Dordrecht-Boston-London, 1989), 87-110. ° The problems presented by Rousseau’s “General Will” are entirely owing to the fact that, starting with an individualist methodology, Rousseau sought to articulate his insight into the fact that there is a non-individual sociocultural context within which the individual functions.

THE SOCIOLOGICAL HISTORIOGRAPHY OF CHARLES TILLY = 299

would be necessary for it to possess the character of science, and that is precisely what bourgeois scholarship cannot possess. Yet, on the other hand, time after time Mogilnitsky seems to be telling us that bourgeois historiography has a job to do. Its job, to be sure, is

not to reveal to us the historical truth about the human past. It 1s, rather, to further the cause of capitalism and western imperialism. It is hard to know how a mere epiphenomenon of social and economic forces can undertake a task with a specifically determinate goal. I suppose that, from the vantage point of Mogilnitsky, the nefarious schemings of bourgeois historiography is no real cause for concern. Why be fearful of a foe who cannot do anything right? Time

after time, he tells us about this or that bourgeois pretension. In particular, there is no science about the scholarship he impugns. It is always pretension. Bourgeois historiography has a pretension to work scientifically or to introduce scientific materials. It 1s prone to crisis. It has placed its unity and autonomy at risk. It 1s, clearly, in a very bad way. But it occurs to me that perhaps I am not reading Mogilnitsky in the way he means to be read. We have, after all, learned from Leo Strauss‘ and his disciples that we must sometimes distinguish between what a writer appears to be saying—aindeed, wants to appear to be

saying—and what he is really saying to those who know how to read. When Spinoza tells us that God creates out of the necessity of his being,” he presumably hopes that the dolts of the world—among

whom he expected that the censors were to be included—would be assured by the idea of a god who creates, and not notice that, whatever that means, it is withdrawn by the idea of the necessity of his nature. Presumably, those who know how to read will grasp the point easily.

While reading Mogilnitsky’s essay, it seemed to me that it could be read as a caricature of Marxism, when, in due course, the point of Straussian reading occurred to me. Where, after all, do we find such pretension among historians to be working in scientific ways— indeed, in the only scientific way it is given to an histonan to work—

but in the Marxist camp? Is it possible that really to understand * See, e.g., Leo Strauss, Persecution and the Art of Wnting (Glencoe, IL, 1952). > Baruch Spinoza, Ethics, Prop. XVII, Corollaries 1 and 2; cf. Prop. XXXII and

Prop. XXXIII note 2. I wish to thank Dr. Amihud Gilead for furnishing these

references to me.

256 THE SOCIOLOGICAL HISTORIOGRAPHY OF CHARLES TILLY

what Mogilnitsky is doing is to recognize it as a piece of self-criticism? Perhaps. We will simply have to wait until some Straussian subjects the paper to the sort of exegesis that Strauss himself provided for Xenophon’s Hiero*—and his disciple, Vaughan, for Vico.’ But there need no cause for fear or alarm. Even if Mogilnitsky’s secret 1s out, we live in a time of glasnost and perestrouka and nothing

untoward can happen.

Il Mogilnitsky’s paper seems to focus on two points: science and historiography. Perhaps it is only one point: scientific historiography. In any event, it seems to have a great deal to say about bourgeois historiography, and it is incredible how large a number of western historlans are cited in a paper of modest size. But even cursory atten-

tion to what the paper actually contains makes it clear that it says nothing about bourgeois historiography. ‘There are many quotations from historians, but it is historians talking about history, not actually

doing it. Carl Becker spent decades of his life producing works of historical scholarship, yet, for whatever reason or reasons, he became skeptical with respect to the possibility of historical knowledge.®

Surely, an account of Becker’s historiography, the principles which inform it, the methods which produce it, cannot be based on some

skeptical assertions. Rather, one must attend to the details of the work itself. We find no such attending in Mogilnitsky’s essay. What

we find instead are all manner of quotations about history; these give expression to their authors’ reflections about history and its character, but that is not the same as actually producing works of history. It is common enough that writers—even really good ones— are not always dependable in presenting accounts of what they do. In any event, historiography is what historians write as historians, not what they say they do or what they say about history. In order to do what he wants to do, Mogilnitsky should have presented detailed accounts of actual works of history. Of course, to deal with all ® Leo Strauss, On Tyranny, rev. edn. (Ithaca, 1968). ’ Frederick Vaughan, The Political Philosophy of Gianbattista Vico (The Hague, 1972). 8 See Leon J. Goldstein, “Historical Realism: The Ground of Carl Becker’s Skepticism,” Philosophy of the Social Sciences 2 (1972): 121-31. (In this volume pp. 106-118.)

THE SOCIOLOGICAL HISTORIOGRAPHY OF CHARLES TILLY 29/7

the historians he cites in his paper would require a tome of massive proportions, yet if he wants to condemn western historiography, he has to do it in terms of what it actually is. And that he fails to do. What about science? And scientific historiography? The Marxists have no doubt that they know what it 1s, and Mogilnitsky has no

doubt that the bourgeois historians he cites do not. This may be unfair, but Mogilnitsky leaves me with the impression that he takes science and scientific method to be rather a settled thing. In those ancient days when I was a student, we seemed to have had a similar attitude. Science is, to be sure, an ongoing and changing enterprise, but what science is—its generic nature, as it were—seemed reasonably stable. Do you want to know what it is like? Simply read the right authors: Carnap, Feigl, Hempel, Popper. Given what science is, if we want history to be scientific we need only figure out how to apply the canons of science to history.” When archeology began to become self-conscious about its identity—wanting to eschew, as it happens, its connection with history and insist that it is a social science—it discovered the philosophy of science of logical empiricism, particularly in the shape it received in the writings of C. G. Hempel.’

The archeology case may reflect a bit of cultural lag, since by the time the “new archeology” came on the scene our general certainty as to what science is was undergoing remarkable change. Did we really think that a scientific historiography required the application to history of the philosophy of science of logical empiricism, particularly its conception of explanation by subsumption under laws? Don’t give it a moment’s thought, we were told by Michael Scriven,

because that model of explanation is not ever to be found in the practice of natural science which is alleged to be its source.'' And then Paul Feyerabend tried to show that there are no specific methods or ways of thinking that were characteristic of science. Anything

that seemed to work, that would push the project forward toward

discernable progress, could be used and would be used in any ” See Hempel’s now-classic paper, “The Function of General Laws in History,” journal of Philosophy 39.1 (1942): 35-48. '" See P. J. Watson, S. J. Leblanc, and C. L. Redman, Explanation in Archaeology: An Expheitly Sctentific Approach (New York, 1971); and M. Alison Wylie’s doctoral dissertation, Positiwism and the New Archaeology, State University of New York at Binghamton, 1981. '' Michael Scriven, “Truisms as the Grounds for Historical Explanations,” in Patrick Gardiner, ed., Theories of History (Glencoe, IL, 1959), 443-75.

258 | THE SOCIOLOGICAL HISTORIOGRAPHY OF CHARLES TILLY

endeavor which purported to be scientific.'? Furthermore, there was no specifically determinate project or program which was science. Rather, the very nature of the enterprise was subject to radical shifts in the course of time, as conceptions of nature and the way in which nature ought to be studied underwent change. ‘There was no one science possessed of an essentially identical character, but in the course

of the history of systematic human study of the world in which we live, different ideals of natural order,’ reflecting different concep-

tions of the way things are and ought to be studied, led to the emergence of a succession of not fully compatible paradigms'* or research programmes’ within which different kinds of science would

be practiced. I’m sure that it comes as no surprise to people like Mogilnitsky that the philosophy of science of logical empiricism is no

longer the methodology it was around mid-century, but there is a lesson for all schools in the manner of its passing. And that is that there is no way to settle for all time what science 1s.

Il I suspect that it is easier to say what history—the disciphne—is, than what science is. Some insist that history is a form of literature and others that it is a means whereby we may correct the mistakes which

inform our social bias. There are those who think that properlydone history is the means whereby we may predict the future, while others are satisfied to believe that it is a way of uplifting the human spirit. ‘Che list could be enlarged. What these views have in common is that, by and large, they focus on the uses to which history is to be put rather than on what history, the discipline, is.’ Surely, the purpose of history is to make known to us the character '2 Paul K. Feyerabend, Against Method (London-Atlantic Highlands, NY, 1975). '3 Stephen K. Toulmin, Foresight and Understanding (New York, 1963). '* Thomas S. Kuhn, The Structure of Scientific Revolutions (Chicago, 1962). '> Imre Lakatos, “History of Science and Its Rational Reconstruction,” in Method and Appraisal in the Physwal Sciences, ed. Colin Howson (Gambridge-London-New York-

Melbourne, 1976), 1-39; see also his “Falsification and the Methodology of Scientific Research Programmes,” in Criticism and the Growth of Knowledge, ed. 1. Lakatos

and A. Musgrave (Gambndge, 1970). '6 Before historians write up their narratives—the only form of historical expression the history-as-literature theorists take to be the legitimate expressions of histoncal knowledge—they must engage in all manner of historical research and historical thought which, to my way of thinking, is what the discipline of history actually is.

THE SOCIOLOGICAL HISTORIOGRAPHY OF CHARLES TILLY 259

of the human past. We may, of course, use that knowledge in whatever way we deem pertinent to our interests—as reformers, revolutionists, or just individuals concerned about the human spirit. ‘here is, thus, no ambiguity about the status of history. Some who think that history ought to be science—whatever that is—become ambiguous about the status of history vis-a-vis sociology and other social sciences. It may be that sometimes history provides facts that social science may put to use,’’ and it is hard to doubt that the social and behavioral sciences contribute to historians’ ideas of the humanly possible.'® But far from suffering from any ambiguity with respect to its status or identity, history is the only discipline that has the task of making known to us the character of the human past. Charles Tilly has identified himself as “a sociologist who often works

with historical materials,”'? but one must recognize that he is also an historian. He may use techniques that are not typically used by his-

torians, but who is to say that future historians may not broaden their repertory of methods precisely owing to his successes. It is, of course, the case that Tilly does not produce narratives, and there are those who take it that history is a special form of story telling; history, W. B. Gallie tells us, is a species of the genus story.”? On that view, Tilly is not to be counted as an historian.?! Gallie—and Hayden White, Louis Mink, F. R. Ankersmit,*? and others—take it

that history is a mode of discourse, whereas for me it is a way of knowing, which is the only way that makes any sense to speak of historical truth and falsity, historical objectivity, factuality, and reference. The manner in which a historian presents his conclusions can-

not be essential to the nature of the discipline. Thus, on the narrational view, the discipline of history is undertaken for the purpose of wniting narratives. '7 Cf. what Guy E. Swanson does with the historical facts of the Reformation in his Religion and Regime (Ann Arbor, 1967). '8 See Leon J. Goldstein, “Toward a Logic of Historical Constitution,” in Epzstemology, Methodology and the Social Sciences, ed. R. S. Cohen and M. W. Wartofsky, Boston Studies in the Philosophy of Science, vol. 71 (Dordrecht-Boston-London, 1983), 19-52.

(In this volume pp. 177-208.) ‘9 The Formation of National States in Western Europe, ed. Charles Tilly (Princeton, 1975), 83. 20 W. B. Gallie, Philosophy and the Historical Understudy (London, 1964), 66.

*! | have had my say on that issue years ago. See Leon J. Goldstein Historical Knowing (Austin, TX, London, 1976), Ch. 5. 2 Hayden V. White, Metahistory: The Historical Imagination in Nineteenth-Century Europe

(Baltimore, 1973); Louis O. Mink, Historical Understanding, ed. B. Foy et al. (Ithaca, NY-London, 1987); F. R. Ankersmit, Narratwe Logic (The Hague-Boston-London, 1983).

260 THE SOCIOLOGICAL HISTORIOGRAPHY OF CHARLES TILLY

The use of sociological thinking, insight, and theories does not affect the character of the work as historical, if the purpose of it is to realize typical historical goals, namely accounts of what the human past was like. The emergence of the social and behavioral science affect our sense of what is and is not possible in the sphere of human activity.** Science in history is simply the using of the best conceptual tools available to make known to us what happened during

some span of the human past. It would, thus, seem to me that, notwithstanding the longstanding tendency to oppose the analytic and the narrative in historical writing as science vs art, to the extent that, in producing her finely-crafted narratives, Cecil Blanche Woodham-

Smith made use of the best available methods for acquiring the knowledge she presents, there is no reason to deny that her work is scientific.“ She makes no appeal to specifically formulated theories and uses none of the mathematicizing techniques that characterize the work of the cliometricians, yet there is method and system to what she does with the evidence before she produces her narratives. IV

But surely this is not how we have come to think about science in history. Ever since Hempel published his classic paper “The Function of General Laws in History,” the advocacy of science in history has been in the hands of those who defend the covering-law point of view: the view that explanation in history—as in everything else— must be by deduction from one or more relevant general laws and antecedent or boundary conditions. There is, however, no reason to identify science with explanation,” and even if one inclines to be more open as to what counts as scientific than allowed by the philosophy of science of logical empiricism, it should be possible to see the point of the nomological-deductive theorists as to the nature of explanation. What can you mean, they seem to be asking, if you say

3 See my “Toward a Logic of Historical Constitution.” 4 As in, The Great Hunger: Ireland 1845-1849 (New York, 1962), or The Reason Why (New York, 1953).

*° Nor, for that matter, with prediction. Those who think of explanation in the nomological-deductive way, identify the logical form of explanation and prediction, and there is a tendency—exaggerated in my opinion—to think of prediction as the goal, or main goal, of science.

THE SOCIOLOGICAL HISTORIOGRAPHY OF CHARLES TILLY 261

that this A explains that B, yet it need not be the case that Bs are always explained by As? I have myself been inclined to be sympathetic toward that view.*° Whatever one may wish to say about the nomological-deductive point of view, one thing 1s clear: it has apphcation only to a retrospective justification of what has already been accomplished. The historian—to limit ourselves to that—has produced an account of something, and the account seems reasonable.

How does one justify it? One way would be to produce the sort of general theory according to what the historian has done in his particular case is rendered intelligible. But this tells nothing about how the historian’s account was produced. It comes into play only after that has been completed, and I suppose that may make it seem like the correct way to go to those who believe that philosophy can only be concerned about the context of justification, never about the context of discovery. But the obvious consequence of this attitude 1s that we simply overlook entirely the actual character of historical knowing, paying no attention at all to what the historian actually does as he tries to determine what actually happened—meaning what it is reasonable to believe happened—in a human past we can never witness. Surely, there is method to that. Surely, there is theory—or common sense informed by theory—in such determination, particu-

larly inasmuch as it is not, contrary to what many people think, simply copied out of surviving documents. I should want to say that,

at its best, the work is scientific, but surely not accomplished by deduction from general laws.?’

Before turning to the task of discussing what 1s scientific in the sociological historiography of Charles ‘Tilly, I should like to strengthen

the position that I have been trying to take here by attending to what Arthur Stinchcombe does in his account of the use of theoreti-

cal methods in the practice of social history.” The problem of his book is to explore the way in which the practice of social history is informed by theoretical commitments which historians bring to their work. The heart of it is the middle two of its four chapters. In each of them, the work of two writers 1s compared. Chapter 2 deals with Trotsky’s History of the Russian Revolution and ‘Tocqueville’s The Old *° See Leon J. Goldstein, “Theory in History,” Philosophy of Sctence 34.1 (1967): 23-40; also in Boston Studies in the Philosophy of Science, vol. 4, ed. R. S. Cohen and

M. W. Wartofsky (Dordrecht, 1969), 277-302. (In this volume pp. 33-57.) “7 See my “Toward a Logic of Historical Constitution,” 24-28. (In this volume pp. 177-182.) *® Arthur L. Stinchcombe, Theoretical Methods in Social History (New York, 1978).

262 THE SOCIOLOGICAL HISTORIOGRAPHY OF CHARLES TILLY

Regime and the French Revolution, works of history by any reasonable

criteria.” Chapter 3 is concerned with contributions to history by two writers who are actually sociologists, Neil Smelser and Reinhard Bendix,°*° but, to the extent that they contribute to our knowledge of what the human past was like, they perform as historians. What Stinchcombe supports is the notion that there is science in the effort to come to know what the human past was like. It is true that he often speaks of explanation and, thus, creates the impression that the focus of his attention is in the use of theory to explain past events, however they come to be known. Thus it would appear that Stinchcombe’s book is an attempt to justify the covering-law point of view, albeit with a knowledge and sophistication with respect to sociology and social-science theory that philosophical defenders of the covering-law position never have. This is not, in fact, the case,”

but the interesting thing about the book is the way it points to the role of social-scientific theoretical models in the acquisition of historical knowledge. With writers like ‘Trotsky and ‘Tocqueville, one would suspect that

their broad visions concerning the character and direction of the sociopolitical world—what Stinchcombe calls their “epochal” theories—would determine how they characterize the historical material with which they deal, but he claims that this isn’t the case. Of course, their epochal visions are to be found in their histories, but the way in which they handle the detail of the course of events is, he argues, independent of them. Rather, embedded in their description are more limited theories which bear on the specific sorts of thing they discuss. And, in Stinchcombe’s view, the two writers are actually in broad agreement with respect to the character of the material and the strategies appropriate to the study of the material.’ The point is as follows. ‘The two historians produce narrative accounts of the material of their respective studies, but the character of the presentation and 2 Leon Trotsky, History of the Russian Revolution, trans. Max Eastman (New York, 1932); Alexis de Tocqueville, The Old Regime and the French Revolution, trans. S. Gilbert

(Garden City, NY, 1955). To be sure, such elements of Trotsky’s history which are in defense of his own role would not be genuine history. 30 Neil J. Smelser, Soctzal Change in the Industrial Revolution (Chicago, 1959); Reinhard Bendix, Work and Authority in Industry (New York, 1956). 3! See my review of Stinchcombe’s book in Foumal of Interdisciplinary History 10.3 (1980): 517-19.

* E.g., see Stinchcombe (49) for the way the two historians treat what he calls “Authority and Dual Power.”

THE SOCIOLOGICAL HISTORIOGRAPHY OF CHARLES TILLY 263

the authors’ respective senses of what the possibilities and actualities

are, are informed in essential ways by the theories they take to be true and relevant. In that way, then, science—not understood as deduction from general laws, but, rather, as the use of the best available conceptual tools for the production of the historical account— is seen to be a characteristic of their work. Stinchcombe explicitly identifies Smelser and Bendix as sociologists, but the focus of discussion is on how they handled what are certainly historical issues. Of course, as trained sociologists their theo-

retical material is not merely implicit. Rather, they are aware of what sociological notions figure in what they do, and one may suppose that their appreciation of the sociological side of historical research is more sophisticated than is more usually the case. But that has nothing to do with the principle at issue. Sociological instruments may be more finely honed in their hands, and they may be more explicitly aware than other historians of what current sociological thinking accepts as genuine possibility. Smelser’s study is concerned with the behavior of workers during the course of the industrial revo-

lution. Over time, one discovers certain kinds of change in that behavior. What actually is happening? Stinchcombe says that “the most original argument of the book is that ‘collective behavior’ of cotton workers in the first part of the nineteenth century was caused by the necessity to reorganize family roles, brought on by the growth

of factory labor separate from the family, and ended when that reorganization had been successfully carried through” (83). That factory behavior and concerns for the family—particularly, as it happens, the children—are elements of the same course of development need

not be all that obvious, and for Smelser to put them together the way he does is presumably owing to the way in which theory—his scientific instruments—leads him to construe them. There is no sociological-deductive derivation of the course described by Smelser, nor of any of the sub-events which make it up, but there is a method, a science, to the way in which his historical account is produced. V Sociological historiography is different than historical sociology. I have

not found in Tilly’s writings any historical sociology, though it does

seem to me that he makes important contributions to sociological

264 THE SOCIOLOGICAL HISTORIOGRAPHY OF CHARLES TILLY

historiography. Let me say briefly what I take the difference in these ideas to be. The job of history is to provide accounts of the human past, to tell us how things were and why. For that reason, historians’

work seems directed at particular historical entities or events, the French Revolution and not revolutions as such, Caesar crossing the Rubicon and not the general problem of upstarts challenging established authority. An historian may use whatever deepens his understanding of his problem and clarifies for him the lhmit of human possibility, but, gua historian, his attention is not on these things: he merely uses them for some other purpose. Sociologists, on the other hand, are oriented toward general theory. ‘They may, to be sure, make use of historical data that, in their opinions, contribute to the testing of the theories they work on, but it is the theory, not the historical fact, that 1s at the center of their attention. Thus, Lewis

Henry Morgan was not interested, as such, in the reforms that Cleisthenes introduced in ancient Attica, but, rather, dealt with those

reforms only as illustrating the theory he had worked out on the evolution of culture.** And certain facts of the religious history of Norway were of interest to John Flint only as they bore on theories in the sociology of religion, and, in order to use them in a way he deemed appropriate, he felt 1t necessary to redescribe the accounts of historians in the language of sociology.** This was not because sociologists are happy only when they have us all drowning in the jargon of their discipline, but because, unless he could present the material in terms of the variables of the relevant theories, there would

be no way to use that material in the way that he wanted. I do not want to exaggerate the distinction, though I think it fundamental. ‘That an historian uses theories from the social sciences in ways that are productive for his own purposes, must surely be understood

as providing some additional reason for taking the theories seriously: additional confirmation, as it were. And to the extent that some piece of historical constitution proves useful to a sociologist in testing or in otherwise enhancing his theoretical work, one may suppose that the

historical work is given some additional support. Thus, there is a mutuality here, not rigid boundaries which cannot be crossed over. Yet it is clearly possible to distinguish between historical sociology, 33 Lewis Henry Morgan, Ancient Society (New York, 1877). # John Flint, “The Secularization of Norwegian Society,” Comparatiwe Studies in Society and History 6.3 (1964): 325-44.

THE SOCIOLOGICAL HISTORIOGRAPHY OF CHARLES TILLY 265

which is a theoretical approach to the social by means of historical data, and sociological historiography, which is the attempt to realize goals of the discipline of history by utilizing the resources and methods of sociology.

The work of Charles Tilly is, clearly, of the latter sort. If it is to prove possible to offer an account of the character and direction of western—bourgeois, if you like—historiography, we are required not, pace Mogilnitsky, to cite general remarks historians may make about

history, but, rather, to attend to examples of the actual practice of history by large numbers of historians. And it seems appropriate to offer an example here. Since so much is made of “science”—by Mogilnitsky and others—it is not unreasonable to examine the historiography of one whose work is clearly informed by the methods of social science.

The Vendeé was illy’s first major work.® After the 1789 French Revolution, the counterrevolution called the Vendeé took place in

the 1790s in the western part of the country rather than in some other parts. It would seem that, by the application of the theories of sociology, Charles Tilly has offered an explanation for it, almost as if he is on the road to applying sociological covering laws to suitably

chosen boundary conditions in order to deduce by means of the techniques of logic an event called “the Vendeé.” All I can say is that this is not how the book reads to me. Even if such an explanation could be carried out, what would be deduced? One doesn’t simply deduce the Vendeé, a global event°®

called by that name. What would be deduced is an historical description of the Vendeé, and given what that 1s, it 1s clear that such

a description would require a very large number of statements in terms of which to be expressed. Indeed, it would require a booklength collection of statements, which is one reason to be doubtful about the realizability of the explanation-by-covering-law project. ‘Tilly’s

book seems not a response to the question, How shall we explain the Vendeé? but, rather, to the question—What is the Vendeé? I should suppose that the accounts we find in different works of history are not different explanations or interpretations of the global event which is the Vendeé, but, rather, different accounts or descriptions of what °° The Vendeé (Cambridge, London, 1964). °° See Henri-Iréneé Marrou, The Meaning of History, trans. R. J. Olsen (BaltimoreDublin, 1966), 314.

266 THE SOCIOLOGICAL HISTORIOGRAPHY OF CHARLES TILLY

the Vendeé consists of.*’ Tilly’s book has a great deal to tell us about

what the Vendeé was like, who participated, who opposed, and so on. What is untypical of historical writing is that, to a major degree, his account is teased out of the evidence by means of the application of sociological theory and technique. In Tilly, we are not confronting a counter-revolutionary outburst as something complete in itself. Rather, the outburst is part and parcel of a socio-economic and religious complex which is expressed in the

outburst. ‘he Vendeé is the totality of it, and in order to have a sense of what it is, we need to know more than that in 1793 there was an armed uprising in the west of France against the Revolution; we need to have a sense of the whole complex of factors which is the Vendeé. The spirit of this research is contained in the following passage from Tilly’s book: It is all too easy to stumble into the assumption that the Vendeé counter-

revolution was simply the natural response to the Revolution of a backward region. This convenient formula explains nothing. It fails because it begs three crucial questions: What do we mean by “backward”? What is the “natural” political behavior of backward regions? Was the Vendeé significantly more backward than sections of France which did not rebel? An examination of urbanization in western France provides some of the elements of answers to these questions. (16)

In order, then, to know what precisely the Vendeé was, we need to focus not merely on a “response,” but, rather, on the total character of the sections of France which rose in revolt. ‘Tilly begins by attempting to determine what urbanization is, and that concept 1s to be applied to the various sections of France. Urbanization 1s, of course,

a notion that is bequeathed to us by sociology, and so we find that sociological methods are used here. By thinking sociologically about the data uncovered from the historical remains, Tilly is able to get some idea of the differential urbanization in the sections of France in the late-eighteenth century. 47 See Leon J. Goldstein, “A Note on Historical Interpretation,” Philosophy of Science, 43.3 (1975): 312-19 (in this volume pp. 95-101), where I try to argue in a similar way about what are alleged to be different “interpretations” of Bacon’s Rebellion. There I show that, far from being different interpretations of the same event, the different account of Bacon’s Rebellion are not compossible in the same historical world inasmuch as each of the historians dealt with describes the rebellion in different ways, such that some of what it is alleged to consist in, according to one historian, is logically incompatible with what is alleged to have happened according to another historian.

THE SOCIOLOGICAL HISTORIOGRAPHY OF CHARLES TILLY 267

Likewise relevant to the outbreak of the counter-revolution are differences with respect to classes in the various parts of the country,

and what we find Tilly doing in Chapter 4 is going through the preserved documentation in order to derive from it relevant information. But this can only be done if he brings to his reading of the documentation the theoretical methods—again, derived from sociology—that enable him to discover the class structure and relations in the communities he studies. That information is there only if one knows how to read it or, more likely, constitute it out of the documentation. Consider the following analogy: I look through a microscope at a culture but see only a glob, but a biologist would discern a bacillus of a determinate strain because a conceptual apparatus enables him or her to see it. But the identification of the bacillus 1s not an explanation of anything. It is, rather, part of a description of what is given, that may, perhaps, be in need of explanation. ‘That 1s what Tilly does in The Vendeé. He is engaged in the historian’s task of constituting or constructing some part of the historical past, and, since what he is dealing with are phenomena of the social realm, the conceptual tools of sociology are relevant and useful to the realization of his purpose. Succeeding chapters of The Vendeé deal with the religious aspect (Ch. 6), economics (Chs. 7 and 10), power (Ch. 8), each carried out through the application of sociological thinking. We get description—

not narrative with heroes and plots—that simply could not be produced without scientific method. A dozen years later, Tilly contributed to and edited a multi-authored book, The Formation of National States in Western Europe, which attempted

to identify the factors that contributed to the rise of the national state. he different contributors focused on different elements: Samuel

E. Finer deals with the role of the military, Gabriel Ardant with financial policy, David H. Bayley with the police, and so on. In addition to a general introduction and a concluding essay, ‘Tilly contributed an essay dealing with food supply and public order. The Vendeé breaks out in 1793 and goes on for a few years. The time perspective of this second book is much longer. ‘The process of national-state formation takes centuries. It begins, obviously, before any states emerge, and while some states—Spain, France, England— emerge relatively early, others—Italy and Germany—do not become fully formed until the nineteenth century. But I do not want to suggest that the analytical work which informs Tilly’s constitution of the

268 THE SOCIOLOGICAL HISTORIOGRAPHY OF CHARLES TILLY

Vendeé is limited only to the small number of years of the counterrevolution. Given Tilly’s view that a proper account of the counterrevolution requires a presentation of the character of the social world

in the different parts of France at the time of the outbreak, he is required to produce tables and charts which show that character over

a period of time prior to the outbreak. The outbreak seems to be not simply a radical deviation from what existed until 1793, but, rather, an expression of that social, political, economic, and religious

reality in reaction to the events of the Revolution. But for all that, the project of the first book is more narrowly circumscribed so far as

its temporal spread is concerned than that of the second book. It is not possible to exaggerate the importance of food and not difficult to understand that the capacity to control the food supply is a source of power. Thus, it is not unreasonable to take seriously that, part of the effort to establish national states and the centralization of political power involves the attempt to control the food supply. Not only is such control itself a source of power, but the growth of the central state created new demands for food where they didn’t exist before—or existed to lesser extent—because it was necessary to feed growing armed forces and to channel food to national capitals where increasing numbers of people who were no longer involved in food production were concentrated to serve the growing state. This

defines the problem that Tilly explores in his contribution to the book on national-state formation. The growing need for food could be fine—especially for farmers—if means exist to increase its supply. That isn’t always easy, and, to the contrary, from time to time, for whatever causes or reasons, food is found to be in short supply. Government may have the power

to take what it wants, but ordinary people do not find the need of the state a sufficient reason to suffer hunger passively. And so we have the phenomenon of the hunger riot. If one thinks of history as simply a sequence of happenings, one might note that, in this place at that time, the masses gave violent bent to their need for food, but that is not the sort of thing we find in Tilly’s essay. Instead, we get a sociologically-constituted account of the nature—Tilly calls it here a “physiognomy” (385)—of food riots. And this is followed by a construal of the “political significance of food” (392ff) which deals with the way in which food and the controversies over it contributed to the rise of the national state. All in all, ‘Tilly’s essay is not a narrative account. It 1s, rather, an attempt to deal with an aspect of a

THE SOCIOLOGICAL HISTORIOGRAPHY OF CHARLES TILLY 269

long-term historical course by means of applying to the surviving manuscripts and documents conceptions of human social possibility and plausibility as those have been worked out within one of the systematic disciplines concerned with such things.

The last of the three works of Tilly that I consider here is The Contentious French.*® This, too, is a work of long temporal spread, namely, four-hundred years, and its focus is upon the character of popular struggle or contention in France during the course of that

period of time. Tilly’s claim is that the elements of contention in France are taken from other aspects of French culture and display considerable persistence (140ff.). The kinds of thing which give rise to contention change over the course of generations as different problems and different interests come to the center of attention, but contention itself is something specifically determinate that has both a history and a tendency to be stable in its expression. Tilly—here and elsewhere—uses the methods of sociology to constitute or reconstruct a swath of the historical past. VI

I use Tilly’s material in order to point to certain conclusions. The first, to go back to the beginning, is that there is no way we can have a sense of the historiography of the west—or the east—simply by noting what historians say about history. There are no such general remarks in ‘Tilly. Besides, no such quotations could begin to give the reader a sense of the rich character and multiple texture of the work that Tilly has produced as an historian. The second has to do with what history as scientific can mean. In “Toward a Logic of Historical Constitution,” I ask, “What makes an historical reconstruction acceptable?” (19); (in this volume 171), and

discuss what conceptions historians bring with them to the task of characterizing the human past. In earlier times, before the rise of the modern social and behavioral sciences, the historian operated in terms of the sophisticated common sense of educated members of his society, but, increasingly, that common sense is now informed by the findings of the social and behavioral sciences, even if the historian is

not particularly trained in any of them. In that paper, I noted that °8 The Contentious French (Cambridge-London, 1986).

270 THE SOCIOLOGICAL HISTORIOGRAPHY OF CHARLES TILLY

these define the limits of human possibility, what can or cannot be accepted (42-45; in this volume 199-203), what may or may not be presumed to be possible given the evidence that exists. In some cases, however, we see historiography move beyond the sophisticated common sense of historians toward the direct application of social science to the historian’s task. What is that task? There is an old, somewhat wooden conception of the relation of history and social science which sees that relation as entirely external to the practice of either one. History provides facts about the human past. Social science produces theories about human possibility and how social systems work. The historian’s facts may be useful for the testing of the social scientists’ theories. ‘The social scientists’ theories may

be useful in explaining the historian’s facts. Certainly, many of us think, if there is to be room for social science in history it can only be in the area of explanation, a term, I fear which is rather promiscuously overused in writings on history (in which respect it resembles “interpretation”). The historian’s task is to recreate the human past,

and what we are seeing increasingly in our time is the use of social science in the realization of that task. In their famous Time on the Cross,*° it does not seem to me that Fogel and Engerman use the cliometrician’s perspective in order to explain pre-established facts. Rather, they use it in order to establish facts not hitherto known, which could not be known until the evidence was subjected to their kind of questioning.* That is the sort of thing we find in Tilly’s work. ‘(he methods and insights of sociology are used to give us new ideas as to what the historical past was hke.

3° Robert W. Fogel and Stanley L. Engerman, Time on the Cross, 2 vols. (BostonToronto, 1974). * See my discussion of Lee Benson’s revision of the idea of Jackson’s democracy (in his The Concept of facksonian Democracy [New York, 1984]) in my Aistonical Knowing,

82-91.

PART THREE

COLLINGWOOD

BLANK PAGE

COLLINGWOOD’S THEORY OF HISTORICAL KNOWING I

In his review of The Idea of History, Maurice Mandelbaum asserts that

“Collingwood’s thesis that historical knowledge consists in the historian’s re-enactment in his own mind of the thought which underlay past actions must, I believe, inevitably lead to scepticism.”! Such a belief is not simply an independent judgment made about Collingwood’s view, but rather rests upon a conception of the nature of history? markedly unlike the one to which Collingwood himself subscribed. In fact, CGollingwood’s conclusion that the historian re-enacts in his own mind the thought that lay behind the historically important actions he is concerned to deal with is precisely the result of reflection designed to overcome historical skepticism. Here, indeed, 1s a paradox. ‘The identical assertion is taken by one as irredeemably skeptical and by another as precisely the way to avoid that undesirable outcome. In any event, it seems to me that Collingwood makes an attempt to understand the discipline of history so as to take it seriously as a method for acquiring knowledge of some sort. In one sense, the point of departure for this effort is the recognition that the discipline which is history has some considerable success to its credit, and that no account of it which makes that success unintelligible or rules it out as merely apparent can be taken seriously. Since Collingwood’s own early views on the subject were unambiguously skeptical, he is clearly rejecting those views and seeking grounds for that rejection.’

' Maurice Mandelbaum, review of R. G. Collingwood, The Idea of History, in Fournal of Philosophy 44 (1947), 187.

* When I talk about history or the nature of history in this paper, I shall be dealing with the discipline of history and not with the course of human events. * Thus, writers like Errol Harris (“Collingwood’s Theory of History,” Philosophical Quarterly 7 [1957], 35 -49) and H. A. Hodges (The Philosophy of Wilhelm Dilthey [Lon-

don, 1952], ch. 10), who think that Collingwood’s views remained essentially unchanged over the years are simply mistaken. Harris attempts to offer a systematic account which encompasses all of Collingwood’s writings on the subject of history and observes that Collingwood never “explicitly abandoned” the doctrine of Speculum Mentis (49). Hodges, too, observes that Speculum Mentis had not been “explicitly

274 COLLINGWOOD’S THEORY OF HISTORICAL KNOWING

If, as we shall see, the method for avoiding historical skepticism characteristic of Collingwood’s mature thought involves a rejection of realism—a theory of knowledge which he says “is based upon the srandest foundation a philosophy can have, namely human stupidity” (Meta., 34)*—it must not be thought that the earlier skeptical position to which he subscribed itself rests upon any sort of realistic foundation. One may suspect that, being dissatisfied with what Speculum Mentis has to say about history, he proceeded to a serious exami-

nation of how that discipline carries out its work, and in the course of that examination came to see how inadequate a realistic theory of knowledge was to any proper understanding of history. But his own initial skepticism was rooted in an altogether different sort of philosophical commitment. The skepticism of Speculum Mentis is rooted not in any conception

of the nature of history, but, rather, in a very restricting conception of what knowledge is. Consequently, we need deal with it only in enough detail to justify my claim that the skepticism is there. The subtitle Collingwood gave to the book is “The Map of Knowledge,” and it is clear enough from its contents that this is intended to suggest that the work attempts to account for the relationships that the

various components of human knowledge have to one another. Collingwood, in this earlier book, conceives of knowledge as divided into art, religion, science, history, and philosophy, and he discusses

them in that order. But the discussion is not merely seriatim. The relationships are taken to be dialectical, and in the course of his account of the essential character of each—except, of course, the last—he tries to show how it points toward the essential character of the next. For whatever reason, science 1s said to point to the need for a discipline concerned to establish particular facts, and this turns out to be fzstory. At this stage of his thought, history is something arrived at by way of a dialectical move of one sort or another; it is not the discipline real historians work at. Only later could he speak of writing down “the lessons of my last nine years’ work in historical research repudiated” (338), and uses it as the touchstone for the interpretation of Collingwood’s later work. But explicitly reyected or not, it is clear enough that the Speculum Mentis conception of historical knowledge is simply not to be found in Collingwood’s later writings, and I do think it unjustifiable to claim that the burden of those writings is precisely to overcome doctrines like that of the earlier book. * Page references given in the text preceded by Meta. will be to Collingwood’s An Essay on Metaphysics (Oxford, 1940).

COLLINGWOOD’S THEORY OF HISTORICAL KNOWING 2795

and reflection upon it” (Auto., 107),° and only later will what he has to say about the subject be important to students of the problems of historical knowing.

Thus, the Speculum Mentis discussion of “history” proceeds with almost no attention to the actual activities of historians, and reaches a culmination in section five of the relevant chapter, “The Breakdown of History,” which is a veritable subversion of the possibility of such a discipline. The opening statement of the section sounds the keynote for the discussion which follows: “History is the knowledge of the infinite world of facts. It is therefore itself an infinite whole of thought: history is essentially universal history, a whole in which the knowledge of every fact is included” (Spec., 231).° History is now being judged by an external standard—a far remove from Collingwood’s later insistence upon the radical autonomy of history—for whatever might be said on behalf of that standard, what is said about history

in the passage just quoted could never have been learned through examination of what historians actually do. In any event, the consequence of what is said about history is drawn almost immediately thereafter: “History is the knowledge of an infinite whole whose parts, repeating the plan of the whole in their structure, are only known by reference to their context. But since this context is always incomplete, we can never know a single part as it actually is” (Spec., 231). The function of history in the dialectical scheme of Speculum Mentis

is to provide knowledge of fact, or “objective reality independent of mind.” Such an object, supposed to be “individual, concrete, infinite, no arbitrary abstraction or unreal fiction, but reality itself in its completeness,” was sought in vain in the previously considered spheres of knowledge. “In history we have found it; and we have found it to

be an illusion.... The progressive alienation of the mind from its object is in history complete” (Spec., 238). It is interesting to see even at this early point in the development of Collingwood’s thought that historical skepticism is associated with a realistic ideal of an object of knowledge independent of and unaffected by the knowing mind. But

here the matter is complicated by the conception of such an object as being an infinite whole and the belief that the structure of each > Page references given in the text preceded by Auto. will be to Collingwood’s An Autobigraphy (Oxford, 1939).

° Page references given in the text preceded by Spec. will be to Collingwood’s Speculum Mentis (Oxford, 1924).

276 COLLINGWOOD’S THEORY OF HISTORICAL KNOWING

part reflects that of the whole. The claim that one knows just a part of the whole—given what the whole is here taken to be, no historian could ever claim more than that—must always be mistaken, for such a part is an abstraction, hence a distortion. A critical examination of these bases for historical skepticism would,

no doubt, be very interesting, but I shall confine myself to showing that so far as Speculum Mentis goes, “History as a form of knowledge cannot exist” (Spec., 238). ‘This is hardly a satisfactory conclusion; surely not for one who was himself an historian. Speculum Ments takes

the problem to be one of how we know an infinite whole: Every part implies the whole, and the whole is presupposed by every part. No part can therefore be known first. No process of thought with respect to such a whole is possible. We cannot come to know it. We must have known it all from the beginning, have known it as a whole before we began to learn any given part; and once the whole is known

any given part must be known too, and therefore cannot be learnt. (Spec., 239f.; italics added)

With this begins “The Transition from History to Philosophy,” as the next section is called—for the problem of knowledge must be solved even if history cannot contribute to the solution. But however the problem is taken in Speculum Mentis, no historian—not excepting Collingwood himself—could leave the matter there. History is, after all, a discipline with some recognized success to its credit, and only a most dogged commitment to the doctrines of Speculum Ment, or systems like it, could lead to denying the claim that there is historical knowledge. It seems to me that the burden of Collingwood’s reflections on history over the course of the years following the publication of Speculum Menizs is precisely the overcoming of the skepticism

to which that work leads. And the solution is to be found not in a dialectical move which leaves history behind, but, rather, in history itself.

II

The strange impact which The Idea of History makes upon readers, even philosophical readers, is in large measure owing to the fact that typically they are not sensitive to what Collingwood is actually doing. Errol Harris is nght to say that “Collingwood believes that it is impossible to determine whether or not a proposition is true without

COLLINGWOOD’S THEORY OF HISTORICAL KNOWING 277

knowing what question it is meant to answer,’ and it is ironic how frequently Collingwood has himself been the victim of critics or interpreters who have not bothered to determine what questions he was trying to answer. Thus, I think it appropriate to assert at once that there are two possible things that The Idea of History is not: it 1s not an attempt to impose an antecedent, say philosophical, conception of knowledge upon history, though anyone who sees the work as simply an expression of idealism or some such doctrine presum-

ably thinks it is;? nor is it an attempt to characterize the course of human events—of the past as past—though it must be admitted that there are passages which do that (cf. Idea, 119).° Virtually from the

time Speculum Mentis was published, and as long thereafter as Collingwood thought systematically about history, he was concerned to answer the question: what is the nature of the discipline of history

that it should be capable of yielding up historical truth? Or, how can the historian, forever caught in the present, speak with assurance about events which must always be beyond his experience? Being

beyond his experience is the case in principle, not simply a contingent circumstance owing to the absence of time machines and the like. Anyone who suggests that the full meaning of an historical proposition (qua historical) is what would be experienced by one who might

actually be present while the event is or was taking place, simply betrays the fact that he has not the slightest appreciation of the central

problem of historical knowing.'® ‘To my mind, the culmination of this long inquiry is to be found in three much-criticized sections of the Epilegomena to The Idea of History: “The Historical Imagination”;

’ Errol E. Harns, “Collingwood on Eternal Problems,” Philosophical Quarterly | 1951), 231. ® That Collingwood did not consider himself an idealist is clear from the follow-

ing: “... any one opposing the ‘realists’ was automatically classified as an ‘idealist,’ which meant a belated survivor of Green’s school. There was no ready-made class into which you could put a philosopher who, after a thorough training in ‘realism,’ had revolted against it and arrived at conclusions of his own quite unlike anything the school of Green had taught. So, in spite of occasional remonstrances, that was how I found myself classified...” (Auto., 56). 9 Page references given in the text preceded by Jdea will be to Collingwood’s The Idea of History (Oxford, 1946).

‘9 What I know through direct observation or the report of a witness, I do not know in the historical way; cf. Collingwood’s remarks on the Greek historians in Idea, 26f. Likewise, what I know about the past because the truth of its having happened is vouchsafed by a revelation I deem to be inerrant, I do not know in the historical way.

278 COLLINGWOOD’S THEORY OF HISTORICAL KNOWING

“Historical Evidence”; “History as the Re-enactment of Past Experi-

ence.” What I shall now try to do is show how some of their most striking doctrines are intelligible only in the light of Collingwood’s attempt to understand what history is.’ The crucial point of Collingwood’s conception is the recognition that the commonplace dual reference of the word “history” is neither a mere linguistic accident nor the occasion for possible punning. The word is used both in reference to what is supposed to have happened in the course of human existence and experience and to the written

accounts of these. Even with respect to this second sense there is room for distinctions to be made; at least we find both wider and narrower senses of “history” as referring to such accounts. Some distinguish between annals, a mere seriatim recording by someone living at the time of what is supposed to have happened during some period, and history taken to be narrative or explanatory, or in some

other possible way a cut above mere annals. On this view, those

masterpieces of Greek literature produced by Herodotus and Thucydides are certainly to be counted as history. Collingwood, however, prefers to restrict the usage of the term even more. The famous Greek historians are, to be sure, discussed in The Idea of History,

but they are not admitted as full historians whose works exemplify fully the “generic essence” of history. Collingwood discusses them as having contributed to the development of the idea of history—clearly, though not explicitly, understood as a “scale of forms’’!*—but once that idea is fully developed, they do not measure up well against it. Such a view of Herodotus, and especially of Thucydides, will surely

be unacceptable to any number of readers; it is not necessary to emphasize here the great esteem with which both have been and stl are regarded. But that Collingwood should have this attitude toward their work is quite important for us here, for it emphasizes '' | know that Collingwood sometimes took extreme positions, not all of which he actually required. And I know that some of the views he expressed would have to be rejected quite apart from the effect such rejection would have on other parts of his doctrine. But since my purpose in writing this paper would not be served by examining these, I do not intend to do so. Interested readers will find some discussion of such matters in Alan Donagan, The Later Philosophy of R. G. Collingwood (Ox-

ford, 1962), chs. 8 and 9. '2 A term Collingwood uses in An Essay on Philosophical Method (Oxford, 1933). Students of that work will recognize here the relevance of his view of the overlap of philosophical concepts and his conception of the scale of forms. See particularly

his remarks on the paradox of saying both that the lower forms of a genus are members of it yet exemplify its generic essence hardly at all (62).

COLLINGWOOD’S THEORY OF HISTORICAL KNOWING 279

the point of his inquiry. It is not that he is unimpressed with the narrative power or psychological insight of the Greek historians. But

he is struck by their limitations as historians: “Their method tied them on a tether whose length was the length of living memory: the only source they could criticize was an eyewitness with whom they

could converse face to face... as soon as Greek historical writing tries to go beyond its tether, it becomes a far weaker and more precarious thing” (/dea, 26); “the Greek historian’s method precludes him

from choosing his subject.... The only thing he can write about is the events which have happened within living memory to people with whom he can have personal contact” (/dea, 26f.); ““Thucydides’ work is a ktfpa é¢ aiet, that of Herodotus was written to rescue glorious deeds from the oblivion of time, precisely because when their generation was dead and gone the work could never be done again.

The rewriting of their histories, or their incorporation into the history of a longer period, would have seemed to them an absurdity” (Idea, 27). Some might wish to say that any criteria according to which Herodotus and Thucydides were not accorded full honors as historians, and which denied that their writings realized almost to the utmost the generic essence of history, are clearly defective and ought to be rejected out of hand. We shall not bother about that. What is to the point is the nature of Collingwood’s judgment. It is based upon recognition that certain limitations were inherent in the fundamental methodology of Greek historical scholarship. ‘There were some things it simply could not do—it is not simply an accident that their subjects were limited in the way they were. Nor were they able to make use of what has since become known as historical evidence.

In large measure, the purpose of the first four parts of The Idea of History is to present an account of the gradual overcoming of those limitations and the emergence in modern times of the true idea of history. '°

Some might suggest that the only way in which to answer the question, what is history? without begging the question is to survey the writings of historians, and to abstract from these those features which seem common to them. Collingwood’s procedure takes for granted that the discipline of history did not arise full blown early in '’ Cf. Arnaldo Momigliano, “The Place of Herodotus in the History of Historiography,” in his Studies in Historiography (Harper Torchbook ed.: New York and Evanston, 1966), 127-142.

280 COLLINGWOOD’S THEORY OF HISTORICAL KNOWING

its own history; he contends rather that the emergence of the fullest expression of its generic essence is the outcome of centuries of effort. But it would not have been possible to trace that development simply by means of careful reading of historical works. The very possibility of the first four parts of The Idea of History having the form they have rests upon Collingwood’s already having come to some definite conclusions about what the generic essence of history is. History turns out to be a discipline that can do certain things, ask certain questions, apply certain techniques. ‘The judgment on the Greeks is based upon Collingwood’s view that they are incapable of asking those questions

or applying those techniques.'* In all this, we have the judging of historical writing not in terms of some theory of knowledge, but, rather, in terms of what history has become.

Il I am trying in the present discussion to provide a context in terms of which certain views of The Idea of History which have perplexed Collingwood’s critics would be rendered intelligible—and even seem essentially correct. And while I have been speaking about his concentration upon the actual practice of historical investigation, I do

think it relevant to look back a bit upon a more specifically philosophical commitment. I refer, of course, to his utter rejection of realism

in epistemology. It may be that very early in his life he actually subscribed to realistic principles (Auto., 25),'° but he soon gave that

up. By his own account, this rejection is for the most part based upon the realization that realism is totally incompatible with what is done in historical research (Auto., 28). I believe that realism is indeed entirely inapplicable to history; but Collingwood’s own retrospective account of the course of his own rejection of that doctrine may perhaps be questioned. His Speculum Ments is by no means a realistic work,'® '* If Collingwood is right about the Greeks, their views would surely lead to some sort of historical skepticism. At most, all we could know is what their historians have recorded; we could not widen the scope of our knowledge nor would we have the means whereby to subject their own writings to critical test. As Collingwood himself would put it, their writings could only be treated as authorities, never as evidence for the exercise of historical imagination. Cf. his discussion of such matters in the Epilegomena to The Idea of History. ') This and other relevant passages of the Autobiography are discussed in Donagan, If.

'§ Indeed, it already contains a version of Collingwood’s anti-realistic view of

COLLINGWOOD’S THEORY OF HISTORICAL KNOWING 281

yet, as we have seen, its account of historical knowledge is clearly quite skeptical.'’ It may well be that the order of development is the reverse of that recounted in the Autobiography, and that his already discrediting realism made it easier for him to discern the characteristic features of historical knowing.

In any event, it would be useful to have some idea of what the doctrine is that Collingwood is rejecting. This is obviously not the place to present a detailed account of the arguments for and against the various kinds of realism in which the history of philosophy abounds.

Even an attempt to present and assess all that Collingwood himself says about realism would take a good deal of effort. His writings are full of all manner of attacks upon it, from comparatively well-reasoned

analyses to passing, merely polemical jibes. It will be enough here merely to state what he takes it to be; with that done it will be easy enough to discover even in his comparatively early reflections on history that it is his growing anti-realism as much as anything that puts him on the road to his mature theory of historical knowing. Or,

perhaps better, it is his anti-realism which opens his mind to an appreciation of what historical knowing is, once he has determined to devote himself to the actual study of it. For this end, some characteristic observations from the Autobiography will serve as well as any.

In the relevant passage, Collingwood talks about the views of the Oxford realists, to which he was exposed as a student, as being that knowing is the simple apprehension of a given reality. In his view, this was essentially the same doctrine held by realists elsewhere, whose

views in general he sums up as follows: knowledge as involving questions and answers (Spec., 76ff.), a conception discussed in such subsequent works as An Autobiography, Essay on Metaphysics, and The Idea of

story. '7 Yet without being realistic it does contain a realistic view of history. ‘That Collingwood’s earliest views of history were realistic is also evidenced by the following, which was published in 1916, eight years before the appearance of Speculum Mentis: “History must be regarded not as a mechanical process, nor yet as a gradual accumulation of truths, but simply as objectwity; as the real fact of which we are conscious. History is that which actually exists; fact, as something independent of my own or your knowledge of it” (Region and Philosophy [London, 1916], 49). (I have not seen this book, but the passage is quoted in William M. Johnston, The Formatwe Years of R. G. Collingwood [The Hague, 1967], 49; the italics are in Johnston’s text.) For all that, the actual argument for historical skepticism in Speculum Mentis is, as we have

seen, based on an ideal of knowledge. It is because history is an infinite world of fact which cannot be truly known ptecemeal that it cannot be known at all. That it is an

objective, real world independent of our minds does not seem to enter into the argument.

282 COLLINGWOOD’S THEORY OF HISTORICAL KNOWING

The condition of a knowing mind is not indeed a passive condition, for it is actively engaged in knowing; but a “simple” condition, one in which there are no complexities or diversities, nothing except just the knowing. They granted that a man who wanted to know something might have to work, in ways that might be very complicated, in order to “put himself in a position” from which it could be “apprehended”; but once the position had been attained there was nothing for him to do but “apprehend” it, or perhaps fail to “apprehend” it. (Auto., 25f.)

Knowledge, on the realistic view, turns out to be an immediate or intuitive grasping by a subject of an object present to its consciousness. It is this idea of immediate or direct apprehension, which he takes to be a characteristic mark of realistic theories of knowledge, that Collingwood thinks is too simplistic to be taken seriously. How

ironic that his critics, in their bafHement over the sense of his well-known aphorism “When he [the historian] knows what happened, he already knows why it happened” (dea, 214), should seek to saddle

such a view on him!'® The objects of historical study, at any rate, are never given to immediate apprehension. IV For all its unfinished character, The Idea of History is the mature state-

ment of Collingwood’s reflections on the nature and problems of historical knowing, as we shall soon see. But there is evidence of the direction his thought was taking in his earlier writings, now happily brought together from the scattered periodicals of their original pub-

lication within the covers of one volume.'? In a paper called “The Nature and Aims of a Philosophy of History,” which goes back to the period of Speculum Ments, but apparently does not reflect the holistic

conception of knowledge we found in that work, Collingwood remarks that “the battle of Hastings is a label for something which, no doubt, did happen in that year [1066]; but no one knows, no one '8 I do not say that if you understand Collingwood’s statement you will accept it; cf. Donagan, 201. But I do say that anyone who cites it in support of the claim that Collingwood believed that one could grasp an event and its explanation in one immediate intuition does not understand it. To understand it means to see it as Collingwoodian, to determine what someone holding Collingwood’s views must mean

by it, not to analyze it as an ordinary English sentence having no context. '9 R. G. Collingwood, Essays in the Philosophy of History, ed. with an introduction by William Debbins (Austin, Texas, 1965); page references to this volume will be given in the text preceded by Essays.

COLLINGWOOD’S THEORY OF HISTORICAL KNOWING 283

ever has known, and no one ever will know what exactly it was that happened” (Essays, 43). But instead of expressing the skepticism we might have expected from the author of Speculum Mentis, what 1mmediately follows is an explicit denial that what he said is skeptical; “for

scepticism implies that no opinion is preferable to any other; and it is certainly possible to choose between different historical views... without implying that one knows all there is to be known about it oneself.”

In Speculum Mentss, historical thought is a dialectically grounded moment in the course of an account of the nature of knowledge. In “The Nature and Aims of a Philosophy of History,” history and its philosophy are the objects of immediate interest. In the first, history is not attended to for its own sake, and, measured against a criterion of knowledge which finds it wanting, it is abandoned. In the second, however, history is too important and taken too seriously to be rejected merely because on some possible criterion—external to itself—skepticism seems to be called for. It is of some interest to see just how

this skepticism is rejected. Soon after the remarks quoted in the previous paragraph, Collingwood says: Ideally, historical thought is the apprehension of a world of fact. Ac-

tually it is the presentation of thought to itself of a world of half-ascertained fact: a world in which truth and error are at any given moment inextricably confused together. Thus the actual object of actual historical thinking 1s an object which is not “given” but perpetually in process of being given. (Essays, 44)

This distinction between the ideal and the real would not have been possible in Collingwood’s later writings, for as he became increasingly preoccupied with what the discipline of history is, he became increasingly unwilling to allow that its autonomy could be limited by standards outside of itself. Here, however, he seems still willing to allow that the common-sense ideal of history as an attempt to describe a past as it was, a reality external to the thought of the historian, is acceptable. But while a goal to be striven for, it is never achieved. ‘The object of investigation is never given fully and finally, but is, rather, “perpetually in process of being given.” This is what history is, and accepting history as a legitimate way of knowing, the Collingwood of “The Nature and Aims of a Philosophy of History” accepts it with all of its shortcomings. It is interesting to note that even here, with his explicit acceptance of the common-sense ideal of historical thought as an “apprehension of a world of fact”—an ideal

284 COLLINGWOOD’S THEORY OF HISTORICAL KNOWING

which is implicitly realistic since such a world of fact is presumably there, unaffected by the consciousness of the knower—he is already moving toward his autonomous and non-realistic conception of history as a discipline whose object is the “presentation by thought to itself of a world.””° Later in the essay we have been considering, Collingwood compares history and perception in such a way as to anticipate what he will say in The Idea of History about the historical imagination. [ shall not develop that comparison fully since his views on perception as such are not germane to what interests us here. For him, perception involves a large component of interpretation—something not at all surprising in a thinker who was increasingly rejecting realism on the one hand, and who so much emphasized the autonomy of history and the central importance of the historian’s imagination in the reconstruction of the past on the other. In any event, it does seem to me that here we see in an early form what is to become his strong emphasis on the historian’s imagination, but unlike the later doctrine, which is concerned with the reconstruction of past events, this earlier form is limited to the interpretation of evidence. It may seem odd to say about history that it “in its fundamental and elementary form is perception” or that it “is perception raised to its highest power” (Essays, 49). This is because “Perception appears *° Collingwood’s conception of the autonomy of history is very important for our understanding of his views and, in particular, is central to his attempt to overcome historical skepticism. Thus, it is worth making certain that we understand what it is and what it is not. As will emerge in the course of our discussion, Collingwood’s position is that the discipline of history is autonomous, in that it is not dependent on things and criteria outside of itself in order for its work to proceed. It is not in the

least what Patrick Gardiner thinks it is. His discussion of it is presented in the course of what is supposed to be an account of views to the effect that history 1s suz generis, a view which he grounds in belief that history deals with its own peculiar subject-matter which, he says, Collingwood treats as “a self-contained world that must accordingly be interpreted by methods bearing little or no relation to those used in other branches of knowledge.” Gardiner makes a number of mistakes here. Collingwood’s view is that the discipline of history is autonomous, not that some peculiar historical entities—the “world of experience’”——are autonomous with respect

to other entities. If Gardiner were correct in his reading of Collingwood, then the discipline of history would not be autonomous at all; instead, we would have the subordination of history to the “dictates” of something external, namely to the separate spheres of existence Gardiner says Collingwood believed in. Finally, far from claiming special methods of its own—’"esoteric methods” as Gardiner calls them— we shall see that for Collingwood the method of history is thinking or re-thinking, surely not a method confined to history alone. See Patrick Gardiner, The Nature of Historical Explanation (Oxford, 1952), 28-34, 49.

COLLINGWOOD’S THEORY OF HISTORICAL KNOWING 285

to the perceiver as immediate” (Essays, 49), and some philosophers

have simply adopted the standpoint of the naive perceiver. For Collingwood, however, reflection makes clear that in perception we have to do with two elements, sensation and thought, and that sensation is only an abstraction, “the limiting case in which we are supposed to receive unreflectively a pure datum. In actual experience

we never get such a pure datum: whatever we call a datum is in point of fact already interpreted by thought” (Essays, 50). In view of this, what at first sight may have seemed to be the strange compari-

son of two fundamentally disparate modes of mental experience markedly changes its character. “The only difference between what we ordinarily call perception and what we ordinarily call historical thinking is that the interpretative work which in the former is implicit and only revealed by reflective analysis is in the latter explicit and impossible to overlook” (Essays, 50). ‘The parallel between these two modes of experience is made even more explicit in the words which soon follow: “in all perception we are making a judgment, trying to answer the question what it is that we perceive, and all history is

simply a more intense and sustained attempt to answer the same question.” The rationale for this reflects his rejection of historical realism, his refusal to take seriously the general view that history is concerned with a real past, complete and simply waiting to be dis-

covered. “The past events which the historian brings to light are only revealed by his thought in its attempt to understand the world

present to his senses: a past event which has left no trace on his perceptible world is to him unknowable” (Essays, 50). In the account of the previous paragraph we see easily enough the direction Collingwood’s thought is taking in his attempt to overcome historical skepticism. ‘The point of the claim that historical thinking is like perception is that it is an affair of the present; that it 1s concerned with what is now given to the historian’s senses. ‘That which

is given to the historian’s senses is not a past event, but rather historical evidence.*’ We shall see in a moment just what Collingwood has to say in this comparatively early essay about historical evidence *! The problem of the past in the context of concern with historical knowing is how we who are always caught in the present, having no direct access to the past, can claim to have knowledge now of what happened then. The whole point of the discipline of history is to enable us to make warranted assertions about a past we can never know by acquaintance. It is failure to attend to this—and, more particularly, failure to recognize how aware of this Collingwood was—that surely lies behind

286 COLLINGWOOD’S THEORY OF HISTORICAL KNOWING

and its interpretation, but it is worth pointing out that the character the discussion takes makes it evident that Collingwood has not yet reached the culmination of his theory of historical knowing, with its central attention to the role of imagination, not in the interpretation of evidence, but in the reconstruction of events. I venture to suggest, and hope that it will become more evident in the sequel, that until Collingwood came to recognize the role that the historian’s imagination plays in the discipline of history, the historical event, with its seeming pointing to a forever unapproachable past-in-itself, may have been somewhat. embarrassing for him. Be that as it may, the discussion which follows, informed as it is by a nonrealistic conception of

perception, deals with the historian’s interpretation of what he qua historian perceives, namely historical evidence. Historical thinking, according to the essay we have been considering, 1s critical thinking about evidence: “We have not only to read, but to criticise” (Assays, 51). Perceptual judgment rests on past experience, “and we perceive more and more accurately according as we become more and more able to compare the present experience with relevant experience in the past” (Essays, 50f.). Similarly, historical thinking requires experience, initially in the form of relevant training

of the sort that will make it possible for the practitioner of the discipline to make critical assessments of historical evidence. Progress in

historical study involved the refusal of critical historians to believe something simply because it is written in some old document. Rather,

by dint of application of the historical disciplines which deal with evidence, historians have learned how to question sources, rejecting assertions which cannot be true and extracting from them all manner of undreamed-of truth. “No competent historian who reflects on the progress of his own thought can overlook the way in which progress has created* masses of evidence bearing on questions concerning which there was once no evidence whatever” (Essays, 52). Several years later,?? Collingwood could still say: Gardiner’s view that the problem for Collingwood was to have a past that was present. If the past could be known by acquaintance, then the discipline of history and its concern with historical evidence would be of no use to anyone. One may presume that the same tendency that leads Gardiner to foist upon Collingwood belief in peculiar historical entities (see the previous note) is at work here too. See ae At his point, Collingwood adds the following footnote: “Created, not discovered, because evidence is not evidence until it makes something evident” (his italics). 3 In a paper called “The Limits of Historical Knowledge,” published in 1928.

COLLINGWOOD’S THEORY OF HISTORICAL KNOWING 287

For historical thinking means nothing else than interpreting all the available evidence with the maximum degree of critical skill. It does not mean discovering what really happened, if “what really happened”

is anything other than “what the evidence indicates.” If there once happened an event concerning which no shred of evidence now survives, that event is not part of any historian’s universe; it is no historian’s

business to discover it; it is no gap in any historian’s knowledge that he does not know it. (Essays, 99)

From the standpoint of Speculum Mentis, this last passage is simply impossible,” for belief in the existence of past events concerning which historians have no knowledge is precisely what leads Collingwood to

historical skepticism in that book. We have seen that the problem his thinking about history is intended to resolve is precisely that of skepticism, and we have seen that the means whereby that resolution is accomplished is the attending to what historians do. And that means,

at least during a large part of the period between Speculum Mentis and The Idea of History, that history seems at least implicitly to be understood not as that discipline which concerns itself about the formulation of warrantable assertions about past events, but, rather, as the discipline whose subject matter is historical evidence—those data which exist now, in the present, which historians have to understand and interpret. ‘Thus, he can say: “The past is simply non-existent; and every historian feels this in his dealings with it” (Essays, 101). Most interesting is the way in which his view of the very basis of

historical skepticism has changed. He does say at one point that historical skepticism has a useful cognitive function—not to call into question the validity of historical knowledge, but rather to point to

its limits (Essays, 91)—but that is not the point I find so striking. What is of interest here is that historical skepticism is laid at the door of historical realism itself. “The discovery that the past as such is unknowable is the scepticism which is the permanent and necessary counterpart of the plain man’s realism” (Essays, 100). ‘The point here is that skepticism is the consequence of any conception of his-

tory which makes the focus of attention not what historians do in the present, but what really happened in the past, a conception which takes as equally factual, hence historical, the question what Caesar had for breakfast some given morning and the question whether he ** And underscores how mistaken are those who think that Collingwood’s views remained essentially unchanged and that his later wntings are to be understood in light of that earlier work.

288 COLLINGWOOD’S THEORY OF HISTORICAL KNOWING

intended to become king of Rome (Essays, 101). Curiously, Collingwood has not forgotten that earlier, Speculum Mentis basis for skepticism, the

view that knowledge is and can only be an indivisible whole. But now he saddles a modification of that view on the realists. Historical realists are taken to be those who measure success in historical work

in terms of the quantities of fact discovered. Their epistemology commits them to a view according to which knowledge is the grasp-

ing or apprehension of objects which exist independently of the knower. In history this is possible only if all of the past events of human experience still exist in some peculiar way,” and historical knowledge 1s the apprehension of them. The more such facts are known, the closer do we come to the ideal of knowing the whole (Essays, O9FF.).

The direction Collingwood’s historical thought was taking during the period from Speculum Mentis to The Idea of History is now clear.

Increasingly he came to think that the way to avoid the pitfalls of skepticism was to move away from an unknowable past-in-itself and pay rather close heed to what historians do. This, of course, rests on the assumption that historical skepticism is to be rejected on the ground

that what historians do and have accomplished justifies the claim that history is a legitimate way of knowing. Historical skepticism, it would seem, must always be consequent upon standpoints outside of history itself—say, an holistic conception of knowledge or a realistic _ view of knowing—but cannot arise from consideration of history it-

self, as it is practiced by historians.” I should not wish to suggest that in Collingwood’s opinion philosophical considerations are merely philosophical considerations, for I think that strewn through his writings are all manner of suggestions to the contrary. But it is nevertheless true that it was only after he began to think about the procedures

of history itself that he was able to see his way clear of skepticism, and it is likewise true that he came early to believe that one virtually fatal flaw of realistic epistemologies is their incapacity to do justice to history. > See Collingwood, “Some Perplexities About Time: With an Attempted Solution,” Proceedings of the Anstotelian Soctety 26 (1925-26), 144, 146, and 150.

6 The skepticism of Charles A. Beard stems not from his reflections on historical method, but from his commitment to goals, rooted in historical realism, with which those methods cannot be reconciled. Note his repeated insistence that the interest of history is in the “past as actuality” in his “Written History as an Act of Faith,” in The Philosophy of History in Our Time, ed. H. Meyerhoff (New York, 1959), 140-151.

COLLINGWOOD’S THEORY OF HISTORICAL KNOWING 289

V

I want now to turn to The Idea of History in order to present and discuss a number of its most important—some would say most outrageous— doctrines. As is well known, this book was published posthumously. Two-thirds of it, the first four of its five parts, are a critical historical sketch of the development of the concept of history—the

emergence through time, as it were, of the generic essence of the discipline of history. The final part, to which our attention here will be limited, consists of a number of separate essays. ‘The actual volume which is called The Idea of History was put together out of these parts by Sir Malcolm Knox, who explains what he did in the “Editor’s

Preface.” All this results in a work the parts of which come from different times of the author’s life—though none from what might be

called his early period—and which did not have the advantage of the author’s final editorial scrutiny. It is very easy to single out passages from The Idea of History for criticism on one basis or another. But for all of that, and for all its incompleteness, the general drift of the book is toward a conception of historical knowing which is essentially sound. Let us set the stage for presenting the main elements of that conception by presenting an extract from the book (/dea, 132f.): The conception of history as dealing with facts and nothing but facts may seem harmless enough, but what is a fact? According to the positivistic theory of knowledge, a fact is something immediately given in perception. When it is said that science consists first in ascertaining facts and then in discovering laws, the facts, here, are facts directly observed by the scientists.... If anyone doubts the fact he can repeat the experiment. ..; and consequently, for the scientist, the question whether the facts really are what they are said to be 1s never a vital question, because he can always reproduce the facts under his own eyes. In science, then, the facts are empirical facts, facts perceived as they occur.

In history, the word “fact” bears a very different meaning. The fact that in the second century the legions began to be recruited wholly outside Italy is not immediately given. It is arrived at inferentially by a process of interpreting data according to a complicated system of rules and assumptions. A theory of historical knowledge would discover what these rules and assumptions are, and would ask how far they are necessary and legitimate. All this was entirely neglected by the positivistic historians, who thus never asked themselves the difficult question: How is historical knowledge possible? How and under what conditions can the historian know facts which, being now gone beyond recall or

290 COLLINGWOOD’S THEORY OF HISTORICAL KNOWING

repetition, cannot be for him objects of perception? ‘They were precluded from asking this question by their false analogy between scientific facts

and historical facts. Owing to this false analogy, they thought such a question could need no answer. But, owing to the same false analogy, they were all the time misconceiving the nature of historical facts, and consequently distorting the actual work of historical research...

The distinction between the facts of history and those of natural science

certainly helps make clear the problem, even though I rather doubt that the distinction as presented here is one that Collingwood really wished generally to make. For the quoted passage would make it appear that Collingwood was willing to concede that realism was an adequate epistemology for the natural sciences, yet that is surely far from the case. Collingwood wants to say, here, that positivist historians assimilate historical facts to natural facts and so never allow themselves to raise the question as to how the former can be known. His realistic critics might urge against him that his own tendency is to do the reverse, to assimilate natural fact to historical fact and, consequently, lose the independent existence of the real world. Yet even if what Collingwood says here is not fully consistent with what he says about realism elsewhere, his way of putting it here is perfectly serviceable. Make the strongest possible case for realistic theo-

ries of knowledge, and accept everything realists tell us about our knowledge of the world around us and the starting points of natural scientific inquiry; yet when it comes to the facts of history—which in

the passage quoted clearly means historical events of the past, not historical evidence—realism has nothing to say. The assimilation of historical facts to natural facts amounts to asserting that the inaccessibility of past events is merely accidental or fortuitous. We can look at what is happening around us, but not at what happened long ago. Some writers seem to think that the difference zs merely fortuitous,

and the full meaning of a description of a past event is precisely what would appear to an observer, or ideal observer, witnessing the event.*’ But a God’s-eye view of the course of human events would not be an historical view of it, nor would an account certified as true by God, i.e., a work of divine revelation, be an historical account. An historical account is one which is the outcome of the application

of the methods and techniques of historical inquiry, and these do 27 Cf. Richard M. Gale, “Dewey and the Problem of the Alleged Futurity of Yesterday,” Philosophy and Phenomenologuwal Research 22 (1961-62), 501-511, and my “The ‘Alleged’ Futurity of Yesterday,” zbed., 24 (1963-64), 417-420. (In this volume

pp. 102-105.)

COLLINGWOOD’S THEORY OF HISTORICAL KNOWING 291

not include the reports of ideal observers. The reports of such observers and the reconstructions of historians have clearly different epistemic statuses. By “the facts of history” Collingwood would understand facts established by historical research, not the actions to which human beings have been party over the course of their existence on earth. From which it clearly follows that historical facts cannot

be facts in the realistic or positivistic sense.” This is a very important issue, and no small part of the critical reception of Collingwood’s philosophy of history may be traced to a tendency on the part of many not to attend to the epistemic differences

between present and past descriptions. What I mean may become clearer if we consider briefly the views of a writer who, far from agreeing with Collingwood, is quite critical of him: A. M. Maclver. In Maclver’s view, “We contrast the ‘generality’ of historical statements with the individuality of the facts on which they are based”” and he goes on to show what he means in the following passage: A typically historical statement is “The Normans defeated the English at Hastings in 1066.” The battle itself was a vast medley of individual actions and experiences—this man shooting this arrow, that man avoiding it or being hit by it, horses stumbling, men feeling pain or fear or exultation—but the historical statement takes it as a whole and selects

for mention just that aspect of it which bears upon the historian’s purpose—in this case, the fact that as a result the Duke of Normandy was able to make himself King of England. It is not the business of the historian to “generalize” in any other sense than this.

As an account of what historians do, the above is patently false, as I shall make clear shortly, but let us at least indicate what about it is right, what would lead most readers to think it sound. And that is that anyone who sought to answer such questions as “what is a battle?”

or “what is an army?” would have to speak about the sorts of thing

8 Hence, pace Gardiner, Collingwood could not possibly subscribe to the view that historical knowing was knowing by acquaintance (Gardiner, 39). Indeed, how anyone as opposed to realism in any sphere of knowledge as Collingwood was could be thought to believe that the acquaintance theory had any application at all is beyond my comprehension. His following remarks are pertinent here: “All theories of knowledge that conceive it as a transaction or relation between a subject and an object both actually existing, and confronting or compresent to one another, theories that take acquaintance as the essence of knowledge, make history impossible” (Idea, 233).

*? A. M. Maclver, “The Character of a Historical Explanation,” Avistotelian Soctety Supplementary Volume XXI (1947), 38. °° Thid., 39.

292 COLLINGWOOD’S THEORY OF HISTORICAL KNOWING

Maclver does. In a battle soldiers behave as they are wont to do, and it would be utterly absurd to think that you can have an army without soldiers or a battle without shooting and being shot—with whatever projectiles could be managed at various stages of military technology.

But an historian’s statement with respect to the Battle of Hastings

is not generalized in the way Maclver says; it is not based on the individual facts to which he adverts. Let us be clear about the fact that we are no longer talking about the nature of armies and battles in general, about the “ontology” of armies and battles. We are now concerned with the nature of historical knowing, with the epistemic status of the historian’s statement. On the basis of evidence, the his-

torian concludes that a battle took place, that the Normans were victorious, and that Wiliam succeeded Harold as England’s king. Far from basing his assertion that a battle took place on the sort of individual facts MaclIver mentions, the historian does not even know

what they were. He does not know anything about who shot what arrow at whom and what resulted therefrom. Except for the major participants, he does not even know who was there, what he did and how he emerged. ‘To be sure, being in full possession of his senses and knowing in general what is involved in warfare of the sort in question, he does believe that individual soldiers performed individual

acts of soldiery. What could he mean by the assertion that a battle took place if not that? Yet his knowledge that a battle took place is not generalized from these acts because he does not know what these acts were. Epzstemically, the basis for the claim that a battle was fought

was the evidence—the precise relation between evidence and the knowledge of historical events need not detain us here.°! If the Battle of Hastings were directly accessible to the historian’s

observation, then perhaps it would be a technique of historians to generalize in the way Maclver says, and there would be no difference in status between past and present descriptions. But then our present discipline of history as a form of knowing would be entirely superfluous. It is the absence of such access to the past that necessitates the historical form of knowing if we are to be able to speak of the past at all. Collingwood was fully aware of this, and it is that awareness which lies back of the views we are now to consider. 3} But cf. my “Evidence and Events in History,” Philosophy of Science 29 (1962),

175-179. (In this volume pp. 3-27.)

COLLINGWOOD’S THEORY OF HISTORICAL KNOWING 293

VI

It is now possible to understand the point of the main doctrines of The Idea of History, and it is possible to understand the way in which they emerge from Collingwood’s reflection both on what historians do and how history is to be justified as a discipline capable of yielding warranted results. This will become clear as we examine the main ideas in the sections of the last part of The Idea of History which were mentioned earlier in our discussion. We shall begin with the notions of the historical imagination and the autonomy of history which are presented in the section called “The Historical Imagination.”

To say of any discipline that purports to be about something or other that it is autonomous, that it is not dependent upon things outside of itself, seems highly dubious at first blush. Surely, we want to reply, no science is free to do just what it wants; rather, it is clearly

dependent upon the nature of that which it studies.” It is evident enough by now how Collingwood would receive such a remark. Since

the objects of history “are events which have finished happening, and conditions no longer in existence” (/dea, 233), it is simply not possible to claim that something now in existence, namely, the enterprise of history, depends upon it. But there is another way in which Collingwood’s critics might oppose the claim that history 1s independ-

ent, and that is by appeal to what Collingwood calls “the commonsense theory” of history, which he presents as follows (/dea, 234f.): According to this theory, the essential. things in history are memory and authority. If an event or state of things is to be historically known, first of all some one must be acquainted with it; then he must remember it; then he must state his recollection of it in terms intelligible to another; and finally that other must accept the statement as true. History

is thus the believing some one else when he says that he remembers something. The believer is the historian; the person believed is called his authority.

I dare say this theory could be expressed in other ways, with this point qualified, that modified, and some other one changed in some 2 However, should we characterize sciences, not in terms of the objects they study, but in terms of their conceptual systems, we could argue that they determine, conceptually, the nature of their own subject matters. But there is nothing to be gained here by pushing this point, and it is perfectly satisfactory to allow the char-

acter of the subject matter of natural science and that of history to continue to appear to be basically different. Cf. my “Ontological Social Science,” Amencan Anthropologist 61 (1959), 290-298.

294 COLLINGWOOD’S THEORY OF HISTORICAL KNOWING

other way. The central point of the theory, which would presumably

survive all modification, is that the historian may not gainsay the actual claims of his authorities.** But this is totally false to the practice of historians. There is nothing contained in any of his documents that the historian cannot doubt, or even reject. And this, of course, means that whatever criteria the historian uses in terms of which to reject or accept the assertions of a document, those criteria are independent of the document itself. In addition, the use historians make of evidence which does not make assertions—ruins, coins, artifacts— makes clear that they have ways of determining what happened that have nothing at all to do with believing the claims of authorities. All of these ways and criteria are part of the present methods of historical research, and that the historian may apply them to his evidence, may subject his evidence to their control, is the measure of the autonomy of history. In Collingwood’s actual discussion of these points (/dea, 236), there is a tendency to ignore an important distinction between two senses of mind-dependence. I have just spoken of the autonomy of history, but he writes of the historian as being his own authority and of hus thought as being “autonomous, self-authorizing, possessed of a criterion to which his so-called authorities must conform and by refer-

ence to which they are criticized.” In the paragraph immediately following, he compares the historian to an artist: “It is the artist, and not nature, that is responsible for what goes into the picture. In the

same way, no historian, not even the worst, merely copies out his authorities. ... It is he, therefore, and not his authority, that is responsible for what goes in” (/dea, 236f.). Soon after this, he talks about the historian’s “a pron imagination” (dea, 240ff), saying that much

of what he takes to be true he constructs by means of his a priori imagination, and then adding: “I am now driven to confess that there are for historical thought no fixed points thus given: in other words, that in history, just as there are properly speaking no authorities, so there are properly speaking no data” (Idea, 243).** Surely, then, those critics who see in The Idea of History a doctrine of skepticism rooted in the author’s inability to get beyond the subjectivity of each historian seem vindicated. * Cf. Gale, 509. ** For an earlier expression of subjectivism, see his 1930 paper, “The Philosophy of History” (Essays, 137ff).

COLLINGWOOD’S THEORY OF HISTORICAL KNOWING 295

In my opinion, this conclusion of the critics is mistaken. My purpose here, however, is to point to the nature of Collingwood’s mistake, for I do not think that the line of thought about history that he developed requires the subjectivism he sometimes seems to espouse.”

The source of the error is failure to keep distinct two senses of mind-dependence, the idiosyncratic and the non-idiosyncratic. Mind-

dependence in the former sense refers to dependence on the peculiarities of an individual, rather like the way in which the blurred look of the world when I am not wearing my glasses depends upon my myopia or someone’s conception of the course of some historical event upon /zs narrow-minded prejudice. In the latter sense, it 1s like the way in which a good deal of human action depends upon shared meanings, values, and institutional arrangements for its intelligibility. By speaking of it as mind-dependent we mean that it is not merely

or directly a creation of nature, but we do not mean to suggest that what we deal with depends upon the idiosyncrasies of particular individuals. It is a regrettable fact that not infrequently these two senses are confused. Perhaps it is because when we think of mind we think

of an individual’s mind, and we are often wary lest we be charged with hypostatizing group minds and the like. And it has often proved difficult in the English-speaking philosophical community to think of

thought or meaning without seeking at once to reduce the whole matter to the thought of some thinker or the meanings of some actor. But it is certainly possible to distinguish the content of thought and the thinking of thought, and any number of writers have held the view that the very possibility of such disciplines as social and intellectual history, cultural anthropology, and sociology depends upon the feasibility of this distinction.* °° Cf. “Everyone brings his own mind to the study of history, and approaches it from the point of view which is characteristic of himself and his generation. ... ‘The attempt to eliminate this ‘subjective element’ from history is always insincere... and always unsuccessful. If it succeeded, history itself would vanish.” This leads Collingwood to what is surely a most vicious form of subjectivism: “This does not reduce history to something arbitrary or capricious. It remains genuine knowledge. How can this be, if my thoughts about Julius Caesar differ from Mommsen’s? Must not one of us be wrong? No, because the object differs. My historical thought is about my own past, not about Mommsen’s past” (Essays, 138f.). *© Dilthey was certainly one of those who held to the distinction, but the reader of the collection of excerpts from his wnitings on history, Pattern and Meaning in History,

transl. and ed. H. P. Rickman (New York, 1961), may sometimes have difficulty in keeping this in mind owing to Rickman’s decision (23) to translate Dilthey’s “das Geistige” as “mental content.” This often makes it seem that Dilthey’s focus was on the idiosyncratic-psychological when actually he was concerned with social-historical

296 COLLINGWOOD’S THEORY OF HISTORICAL KNOWING

What Collingwood ought to say is that history is mind-dependent in the non-idiosyncratic sense, as can be seen from a summary of the argument so far. ‘The task Collingwood set himself is the vindication of history against any forms of skepticism which would call it into question, and this he does by attending to what history is. He does not ask, “is history possible?” but, rather, “what is history that it is possible?”*’ The answer which emerges in the course of his work is that it is a discipline having certain procedures and methods, which makes use of what is called historical evidence, which is autonomous in that it is not dependent on criteria other than those it establishes, and so forth. In all these ways it is mind-dependent in the non-idiosyncratic sense. But hke anything mind-dependent in that sense, it

has no existence apart from individuals: it would be ludicrous to think of the discipline of history without historians. Yet the discipline

is not dependent upon the peculiarities—much less the biases—of any of its practitioners, and the mistake Collingwood makes in the passage we have been considering is the failure to keep distinct the two senses of mind-dependence and the failure to keep from slipping from the discipline and its characteristics to its practitioners and theirs. But in view of the way I have been supporting Collingwood’s rejection

of historical realism, virtually daring his critics to produce a real fact-in-itself to be the subject matter of historical research, some may wonder if I have a nght to appeal to the distinction of the two senses in order to keep non-realism from falling into subjectivism. Briefly, such

wondering might proceed in the following manner. If the real past is, indeed, not available to be examined or studied, what, in fact, is? Clearly, historians—individual scholars—engaged in historical research.

Each historian is a subject embodying a standpoint in terms of which his work is carried out. Each does his own investigating, makes his own judgments, accepting or rejecting according to whatever criteria

he chooses to accept. Having rejected historical realism, the critic claims, Collingwood has at the same time made it impossible for us to have a non-subjective history; each historian, his imagination as unrestrained as that of a novelist, may simply do as he pleases.

content; cf. 69, 71, 72, and 76. I cite this not to be critical of Rickman, but only as an example of how easy it 1s to slip from the non-idiosyncratic kind of mind-depen-

dence to the idiosyncratic kind. *” Dilthey, too (zbid.), takes the possibility of history for granted, but his approach to answering this second question is, in the Kantian manner, to look for the categories which are constitutive of the historical world.

COLLINGWOOD’S THEORY OF HISTORICAL KNOWING 297

But while we have historians—individual subjects, to be sure——we

also have history. We have an organized structure of historical research. We have widely-shared techniques and generally-agreed-to results. We have, in sum, an inter-subjective discipline, and while it is carried on by individuals, they are committed to a common enterprise having determinate features. The claim, then, that history is autono-

mous need only be made with reference to that discipline. History 1s

autonomous in a way historians may well not be. It is, then, the autonomous discipline of history which controls and has the final say

about historical evidence, and that discipline is not subject to the authority of old texts. With this much said, we may now turn to the next section of The Idea of History, “Historical Evidence.”

VU

The relation of history—the discipline—to historical evidence is one of the pervasive themes of The Idea of History, and perhaps more than anything else it is the changing character of that relation which marks

the different stages of the evolution of the idea of history. I noted earlier that in Collingwood’s view the concept of history is not sub-

ject to precise definition but is rather a scale of forms the generic essence of which is increasingly realized in the course of the history

of history. At first, there is virtually complete dependence of the historian upon his sources, and, it will be recalled that Collingwood says of the Greek historians that they were capable of using only one kind of source, the reports of witnesses. Even when historians could

use written reports of witnesses no longer living, hence no longer available for questioning, at first, Collingwood claims, they could either

incorporate the content of the report or reject it, but it took some time before they arrived at techniques for dealing with it critically. As it matures—-and more perfectly expresses the generic essence of its concept—history learns to subject its sources to 1ts own purposes,

to extract from a text all manner of information not explicitly asserted in it, to deny its sources all claims to authority. History, itself, as it becomes increasingly autonomous, becomes the sole authority:

far from relying on the reports of others, it is history itself which decides what is and what is not historical evidence and what possible use this or that historical evidence may have. According to a widely-held, yet clearly naive, view, the historian discovers his facts by reading sources: what he may assert with warrant

298 COLLINGWOOD’S THEORY OF HISTORICAL KNOWING

is limited by the testimonies they preserve. In Collingwood’s view, such testimony cannot be historical knowledge: What I assert is that it can never be historical knowledge, because it can never be scientific knowledge. It is not scientific knowledge beCause it cannot be vindicated by appeal to the grounds on which it 1s based. As soon as there are such grounds, the case is no longer one of testimony. When testimony is reinforced by evidence, our acceptance of it is no longer the acceptance of testimony as such; it is the affirmation of something based upon evidence, that is, historical knowledge. (Idea, 257)

Let us see what this really means. First of all, we must be careful not to be bewildered by the statement just quoted through a possible confusion of evidence and testimony. For many, the testimony found in our documents and other written sources is just what we mean by the best possible sort of historical evidence, though we recognize that there are other kinds of evidence too. But in Collingwood’s view the mature and autonomous science of history itself determines what is or is not evidence. That some assertion is transmitted over the course of time and exists preserved in writing in some archive or library 1s of no interest at all to historians, in Collingwood’s view, until such time as it proves relevant to answering questions historians seek to answer. ‘The testimony preserved in the documents qua testimony, 1s

not, then, historical evidence. Its becoming historical evidence 1s entirely dependent upon the historian and the use he has made of it. The point of the quoted statement should now be clear. To be

scientific—in the older, broader sense of the term preferred by Collingwood, rather than in the narrower sense more commonly used these days—historical propositions must rest upon evidence, grounds for the admission of which are rooted in the character of the inquiry itself. Assertions which rest upon testimony depend upon the merely fortuitous or accidental, and that is hardly a firm basis for scientific knowledge.

At first blush, we seem to face a paradox. It is no secret that history has had its credentials as “objective” challenged much more than most disciplines, and if it is to have any hope at all of holding up its head in the company of epistemically respectable disciplines

that will surely depend on its ability to base itself upon evidence which is in some sense given and does not depend upon the vagaries—the subjective vagaries—of historians and their questions. But if Collingwood is right, this support is no longer there. The documents

COLLINGWOOD’S THEORY OF HISTORICAL KNOWING 299

and testimonies which have been transmitted from bygone days, far from being the touchstone in terms of which to determine historical truth, now appear themselves to depend for their status on the workings of an autonomous discipline of history. Yet a moment’s reflection makes it clear that the paradox is merely apparent, that it rests upon the plausibility of some sort of common-sense realism and cannot survive its rejection. It would not be appropriate here to open this question to discussion; our concern, after all, is to make clear what Collingwood’s views are without detailed exploration of all the problems which would have to be cleared up before one could claim to have a worked-out theory of historical knowing. In any event, the anti-realistic line relevant to removal of the seeming paradox is to urge that knowledge and knowledge claims do not rest simply upon

a direct interaction of knower and the known, but depend upon acceptance of criteria in terms of which knowledge is delineated by those jointly engaged in some knowing enterprise—the small number of specialists when that enterprise is the disciplined pursuit of scientific or theoretical knowledge, the total community or the largest part of it when what is involved is the everyday knowledge of common sense and social existence. It is only in terms of such a conception that the historian’s enterprise 1s possible or intelligible, since the realistic interaction of knower and known is clearly not possible in this sphere. It turns out, then, that what history in fact does is precisely what Collingwood says a discipline must do if it 1s to be worthy of the name of knowledge: it does not simply accept what has been fortuitously preserved. Rather, its claims to knowledge are vindicated by appeal to grounds. And that such-and-such are indeed the grounds for this-or-that claim to historical knowledge is determined in terms of the criteria that the discipline uses. History is a responsible discipline, and the historian is not at liberty to make up the facts as he needs them. Nor is a work of historical narrative so self-complete and independent of everything else as

to require no reference to anything beyond the needs of its own inherent logic.* If nothing could serve as historical evidence, then clearly there would be no possible discipline of history, and it is worth

remembering that the main emphasis in the first two-thirds of The Idea of History, the part concerned with the development of the idea °° For an opposite view, see W. B. Gallie, Philosophy and The Historical Understanding

(London, 1964), chs. 1-5.

300 COLLINGWOOD’S THEORY OF HISTORICAL KNOWING

of history or the coming to maturity of its generic essence, is the varied and changing ways in which historians deal with historical evidence. But, in the end, it is the discipline which is the master; it is not subservient to what has simply managed to survive. Nor is this particularly strange. In terms of what, after all, are we to decide that something is evidence for something else, or that of two somewhat divergent accounts of what purports to be the same occurrence one

is better than the other? It is clearly not in terms of what really happened when the past was present, since that is not available for the historian’s inspection.*” And this leaves only the discipline itself, the principles and criteria in terms of which it assesses the cogency and acceptability of putative historical evidence. With this in mind, we can now turn to the last of the three sections of The Idea of History we are considering, “History as Re-enactment of Past Experience.”

Vill The section begins with the question, “How, or on what conditions,

can the historian know the past?” (/dea, 282). The direction of Collingwood’s thought makes it clear that the main condition is that there be something to engage the historian now. The actual events

which transpired when the past was present are clearly no longer available. And we have already seen why he cannot allow that it is sufficient for the historian to find testimonies or other accounts in his sources, which accounts he may simply—acceptingly—copy in his

own narrative. What, then, is left for the historian to do?

It is here that we get Collingwood’s view that history is the re-enactment of past thought. Thus: “The difference between memory

and history is that whereas in memory the past is a mere spectacle, in history it is re-enacted in present thought” (Idea, 293).*° Taken out

of context, this seems absurd, and even in context it may well be incorrect. But it is at least intelligible when one sees what Collingwood

attempts to accomplish with it. It emerges as an attempt to answer the question: Since history is being done and is, therefore, patently

possible, although the events of the past are forever beyond our 9 See my “Evidence and Events in History,” 175-179. (In this volume pp. 3-9.) ” Cf. “If we raise the question, Of what can there be historical knowledge? ‘The answer is, Of that which can be re-enacted in the historian’s mind” (/dea, 302).

COLLINGWOOD’S THEORY OF HISTORICAL KNOWING 301

acquaintance, what is there for the historian to deal with? ‘The answer turns out to be thought. Without dealing in detail with Collingwood’s

view of thought or thinking, it will suffice to point out that for him it is something which can survive in the way that events—social events

and psychological ones—cannot. The historian understands what transpired when he is able to re-think the thought of the historical actor he is dealing with. In some relevant sense, Collingwood takes the actor’s thought and the historian’s re-thinking it to be identical (Idea, 288ff.). In seeking to understand a theorem of Euclid’s Elements

or the point of the argument in some dialogue of Plato, to use examples he uses himself, one has, in effect, to re-think their thoughts. There is no significant sense in which the original thought and its present re-thinking are two, not one. Before bringing this long discussion of Collingwood’s theory of historical knowing to an end by completing our account of what 1s involved in this idea of re-thinking, we had better be sure we fully understand how it works. It may seem to have some point in intellectual history—understanding Euclid or Plato—but there 1s more to history than that. Collingwood, himself, says that politics, warfare,

economic activity, and religion are among the proper subjects of historical inquiry (/dea, 309ff.), but I dare say that any number of his

readers have failed to be convinced that, given his views, he could make that out. I do not wish to make up arguments which might be offered in a Collingwoodian spirit in support of these assertions.

Rather, I prefer to turn to an actual instance of Collingwoodian history, written by Collingwood himself. Though Collingwood as historian worked on the seemingly unlikely brute vealza of archaeology, his way of dealing with them 1s quite consistent with his conception of historical knowing as present re-thinking and gives substance

to his claim (Idea, 296) that one recovers the thought of others with the aid of evidence. _As is well known, Collingwood combined a career as philosopher with that of historian, being for some time University Lecturer in Philosophy and Roman History at Oxford (Roman, vii).*! His own special field was Roman Britain, and for a time he did quite a bit of writing on the subject. I do not know what his reputation was among *! Page references given in the text preceded by Roman will be to Collingwood’s “Roman Britain” in R. G. Collingwood and J. N. L. Myres, Roman Bntain and the English Settlements, 2nd ed. (Oxford, 1937).

302 COLLINGWOOD’S THEORY OF HISTORICAL KNOWING

his fellow historians,** but one may suspect that his work was held in some esteem, since he was invited to contribute the more than three

hundred pages devoted to Roman Britain in the Oxford History of England. But whether esteemed or not, his historical work does show that he himself felt no difficulty about the possibility of writing his-

tory which was anything but intellectual history or the history of ideas. And if the sort of history he writes can be treated as being the re-thinking of past thought, it seems only reasonable to conclude that that 1s what he meant by his famous phrase all the time. It seems

considerably more sensible to explicate the phrase in the manner here suggested than to determine what it means by a contextless analysis of it as simply a phrase in the English language. One of the major preoccupations of Collingwood’s Roman Britain is to make sense of archaeological remains; he seems particularly interested in fortresses, walls, and the like. ‘The question frequently arises:

“why should anyone build something like that?” and, of course, the attempt to provide an answer leads Collingwood to try to discern a purpose that the construction could serve. Although I do not wish to repeat Collingwood’s detailed descriptions of the remains he deals with—which are hard to visualize in any event, since the volume has

no plates or illustrations—I do think that one good example will make clear what is involved. Somewhat to the south of Hadrian’s Wall, but following the line of the Wall, there is an earthwork which, Collingwood tells us, English antiquaries call the Vallum (Roman, 124). After presenting an account of its chronological relations to the building of the Wall itself, the question of why it was built 1s dealt with, and we are told that this has always perplexed scholars who have formulated hypothesis after hypothesis only to have them rejected as incompatible with the actual features of the Vallum (Roman, 133). Collingwood proceeds

to detail some of the features in order to determine what it might or might not be used for. We are told (Roman, 133), that the Vallum in its original shape was a formidable obstacle to traffic, but incapable of military defense, and so designed, indeed, as to look ostentatiously unmilitary; that this obstacle is carried with remarkable thoroughness, admitting no interruption whether from hard rock sub* Cf. I. A. Richmond’s generally very favorable “Appreciation of R. G. Collingwood as an Archaeologist,” Proceedings of the British Academy 29 (1943), 476-480. I am

grateful to W. von Leyden for calling this to my attention.

COLLINGWOOD’S THEORY OF HISTORICAL KNOWING 303

soil, morass, or ravine, right across the country from Tyne to Solway, close behind the Wall; that, according to the latest results of excavation, it was made at the same time as the Wall itself; and that the only original ways across it are solid causeways opposite the Wall forts and

perhaps also opposite the milecastles, each surmounted by a stone gateway. In sum, the Vallum is a second obstacle parallel to the Wall and provided with a corresponding series of controlled openings for traffic, differing from it in its deliberately unmilitary design.

If at first blush a second obstacle seems a redundancy and one 1s, hence, tempted to think of it as an earlier, perhaps a more primitive, less satisfactory attempt to accomplish what was better done subsequently, we are forced to face both the archaeological evidence of their contemporaneity and that the unmilitary character of the Vallum would presumably preclude its serving the defensive function the Wall

of Hadrian is known to have had. The question, then, is “what function could the Vallum serve?”— or, indeed, “what is the Vallum?” Drawing upon what he knows of Roman frontiers, administrative arrangements, and the like, Collingwood offers an account concerning which he says, “There is no proof that this explanation of the Vallum is correct. All that can be claimed for it is that it fits the facts” (Roman, 134). Yet it is so excellent an example of the kind of historical thinking one finds throughout Ao-

man Britain that it is worth quoting in full the three paragraphs involved (Roman, 133f.).

Now, a Roman frontier has two functions, one military or defensive, the other financial, as a line where trafhc passing between the province and the unconquered country outside it passed through supervised openings and paid duty. And it is a peculiar feature of Roman administration that the financial service under the procurators was entirely separated from the military service under the provincial governors. The sentry on guard at the gate of a fort was responsible to the commandant; he, whether or not through the mediation of a legionary commander, to the governor of the province; and he to the emperor. A customs officer was responsible to the procurator, and he to the emperor directly, without any intervention of the governor. And not only were these two services separate, but relations between them were delicate: friction and jealousy were not unknown. Before the building of Hadrian’s Wall, contnuous frontierworks, where they existed at all, had been structurally separate from the forts of their

garrisons. It may be conjectured (we have no proof) that the openings in the barrier were controlled by customs officers, while obviously the sentries of the garrisons looked after their own fort gateways. But on Hadrian’s Wall the forts, with one or two exceptions, formed part of

304 COLLINGWOOD’S THEORY OF HISTORICAL KNOWING

the barrier itself, so that a man passing through the fort was passing the line of barrier; and (again with one or two exceptions) there was no way of crossing the barrier except by going through a fort, unless, indeed, non-military traffic was allowed to pass through milecastles; if it was, the same problem would arise there. From a military point of view, this new method of planning the forts in relation to the barrier was no doubt an improvement. If traffic crossing the line of the barrier was compelled to pass through a fort, the military control over such traffic was tightened. But the question must now have arisen, how to provide for the customs officers? Hadrian, a stickler for military discipline, may very well have thought it unwise to give the procurator’s men an official position at fort gateways, where

the authority of the commandant should be undisputed. The simplest solution on paper, though a cumbrous and expensive one, would be to have a second barrier behind the Wall; to make this barrier look as unmilitary as possible, consistent with efficiency; and to provide it with a crossing opposite each fort, where the customs officers could do their work. ‘The Wall as a whole would be controlled by the governor, the Vallum by the procurator; the distinction between the two reflecting and symbolizing the separation between the military and financial services.

As I have indicated, this kind of thinking is typical of what Collingwood

does in this study. I think by itself it will prove sufficient to help us elucidate what he means by the historian’s re-thinking.* Since The Idea of History is not part of any actual instance of his-

torical research, at least not in the sense of an historian’s reconstructing some past event,** what Collingwood can mean by some of his assertions may seem perplexing at times. He has been telling us

that history is the re-enactment of past thought, and then says: The historian cannot apprehend the individual act of thought in its individuality, just as it actually happened. What he apprehends of that individual is only something that he might have shared with other acts

of thought and actually has shared with his own.... It is the act of thought itself, in its survival and revival at different times and in differ-

ent persons: once in the historian’s own life, once in the life of the person whose history he is narrating. (/dea, 303)

*8 For other instances of Collingwood’s historical thinking see Roman, 140, 142, 159, 160 (Postscript), and 172f., as well as his extended discussions of the strategy of Julius Caesar, Claudius, and Aulus Plautius. “* Of course, the first two-thirds of the book is an fustorical sketch of the development of the idea of history, but that is not the sort of work I have in mind. In any event, the sections of the last part of the volume were separately written essays or lectures.

COLLINGWOOD’S THEORY OF HISTORICAL KNOWING 305

One can imagine philosophical critics wondering about what sort of

thing an act of thought is which is to be apprehended by another who can share it “itself” but cannot apprehend it in its “individuality.” But it does not really matter. Whatever these sentences may be thought to mean, or whatever someone else might mean who through some extraordinary coincidence happened to express himself in these very same words, what Collingwood means to convey by them 1s clear enough in light of his account of the Vallum and its purpose. Collingwood does not claim—indeed, he denies—that he 1s able to

experience—in whatever sense is suitable here—the actual act of thinking that took place in antiquity.” His claim is that he can reconstruct the thought behind the decision to build the Vallum, which

means, of course, that he is able to re-think the thought—though not, to be sure, exactly in the same order in which the various elements of the project were taken into account, and not in Latin but in English. The problem of the Vallum cannot be solved until the historian concerned with it can figure out what human purpose it served or could serve. What that purpose 1s, is not self-evident; one does not immediately hit upon it as soon as one determines its physical

character. One has to think about what could be done with it, and not by one of us but, rather, by a Roman ofhcial highly enough placed as to be able to have such a thing made.” If Collingwood’s * Cf., “To return to our supposed objector. Why did he think that the act of thought, by becoming subjective, ceased to be objective? ... It is because he understood by subjectivity not the act of thinking, but simply consciousness as a flow of immediate states.... What the objector was doing, therefore, was to assume that all experience is immediate, mere consciousness, devoid of thought” (/dea, 294). It 1s thought, not immediacy, which Collingwood takes to be the object of historical knowledge. *6 One has to put oneself in Hadrian’s place and make sense of what he did in terms

of the problems he was seeking to deal with, the questions he sought to answer. Cf. C. V. Wedgwood’s criticism (in The Sense of the Past [originally published as Truth and Opinion], Collier Books ed. [New York, 1967]) of Gibbon’s condemnation of the ac-

tions of Sulpicianus after the murder of his son-in-law, the Emperor Pertinax, as reflecting Gibbon’s inability to see the problems from Sulpicianus’ own standpoint: “Gibbon cannot and does not imagine himself in the predicament of Sulpicianus and an obvious element in the situation therefore escapes him. Gibbon’s method enables him to describe and explore the surface of events with incomparable brilhance, but it rarely leads to any penetration below the surface” (34). Clearly, Miss Wedgwood’s point is that Gibbon simply failed to re-think Sulpicianus’ thought in Collingwood’s sense. Note also her following remark in an entirely different context: “The wars of religion have left their mark on the institutions, the society and the prejudices of most of the peoples of western Europe. These bitter and distracting conflicts are not only dismal and deplorable but largely incomprehensible until we can bring to them some understanding of the beliefs which guided the protagonists” (46).

306 COLLINGWOOD’S THEORY OF HISTORICAL KNOWING

solution to the Vallum problem is correct, there is a clear sense in

which he has re-thought the thoughts of Hadrian in all of their historically-relevant character. The essential considerations which presumably passed through Hadrian’s mind as he came to the decision to have the Vallum built have passed through Collingwood’s as well. There is no suggestion of his having entered fully into the existential experience of the historical actor in the sense of reproducing the feelings,

emotions, and other appurtenances of existing-and-experiencinghere-and-now in the way that an historical novelist might seek to do.*’ Here, it seems, is the central feature of what Collingwood thinks history is: it is re-thinking thought on the basis of evidence without ever becoming psychology. As re-thinking thought, its object is what

can be detached from the original context of action and be reproduced in the later context of historical inquiry, hence its object is universal and not existential.” We are now in a position to offer an account of what is involved in one of the most quoted of Collingwood’s assertions: “For history, *” Hugh J. Schonfield’s fascinating and suggestive The Passover Plot (New York, 1966) is frequently flawed by just such a pretense that he is able to write or speculate about how Jesus felt at some time or what he might have said at another. *® It should be clear from all this that when Collingwood says that historians rethink past thought he means to have “thinking” taken literally. But the sheer incapacity of his critics to recognize thinking taken seriously as thinking taken seriously is, indeed, a wonder to behold. Thus, Gardiner says that “the suggestion of some sort of telepathic communication with past thoughts is too insistent to be entirely disregarded” (op. cit., 39) and that “the subject-matter of history is made to appear very mysterious indeed, demanding tentative handling and esoteric methods” (zbid., 49). F. D. Newman speaks of “Collingwood and his infamous doctrine of empathic understanding” (Explanation by Descnption [The Hague and Paris, 1968], 51; the text actually says “emphatic,” but this is clearly a typographical error). And W. H. Walsh

proclaims that “It is not true that we grasp and understand the thought of past persons in a single act of intuitive insight” (An Introduction to Philosophy of History [New

York, 1951], 58) as if Collingwood ever held that it was. (Walsh has discussed Collingwood’s views elsewhere as well: “R. G. Collingwood’s Philosophy of History,” Philosophy 22 (1947), 153-160, and “The Character of a Historical Explanation,” Aristotelian Socuety Supplementary Volume XXI (1947), 51-68. One occupational

hazard of doing philosophy is that your words may be given the worst possible interpretation by a critic who then saves the day by claiming what you had been trying to say all along. The treatment of Collingwood in Walsh’s book exemplifies this very nicely.) Other such critics include L. Jonathan Cohen, “A Summary of Work in the Philosophy of History, 1946-1950,” Philosophical Quarterly 2 (1952), 173-

177, and “Has Collingwood been Misinterpreted?” zbid., 7 (1957), 149-150 and G. J. Renier, History: Its Purpose and Method (Boston, 1950). How these direct, immediate, non-thinking forms of mental communication have been foisted upon a writer so unambiguously committed to thinking must be one of the most remarkable intellectual accomplishments of our time.

COLLINGWOOD’S THEORY OF HISTORICAL KNOWING 307

the object to be discovered is not the mere event, but the thought expressed in it. To discover that thought is already to understand it. After the historian has ascertained the facts, there is no further process of inquiring into their causes. When he knows what happened, he already knows why it happened” (/dea, 214). Actually, it 1s only the last of these sentences which is really famous, but I have chosen to quote a somewhat larger extract because my interest here 1s different from that of any number of others who have discussed these words. For the most part, they are discussed in writings on historical explanation,

but my interest here is not in Collingwood’s views on that subject, but, rather, on his conception of historical knowing. Hence, I begin the extract quoted with Collingwood’s statement of what the object to be discovered is—that is, the object to be discovered no matter how it is to be explained. I should not wish to deny that the last of the sentences quoted contains a conception of historical explanation, but it is a very unusual sentence. It takes all of its words to express its conception of historical explanation, yet it manages to say something

else as well, something about the nature of “what happened,” that is, about “the object to be discovered” in historical research. It should be possible to provide a complete description of that earthwork which is the Vallum in physical and chemical language. Likewise, in principle, as the saying goes, an ideal observer*’ or a whole research team of ideal observers might have been able to provide

a complete behavioristic description of the movements of Caesar’s forces in the course of his invasion of Britain. But the former would

not be an account of the Vallum gua the Vallum, and the latter would not be a description of an invasion qua invasion. The Vallum

as such is a product of human purpose; without that purpose it is something simply there, and unintelligible. Consider the following well-known example: “Suppose that I enter a bank, I then take a withdrawal slip and fill it out, I walk to a teller’s window, I hand in my slip, he gives me money, I leave the bank and go on my way. Now suppose that you have been observing my actions and that you are accompanied by, let us say, a Trobriand Islander. If you wished to explain my behavior how would you proceed?””’ Pre-

sumably, the Trobriander sees the course of what takes place as a * Who, of course, would not be an historian. °° Maurice Mandelbaum, “Societal Facts,” in Theortes of History, ed. Patrick Gardiner (Glencoe, Ill., 1959), 479.

308 COLLINGWOOD’S THEORY OF HISTORICAL KNOWING

string of unintelligible bits of behavior. The other observer sees it as an instance of withdrawing money from one’s savings account. The problem is to get the Trobriander to see it in the same way, so that when, at last, he knows what happened, that money was being withdrawn, that each of these discrete—describably discrete—bits of be-

havior are parts of a certain course of action, he also knows why each describably discrete bit of behavior happened. When Collingwood

knows what the Vallum really is, he knows why it was built. When he knows what Caesar’s strategy was, he knows the record of the ideal observer noted above not as behavior, but as the movement of troops. In sum, what 1s involved in the passage quoted from Collingwood is simply the claim that the object of the historian’s research is human action, not behavior understood physicalistically.*! This is no place to raise the issues of behaviorism. Whether behaviorism is tenable or not, Collingwood’s point is that its object is not that of the historian. If all that we could have were the observations of behavioristic psychologists—if, that is, the methodological requirements of behavioristic psychology became the elements of the only possible theory of human nature’?—then, in Collingwood’s view,

history as a discipline would be impossible. The object of historical research is human thought, because only thought can be abstracted from its original context and reproduced in the historian’s mind. What Collingwood is trying to say is that the historian studies action—an “outside” informed by an “inside’”—not behavior. Anything exhausted by its immediacy cannot be reproduced in the mind of the historian.°° °! Critics (cf. Walsh, An Introduction to Philosophy of History, 52) have been quick to

cite Collingwood’s “When he knows what happened, he already knows why it happened” (/dea, 214) as evidence of his intuitionism, and taken out of context, it does seem to express intultionist sentiments. But we have already seen that Collingwood

was not and could not be an intuitionist, a conclusion surely reinforced by my non-intuitionist interpretation of the offending passage. (Actually, the argument really

goes the other way. It is precisely because Collingwood cannot be an intuitionist that seemingly intuitionist assertions require to be given non-intuitionist interpretations. We are not, after all, dealing with contextless statements in ordinary English.) *’ See my “The Phenomenological and Naturalistic Approaches to the Social,” in Philosophy of the Social Sciences: A Reader, ed. M. Natanson (New York, 1963), 289ff.

°° Collingwood’s use of the “inner-outer” metaphor to characterize human action has provided another occasion for critical attack. All he intended by the expression is that human action consists of behavior informed by thought, and it is difficult for me to see how any fair-minded reader could fail to see that. Gardiner says that the expression is “artificial and misleading”: artificial because we do not talk that way about human action—a petty observation of no importance—and misleading because “the introduction of a spatial metaphor gives the impression that what are called ‘insides’ of events are queer objects, invisible engines that make the wheels go round”

COLLINGWOOD’S THEORY OF HISTORICAL KNOWING 309

IX The burden of Collingwood’s investigation into the nature of historical knowing is to discover the grounds of the possibility of knowing

now what happened then. That historical knowledge is possible, (tbid., 47). To say that Collingwood has used a metaphor that may be misleading is both to recognize it for what it is and not to be misled by it. But what follows in Gardiner’s book is an attempt to show that if anyone was misled it was Collingwood

himself; or perhaps “inside-outside” is not a metaphor after all. I would only say that we have already found in Gardiner (see notes 20, 21, and 28) an inclination to convert poor Collingwood into the one thing he would not be, an historical realist believing in the existence of strange objects which the historian knows directly by intuition or acquaintance. This is simply another instance of the same tendency. Given what Collingwood’s actual views were, he could never have intended that the metaphor be taken literally and that human thought be something spatial. (Indeed, I cannot imagine anyone of any philosophical persuasion not bereft of sense believing in the various entities—pasts which are present, thoughts which are spatial, and the like—that Gardiner attributed to Collingwood). Gerd Buchdahl’s treatment of the “inner-outer” matter is also worth our attention (“Logic and History, An Assessment of R. G. Collingwood’s Idea of History,” Australasian Fournal of Philosophy 26 [1948], 94-113). Collingwood notes that for “the scientist, nature is always and merely a ‘phenomenon,’” (/dea, 214) but though he uses “phe-

nomenon” there is no reason to think that he has Kant in mind. For he immediately explains what he means: “not in the sense of being defective in reality, but in the sense of being a spectacle presented to his intelligent observation; whereas the events of history are never mere phenomena, never mere spectacles for contemplation, but things which the historian looks, not at, but through, to discern the thought within them.” The point is two-fold: human beings act, not simply behave, and one must get to the thought to know what the act is. And given the pastness of the past, its non-observable character, it is this, the possibility of getting to the thought, that

makes historical knowledge possible—assuming, that is, that thought is what Collingwood takes it to be and his belief that the same thoughts may be thought by many. But in Buchdahl’s treatment all this is lost or ignored. Instead the distinction between inner and outer becomes Kant’s distinction between things-in-themselves and appearances. Thus, Collingwood “allowed himself to be misled by the metaphysical use of the adjective ‘inner,’” and owing to his preference for internal relations—a doctrine Buchdahl foists on Collingwood from Bradley on the ground that “Like most philosophers, Collingwood is not too clear on the nature and content of his ideals, is not clear how what he feels as a deficiency determines the choice of his ideal” (ibid., 96)—he thought that thought presumably the historical thing-in-itself,

would enable him to have them. Having saddled these incredible views upon Collingwood, Buchdahl now criticizes them: “But it seems to me that in the end the other person’s thought, particularly when this thought .. . 1s re-constructed, is just as

much reached from the outside... is as much ‘an appearance’ as anything else. That Collingwood did not realize this may be due, among other things, to his peculiar views of thought...” (zbzd., 98). Buchdahl’s treatment of Collingwood is altogether an incredible performance. To

understand Collingwood, he finds hints in Bradley and others in Kant; what he does not do is allow Collingwood’s own concerns, together with the way the actual discipline of history is practiced, to provide the context in terms of which to interpret The Idea of History. This will be clear from a moment’s consideration of the way he

310 COLLINGWOOD’S THEORY OF HISTORICAL KNOWING

Collingwood did not doubt. His Speculum Mentis view was that history is concerned with a world of fact. ‘The holistic view of knowledge he

then held led him to conclude that historical knowledge was not possible, but even without that, the objectively-there character of that world of fact would have led him to reject it as soon as his anti-realistic views crystallized. He could not, however, simply deny history, and seeking to discover just what it could possibly be about—what subject

matter one could find for it to deal with—he came to think of it as the science of historical evidence. Any number of essays between Speculum Mentis and The Idea of History take this view. But while much

is made of the critical treatment of historical evidence, that middle position is not fully satisfactory inasmuch as it fails to account for the historian’s treatment of events. Without some way of bridging historical evidence and historical events one of the central concerns of the historical enterprise remains unintelligible. His quest for the solution of this problem, beyond doubt accompanied by very serious attention to what he was himself doing as an historian, led him back once again to a world of historical fact, but not to the lived expenence of the historical agents he was concerned about. Packed into The Idea of History is a conception of action and a conception of thought, which

together hold that human action is behavior informed by thought, that history is concerned only with action, and that by making proper

use of historical evidence the historian can bring to his own mind thoughts once thought before. History can be about only what the historian can know: not the real past, but thought he can re-think. I do not wish to say that Collingwood’s most mature doctrine is free of problems, though it is always fair to remember that illness brings in the doctrine of internal relations mentioned parenthetically in the previous paragraph. Making the usual realistic assumptions, Buchdahl thinks that Collingwood “loosened the facts from a solid background” and was then confronted by the need

to tie them back together in an even more intimate way than they were before. Collingwood’s motive is said to be that he 1s committed to the doctrine of internal relations, and it is that which led him to emphasize that thought is the material of history: ““Thought’ you will find is the model making the assertion of internal connection plausible” (zbid., 96). But all this is simply wrong. It is only because he has not paid any attention to how the historical past is constructed by historians that Buchdahl can say that Collingwood loosened the facts from solid moorings: as if the historian ever confronts real, solidly-moored historical facts, and as if he does not himself construct his facts on the basis of historical evidence. As for thought, we have already seen how, far from being introduced to justify the Bradleyan doctrine with which Buchdahl saddles him, its central place in Collingwood’s theory is entirely owing to his view that it alone can survive its original context to be re-enacted in the historian’s present.

COLLINGWOOD’S THEORY OF HISTORICAL KNOWING 311

and premature death prevented his bringing his philosophy of history to a conclusion. In particular, even if he is nght in thinking that history can only be concerned with what is informed by thought, to be informed by thought is not to be exhausted thereby. One still requires a non-realistic account of the relation of the entire historical event that the historian reconstructs to historical evidence. But even as far as we have it, the doctrine of The Idea of History is certainly not the work of an historical skeptic, and only one who 1s either a realist in epistemology or a behaviorist in philosophical anthropology could

credit that charge against it.

+ Cf. my “Evidence and Events in History.”

COLLINGWOOD ON THE CONSTITUTION OF THE HISTORICAL PAST I

One of the pervasive features of Collingwood scholarship, particularly as concerns the interpretation of his philosophy of history, 1s the virtually unanimous refusal of his critics and commentators to approach his view on his own terms. Thus, almost no effort is made to place The Idea of History in its own proper context, the two main elements of which are Collingwood’s work as historian and his philo-

sophical preoccupation with what he had been doing as historian. The overwhelming interest of present-day philosophy of history, at least in the English-speaking philosophical world, is with the nature of historical explanation, and it is clear enough that most writers who discuss Collingwood on history purport to present, modify, or criticize his conception of that. Almost no one, I am sure, doubts for a moment that the well-known dictum about the historian rethinking or re-enacting past thought is intended as a contribution to a theory of historical explanation. Unfortunate for this interpretation—shared both by hostile and friendly critics alike—is that there are two sorts of evidence that may be cited against it: the actual language of The Idea of History and the actual practice of ‘Roman Britain’.’ In addition, the usual view would be entirely inconsistent with Collingwood’s conception of human action. Thus, he says: “Since history proper is the history of thought there are no mere “events” in history: what is miscalled an “event” is really an action, and expresses some thought (intention, purpose) of its agent’ (A, 127f.). In The Idea of History he says: ‘By the outside of the event I mean everything belonging to it which can be described in terms of bodies and their movements. . . .

By the inside of the event I mean that in it which can only be described in terms of thought. ... The historian is never concerned with

either of these to the exclusion of the other. He is investigating...

' Collingwood, ‘Roman Britain’, in R. G. Collingwood and J. N. L. Myers, Roman Britain and the English Settlements, 2nd ed. (Oxford, 1937).

COLLINGWOOD ON THE CONSTITUTION OF THE HISTORICAL PAST 313

actions, and an action 1s the unity of the outside and inside of an event (LH,

213.).2 But if the usual understanding of Collingwood is correct, he

must hold that the two are frequently separated in the historian’s work, and that the historian reconstructs outsides of events—understood in a way at which no behaviourist could cavil—which he must then explain by appeal to an inside. In principle, it should be pos-

sible for historians to agree about the outside and disagree about which inside explains it adequately. But all this clearly flies in the face of Collingwood’s plainly expressed intention. In his view, the two sides of the action are inseparable aspects of it and are not to be treated as different things. To understand what a philosopher of history who was himself an historian means by his central notions, it seems most reasonable to look for guidance to his own historical writings. There are quite a number of passages in ‘Roman Britain’ which clearly exemplify the re-enactment of past thought, and I shall be presenting a number of them further along in my discussion. But it will be most instructive, I think, to attend to the character of something Collingwood actually offers as an explanation. I think that some will find it rather surprising. The problem is set by evidence of the decline of villas all over Britain late in the fourth century. There is very little directly to go on, but we do have an account, by a writer named Salvian, of similar difficulties in Gaul, and these are related to a series of peasant revolts. Collingwood thinks that the Gaulish situation is relevant to the problem which confronts him. ‘Uhus, he writes: The underlying cause of these recurring peasant revolts, as expounded by Salvian, was the contrast, not in wealth alone but in security, between rich and poor. The great landowners were favoured by the incidence of taxation, and could pass on their burdens to the poor. The legal and administrative system of the late empire favoured economic tyranny. It was in the power of a rich man to deprive a poor man of

all he possessed; and Salvian gives examples where the power was exercised without pity and without appeal. Hence, to the poor, ‘the enemy is kinder than the tax-collector’; it needed only the occasion of a barbarian inroad to convert exasperated peasants into Bacaudae’ and bring into existence wandering bands of broken men, escaped slaves, and despairing debtors.*

* Italics added. * The Gaulish term for the peasant rebels. * ‘Roman Britain’, p. 304.

314 | COLLINGWOOD ON THE CONSTITUTION OF THE HISTORICAL PAST

Thus the situation in Gaul. But what has that to do with Britain? Collingwood proceeds to make the connection: The same legal and administrative system, the same distinction between rich men in great villas and poor men in village huts, and the same barbarian invasions, were present towards the end of the fourth century in Britain. Causes being identical, it is hardly to be doubted that effects were identical too; and that the wandering bands which Theodosius saw in Britain included large numbers of Bacaudae. But every man who became a Bacauda ceased to be a productive labourer. Consequently the rich estates, in addition to suffering actual plunder and the deprivation of trade, suffered a diminution in their own productive powers.”

Here, surely, is an explanation to warm the soul of the most toughminded covering-law theorist. ‘There 1s no obvious way in which this explanation rests upon re-thinking thought. Instead, we find reference to social and economic causes of social and political disorganization with particular emphasis upon the explanation of recurrent effects by appeal

to recurrent causes. It would, of course, be a mistake to claim Collingwood for nomothetic social science and naturalism. ‘There are

too many other things he has said that would preclude that. But here we do have a perfectly tenable historical explanation which is quite clearly unlike what we have been led to expect that such an explanation offered by Collingwood would be like. And I think it provides still another reason to doubt the general view that for him historical explanation 1s accomplished by rethinking. II

If one turns directly to that part of The Idea of History in which the doctrine of rethinking gets its most systematic exposition and reads it without expectations built upon our own interests—particularly our interests in explanation—that anyone should think that it is intended by its proponent as a theory of explanation becomes more surprising still. ‘The section in question, called ‘History as Re-enactment of Past Experience’, opens with the question: ‘How, or on what conditions, can the historian know the past?’ And the beginning paragraph ends with the question: ‘If then the historian has no direct or empirical 5 Tbid., p. 304.

COLLINGWOOD ON THE CONSTITUTION OF THE HISTORICAL PAST 3195

knowledge of his facts, and no transmitted or testumoniary knowledge of them, what kind of knowledge has he: in other words, what must the historian do in order that he may know them?’ He immediately tells us that his ‘historical review of the idea of history has resulted in the emergence of an answer to this question: namely, that the historian must re-enact the past in his own mind’. And he

thinks what he has been saying may be easily understood in the following way: ‘When a man thinks historically, he has before him certain documents or relics of the past. His business was to discover what the past was which has left these relics behind it... . This means discovering the thought... expressed by them’ /H, 282f.). It is hard to see how Collingwood could have made his intentions any plainer. Re-enactment of past thought is part—indeed a central part—of his conception of how the historical past is known or constituted in historical research. It has nothing to do with explanation at all. The point to be kept in mind is that historical events are human actions. Thought, the inside of the action, is an inseparable part of it, and in order to know at all what the action was—a determinate piece of military strategy or the attempt to solve a specific mathematical or philosophical problem—-you have to know the thought in terms of which some bit of overt behaviour may be recognized as action.® I shall not insist upon this any more at this point other than to reit-

erate than what I have quoted makes it clear that Collingwood’s own intention was to use the idea of re-enactment in his account of how the historical past is known, not explained. Instead, I shall turn directly to the presentation of some actual examples of re-enactment from Collingwood’s ‘Roman Britain’ in order to show that the purpose he said re-enactment serves is precisely the purpose he tried to make it serve in his own historical writing. We shall begin by considering what Collingwood has to say about the activity in Britain of the emperor Severus early in the third century. We are told about the restoration of fortifications, campaigns ° I venture to predict that any attempt to save the peasant-slave revolt example from exploitation by covering-law theorists by showing that it may be conceived in terms of re-enactment, would, in effect, prove to be dealing with the event being explained and not the explanation. And that would only reinforce my belief that there is nothing about the covering-law thesis that weds it necessarily to behaviourism. Because most covering-law theorists are behaviourists, it is assumed, without argument, that they have to be; cf. A. R. Louch, Explanation and Human Action (Berkeley and Los Angeles, 1966) pp. 64, 65, 68, 69, 70, 71, 83, 90ff., 101. At very least, this ought to be discussed and not settled by tacit agreement.

316 = COLLINGWOOD ON THE CONSTITUTION OF THE HISTORICAL PAST

in Scotland which the emperor came to Britain to lead in person, heavy casualties as a result of the successful use of guerilla tactics by the enemy, and so on. After that, we know ‘that Severus died, worn out by his labours, at York in 211, and that his sons forthwith broke off the Caledonian war and returned to Rome’.’ Collingwood now turns to a final summary account of the situation he has been talking

about, and it will be helpful simply to quote a bit at length: It looks like the story of a misguided and wholly unsuccessful war. Yet after its conclusion the British frontier enjoyed unbroken peace for nearly a hundred years. If the war ended in complete failure for Rome, why did the Scottish tribes do nothing to follow up their victory? ... But it

is difficult to say that the second phase of his frontier policy was a failure unless we are first sure what it was meant to achieve. Certainly not the complete conquest of Caledonia; if that had been its aim the elaborate reconstruction of Hadrian’s Wall would have been sheer waste. Certainly not the restoration of the Antonine Wall, or he would have

got to work on that... as soon as he had possession of the ground on which it stood... It seems that he did not intend a permanent occupation of any part of Scotland... The evidence... suggests that his campaigns were meant as wars of devastation, . . . punitive expeditions, visiting the wrath of Rome on enemies outside her grasp but not out-

side her reach.. .°

To some it might appear that having presented an account of what Severus had done, Collingwood was now attempting to explain it by eking out the thought behind it. But this is not the case. The earlier remarks which may seem to describe what Severus did are, in fact, only limited to the event—in the sense we have seen that Collingwood

gives to this word—the external husk which is only one aspect of the action in which he was engaged. ‘To be told that Severus sent troops to Britain, had some fortifications restored, and came, finally,

to lead the troops himself, is still not to know what Severus did. That, surely, is the point of his saying that ‘it is difficult to say that the second phase of his frontier policy was a failure unless we are first sure what it was meant to achieve.’ Knowing only the sorts of thing mentioned two sentences ago is not knowing yet what the action upon which Severus was engaged actually was. What Collingwood

tries to do in the paragraph from which the above is excerpted is to examine several possibilities in order to determine which of a ” ‘Roman Britain’, p. 159. 8 Ibid., p. 159£

COLLINGWOOD ON THE CONSTITUTION OF THE HISTORICAL PAST 317

number of seemingly plausible alternatives could be the nght one. In other words, Collingwood is not attempting to explain what Severus

did as much as he is seeking to determine what the policy he was pursuing actually was. It is only once that has been established, once,

that is, we know what the action we are concerned with happened to be, that we can raise the two obvious subsequent questions: (1) ‘Was it successful?’ and (2) ‘Why was such a policy adopted at all?’ The second of these is the explanation stage, and while we may expect that Collingwood would disagree with a good bit that has been said about explanation in history in the last decades, it 1s clear from the peasant-slave-revolt example discussed earlier that he was open to a wider variety of possible explanations than those believe who think re-enactment of past thought is his conception of explanation. It is, in any event, at the stage of inquiry during which what the historical action was is being determined that re-enactment takes place. What Collingwood does in the material just presented in the long excerpt is attempt to re-enact Severus’s thought, to try to see his problems from the vantage-point of his situation and in terms of the alternatives of policy which were open to him, thus to determine what in fact he did. This, of course, requires that Collingwood make use of everything he knows about Roman colonial and military administration at the period in question. Rethinking requires a good deal of preparation,’ and one cannot expect to rethink the thought of any historical actor without knowing a good bit of what he must have known before determining his own course. ‘To re-enact Severus’s

thought is to determine what he set out to do; it is not to explain why he set out to do it. The only way to avoid this conclusion, it seems to me, 1s to take what he did as something to be characterized

in behaviouristic terms: a course of behaviour to be explained in terms of its rationale.'° That is, only if the historical event'! has already been fully described before we reach the paragraph from which

the excerpt was made can the excerpt be taken to explain it. But if the event is fully described before the excerpt, it is fully describable

without reference to thought. Whatever merit that view may be thought by some to have, it is not Collingwood’s view. ” It is certainly not the immediate, intuitive grasping that so many critics have imputed to Collingwood. '° Cf. William Dray, Laws and Explanation in History (Oxford, 1957), ch. 5. '' When speaking for myself, I obviously do not feel that there is anything compelling about Collingwood’s usage with respect to this word.

318 | COLLINGWOOD ON THE CONSTITUTION OF THE HISTORICAL PAST

The next instance of re-enactment which we shall consider has to

do with the Antonine Wall. It is perhaps worth noting in passing that much of Collingwood’s ‘Roman Britain’ is taken up with such things as walls and fortifications, not because this is particularly to be wondered at in a book about a period so much of our knowledge of which depends upon the survival of remains that can be uncovered by archeology, but because given what seems to be Collingwood’s

inordinate emphasis upon rethinking one might expect that intellectual history-——construed as broadly as you like—might have been the

branch of history that would best exemplify what he takes history to

be. If it can be shown that even the treatment of such stuff as archaeology deals with can be treated in accordance with Collingwood’s

conception of historical method, then it is clearly the case that his critics have construed Collingwoodian history far too narrowly, and there is no branch of historical scholarship that Collingwood’s view requires us to abandon." Having presented a detailed account of Hadrian’s Wall and the frontier-defence system of which it was the key element, Collingwood

thinks it ‘puzzling to find that by 140, two years after Hadrian was succeeded by Antoninus Pius, drastic changes were being made in the British frontier-system’.' This is revealed by the character of a new wall built in Antoninus’s reign. Like the wall of Hadrian it was continuous, stretching from sea to sea, but it was remarkably different from the earlier wall in a number of ways, having a ‘simpler design and less elaborate structure’.'* Collingwood goes on to detail the ways

in which the Antonine Wall appears to be inferior to its predecessor—absence of milecastles and turrets, small size of the forts, and such hke—but we need not be concerned with the details of his account. I take it that Collingwood is trying to answer the question, ‘What was Antoninus doing here?’—trying to determine what it is that this thing being examined is—and at this point we get at least a partial answer: ‘Evidently, then, the forts of the Antonine Wall were not meant to be occupied each by a complete regiment’.'!’ To '2 Which is entirely what one expects when one realizes that Collingwood’s is a theory about what the discipline of history is and how it carries out its work; it does

not purport to say what historians ought to do, other, of course, than to say that they ought to do historical work to the best of their ability. See my ‘Collingwood’s Theory of Historical Knowing’, History and Theory, ix (1970), pp. 3-36. (In this volume pp. 273-311.) '5 “Roman Britain’, p. 140. ‘+ Tbid.

'° [bid., p. 141.

COLLINGWOOD ON THE CONSTITUTION OF THE HISTORICAL PAST 319

this extent, at least, he was doing something different from what Hadrian had done. Hadrian had built a wall with fortifications that could be manned by larger units than those Antoninus seems to have had in mind. We shall come back to this question and its answer in just a moment. Having reached his initial conclusions about the nature of the units for which the fortifications were designed, Collingwood turns to what he takes to be evidence of economy in the construction. Presumably, the size of the forts is also evidence of economy, but in addition the choice of materials—‘the rampart itself, instead of stone, is made of turf in the western and central part, of clay in the eastern’ and ‘the official central buildings in the forts were not uniformly of stone, and the barracks were of the cheapest, wooden hutments which in some cases had thatched roofs’'®—underscores this still more. Summing up his conclusions about the Antonine Wall, Collingwood says the following: Both in construction and in organization, then, the Antonine Wall bears the marks of a deliberate effort after cheapness, at the cost of a serious decrease in efficiency. The same thing is to be seen if we consider its

strategic position. Both its flanks, especially the left, lie unprotected upon narrow estuaries, easily crossed by the smallest craft in almost any weather. If we recollect the care with which Hadrian... fortified the whole of the northwest Cumberland coast for thirty or forty miles beyond the terminus of the Wall, the complete absence of Roman posts on the Clyde below Old Kilpatrick becomes so striking that we cannot put it down to negligence. These various features of the Antonine Wall

when considered together, seem less like a series of oversights than parts of a deliberate policy, based on the assumption that a powerful frontierwork on that line was not needed."

This conclusion is reinforced by comparing the wall we have been discussing with the same emperor’s fortifications made only five years later in Germany, where, presumably, stronger defence was deemed necessary.

At first appearance, what Antoninus did seems obvious: he built a wall. ‘That thing there which presents itself to the sight of the archeologist is what was done in the second century, and historians of the

period must explain it. But I very much doubt that this is Collingwood’s view. That thing there is just something inert, and just seeing it there does not tell us what it 1s. ‘To be sure, we recognize 6 Tbid., p. 142. 7 Tbid., p. 142f

320 COLLINGWOOD ON THE CONSTITUTION OF THE HISTORICAL PAST

it as a wall or a fortification, but without wishing to raise all manner of issues tangential to our present interest, I would suggest that so to

identify it as a member of a class is not to know what it is as a particular historical construction. Even to know that it was built during

the reign and, presumably upon the instruction of Antoninus Pius is

not really to know what it 1s. What, after all, does the statement, “That is the Antonine Wall’, tell us about the object it designates? To know what it 1s as something historical is to know what purpose it served, what thoughts—policies—it embodies. Short of that, one really does not know what that thing stretched across Scotland from

sea to sea is. One might look at the thing and say: “There is Antonius’s Wall; note the shabbiness of its construction. Standards had certainly declined since Hadrian’s day.’ Or, ‘Notice the cheap-

ness of that wall’s construction; Antoninus was certainly not as munificent as Hadrian.’ But neither of these would be right if Collingwood’s re-enactment of the thought of which the wall is an expression is correct. For this is not just a wall; a cheap wall; a shabby wall. It 1s a wall erected as part of a certain policy which is based upon some determinate assumptions. To call it inferior to Hadrian’s presumes that Hadrian’s is a standard in terms of which such walls are to be judged. But that is not Collingwood’s view. Collingwood thinks that the adequacy of the wall is not to be judged in terms of its semblance to some standard wall, but, rather, as to whether or not it is adequate to the policy it is intended to help carry out. Without

that policy—which Collingwood expects that his re-enactment has brought out—we simply do not know what confronts us at the site of the wall, other than merely the member of a class. ‘The answer to the aforementioned question, “What was Antoninus doing here?’

turns out not to be ‘Building a wall’, to which we then respond ‘Why?’ It turns out, rather, to be carrying out a certain defensive policy, on the basis of these and those assumptions, requiring, zier aha, the construction of a wall having certain minimal characteristics, and so on. All this is established by means of the re-enactment of thought— together, of course, with evidence, the wall itself, to be sure, but all manner of other sources about Roman administrative and military policy. As I understand Collingwood, thus far we have been learning what happened, but not why. That is, Collingwood has been constituting a past historital event, but he has not yet attempted to explain it. This conception of what he was about is somewhat strengthened by

COLLINGWOOD ON THE CONSTITUTION OF THE HISTORICAL PAST 32]

his own words. After presenting the outcome of his re-enactment of the thought involved in the decision to build the wall and making the brief comparison, mentioned above, of the Antonine Wall with the frontier-line constructed by Antoninus in Germany, he adds the following: ‘It is evident, then, that the new features of the Antonine Wall are not due to any general cause, such as a cheeseparing policy on the part of Antoninus Pius, or a decline in the efficiency of Roman military engineering, or a less exacting conception of what was demanded in frontier-works. Their explanation must be sought in conditions peculiar to Britain.’'® The last sentence makes it clear that Collingwood does not yet consider himself to have offered an expla-

nation for the event he has been presenting. In the pages which follow, he attempts to do that. He goes back to the situation in Hadrian’s day, traces the development of the frontier situation and brings the matter up-to-date, as that would be in Antoninus’s time. To some extent there is repetition in that certain of the elements which entered into Antoninus’s thought must now appear as elements which affected that thought: Antoninus, obviously, had to consider the situation in terms of its actual features. I cannot offer a detailed account of Collingwood’s explanation, and I do not wish to decide at all as to whether there is a consistent carrying out of the distinction between constitution and explanation in Collingwood’s text. But I do believe that on the whole, the material preceding the passage I have just quoted is generally taken up with the task of determining what happened by use of the historian’s technique of re-enacting past thought, and that the material which follows the passage is for the most part taken up with explaining why it happened by dealing with the circumstances which Antoninus and his advisers must have taken in account, that is, with the circumstances which must have conditioned”” their thinking to begin with. We have so far discussed two kinds of historical reconstruction, one involving a military campaign and the other an artifact, and | should like to present one more, an example of strategy. I do not wish to argue that strategy is something distinct from a campaign. I think it is, but there is nothing to be gained here from an attempt to distinguish conceptually between them. Anyone who does not sense '® “Roman Britain’, p. 143. ' I do not mean by this ‘caused’ or any other term or notion which suggests that thought and thinking are not taken seriously on their own terms.

322 COLLINGWOOD ON THE CONSTITUTION OF THE HISTORICAL PAST

or feel that there is a distinction here will simply conclude that I have offered two examples of the same sort. The main value I should think, of adding still another example of whatever sort is to reinforce the claim that re-enactment functions in Collingwood’s history in the way I have been saying. It would be interesting, given the historical importance of Julius Caesar, to present in detail Collingwood’s account of the strategy of his invasion of Britain,” but that would require a good deal of effort and a good deal of space. Instead, I shall discuss Just one small point of the strategy of Aulus Plautius who commanded an invasion during the reign of Claudius. Existing evidence makes it clear that the invading fleet left Gaul in three divisions each under a separate commander. It has been suggested that this was so

that each fleet could land at one of three ports which it is known were used by the Romans later in history and then, according to their orders, converge at Canterbury. Collingwood thinks there are two reasons why this cannot be so. To begin with, Plautius’s campaign was modelled upon that of Caesar, and the latter’s account of his campaign gives no reason to believe that good routes were to be had from two of those ports since he only explored one. The other is an apriorl argument based upon ‘a maxim of strategy’, namely, ‘that forces should not be divided in face of the enemy’. The point is that any one of the separate units might encounter an enemy force large enough to cause it to suffer a serious setback at the very start of the campaign.’ In Collingwood’s view, what Plautus had in mind was as follows: It is more likely that when Plautius divided his forces ‘in order that they might not be prevented from landing anywhere’ his intention was to confuse the Britons by making feints at two other possible landingplaces ... while intending actually to land at one only. Recent excavations have proved that this one was Richborough, where the remains of a large camp of Claudian date have been found... The discovery of this large land-locked harbour... was a triumph for the intelligence service of Claudius’s army. It suggests that his staff realized where lay the fundamental weakness of Caesar’s campaign, namely, in

his failure to find a safe and commodius harbour . . .”

Some might think that what Plautius did was distribute his ships and

troops in a certain way, and that what Collingwood now does is 20 ‘Roman Britain’, ch. 3. 21 Ibid., p. 79f. 2 Tbid., p. 80.

COLLINGWOOD ON THE CONSTITUTION OF THE HISTORICAL PAST 323

explain the distribution. But this would seem to require that the mere physical arrangement of people can be treated as human action. That is not Collingwood’s view: given his conception of action no histoncal analogy to behaviourism* could possibly be acceptable to him.

The arrangement of the invading forces represents or embodies a plan, a well-thought-out piece of military strategy, and the task of the historian is to determine what that is. It is the arrangement of the forces as embodying Plautius’s strategy that is the human action Collingwood seeks to reconstruct. The procedure is to try to take account of the sorts of thing that Plautius had in mind—the earlier campaign of Caesar as a model, the need to overcome the limitations of that earlier campaign that led to its failure,** the particular goal of Claudius, and so on. ‘The passing remark about the archeological discoveries at Richborough underscores Collingwood’s view that rethinking must take into account everything the histoman can think of which may be relevant—rethinking must be rooted in knowl-

edge of historical evidence. And it is, I think, worth reiterating the point that this whole procedure, the piecemeal way it 1s carried on, the need to take up each bit of relevant information, and the taking up if only to refute other views that historians have offered, all make evident that rethinking is systematic and conceptual. All those critics

who have treated rethinking as a species of empathy or intuition have simply no idea at all of what it involves.”

Ill Nothing is rarer in the critical literature on Collingwood’s philosophy of history than consideration of his actual work as an historian;

*3 I say ‘analogy to behaviourism’ because the historian does not ever observe the overt behaviour of the people he deals with, but simply postulates it as required by the evidence he has. Far from being based upon observation— in the crude sense— historical reconstruction depends upon intellectual operations which completely defy analysis in behaviouristic terms, namely, the operations of historical research. ** Caesar’s failure to state what the objective of his invasion of Britain was, leads Collingwood to conclude that it was not fully achieved; see “Roman Bnitain’, p. 33. * Among others, the following may be cited; Patrick Gardiner, The Nature of Historical Explanation (Oxford, 1952), pp. 39 and 48; Fred D. Newman, Explanation by Description (The Hague and Paris, 1968), p. 51; W. H. Walsh, An Introduction to Philosophy of History (New York, etc., 1951), p. 58; L. Jonathan Cohen, ‘A Summary of Work in Philosophy of History, 1946-1950’, Philosophical Quarterly, 11 (1952), 172-86, at

324 |= COLLINGWOOD ON THE CONSTITUTION OF THE HISTORICAL PAST

indeed, one almost never sees even the slightest citation of his historical writings.*° Alan Donagan is one of the very few writers to take seriously Collingwood as historian,*’ and it is thus with some

regret that I must take issue with his account of one of the very interesting chapters of ‘Roman Britain’, one which, as Donagan reminds us, Collingwood himself so esteemed that he described it as ‘a

chapter which I would gladly leave as the sole memorial of my Romano-British studies, and the best example I can give to posterity of how to solve much-debated problems in history, not by discovering fresh evidence, but by reconsidering questions of principle’.”? The chapter in question is the one called ‘Art’, and the problem is set by what appears to be the sudden disappearance of Celtic styles of art during the period of the Roman occupation of Britain only to return

in full bloom when that period comes to its end. I shall attempt to show that what Collingwood does in that chapter is the same sort of thing we have seen in the examples of re-enactment we have just been examining. That 1s, I shall try to show that what Collingwood tries to do in his treatment of this problem is to determine what happened with respect to art in Britain during the period in question. Donagan, on the other hand, thinks that Collingwood’s problem is

one of explanation, that he has already an idea of what happened, and is now faced with the need to render what happened intelligible. This is quite clear from the way he formulates the question that he thinks Collingwood sought to answer: “Chere are two problems: why,

if pre-Roman British art was of such high quality, was RomanoBritish art so bad? And why did Celtic art revive after its apparent extinction?’ I take it that the use of ‘why’ in both of these formulations makes it apparent that what is taken to be wanted is an explanation of something already known. A bit further on in his pp. 173-7; and G. J. Renier, History: its Purpose and Method (Boston, Mass., 1950), see Index under ‘Collingwood.’ *© In a footnote to the Introduction to his edition of F. H. Bradley, The Presuppositions of Critical History (Don Mills, Ont., 1968), p. 72, n. 83, Lionel Rubinoff writes,

‘Examples of the reduction of universal principles to concrete individuals can be found throughout the historical wntings of R. G. Collingwood.’ But instead of telling us where, he adds: ‘See, for example, Collingwood’s discussion of the “principle of Incapsulation” in the Autobwgraphy ch. XI...’ (fo make matters worse, the chapter in question is X.) ”? Alan Donagan, “The Verification of Historical Theses’ Philosophical Quarterly, vi (1956), 193-208.

8 Ibid., p. 198; quoted from A, 144f. 9 Donagan, op. cit., p. 198.

COLLINGWOOD ON THE CONSTITUTION OF THE HISTORICAL PAST 325

discussion, Donagan says the following: “To re-think significant past

thoughts is part of the end an historian strives to accomplish; it 1s not even the whole of it, for he must also both demonstrate that he has re-thought them, and use them to explain past actions.” The first of these two additional tasks means simply that the historian must take pains to ensure the reasonableness of his claim that rethinking of past thought has actually taken place and that he is not imposing upon past actors what is simply his own conception of what

might have been or ought to have been thought. Collingwood was himself very much aware of this danger and discusses it in his ac- __ count of re-enactment in The Idea of History IH, 292 and 296). But it is the second of the tasks which interests us here, and it is unambiguously clear that Donagan believes that rethinking is used in historical research in order to explain historical events. If he could show that this was the case, the consequence for what I have been saying

so far would be very grim indeed. On the one hand, what we have seen to be Collingwood’s conception of an historical action, as distinguished from what he takes to be merely an event, would be very hard to accommodate to what Donagan would have succeeded in demonstrating. And, on the other, the three examples presented in the previous section of this paper would perhaps be in jeopardy, for if re-enactment is, indeed, a technique of explanation what I have been saying about it may turn out to be all wrong. Thus, it becomes obviously imperative to re-examine the chapter on art in order to determine for ourselves whether Collingwood’s problem there is one of explanation or, as I incline to suspect, one of the actual constitution of the historical past. Of the situation to which he addresses himself, Collingwood says: ‘At its lowest terms, the history of Romano-Bnitish art can be told in a couple of sentences. Before the Roman conquest the Britons were a race of gifted and brilliant artists: the conquest, forcing them into the mould of Roman life with its vulgar efficiency and lack of taste, destroyed that gift and reduced their arts to the level of mere manufacturers.”*' This rather terse account, rather more judgemental than descriptive, is followed by an attempt to tell us something about the sorts of thing one sees when examining examples of the art-work of

the periods being considered, but that will not actually concern us ® Ibid., p. 199. *' ‘Roman Britain’, p. 247.

326 COLLINGWOOD ON THE CONSTITUTION OF THE HISTORICAL PAST

here. Perhaps it should, but we have no alternative to ignoring it short of digging up the reports cited in Collingwood’s footnotes, for his account is not graced with any sort of plate, figure, or illustra-

tion. In fact, nowhere in the entire volume is such a thing to be found, surely a serious defect in a work so much concerned with and heavily dependent upon the results of archeological exploration. In any event, since Collingwood does not present the problem with which

he is here concerned in any but judgemental terms—good art and bad art—and has not provided us with the means of doing anything about it, the discussion which follows will simply continue to use his own manner of presentation. So far as our present problem is concerned, namely, to decide what Collingwood himself is attempting to accomplish in his discussion, it really does not matter. His problem

is set by the situation characterized in the brief passage quoted in the beginning of the present paragraph with one additional matter: ‘we find the same Celtic art which disappeared at the beginning of the Roman period rising mysteriously from its grave, enfeebled and uncertain of itself, but unmistakable, when that period is over.”*? One

would have expected that after a period of more than three centuries for which archeology has not uncovered any evidence of the preservation of old Celtic art that kind of art must have been fairly well extirpated; yet, we are now told, that as soon as the Romans left Britain it begins again to find expression. ‘Where have its seeds been preserved? And why have they been hidden so completely for three centuries and a half?’ Immediately after this, Collingwood sums up the perplexities which confront him in the form of two questions to be answered: ‘Why did anything so well established and

well developed as British La Teéne art fail to survive the Roman conquest?’ And since ‘the artistic talent that had produced it was not

extinguished: why did it not turn its powers to the production of works in the Roman style’ and produce a local school of art in that style of a better sort than what Collingwood thinks actually was produced? I have been quoting somewhat repetitiously, but not inadvertently.

As I have been making the distinction, I want to argue that Collingwood’s task in the present chapter is not to explain what happened but to tell us what in fact did happen, yet it might well 2 Ibid., p. 250. °° Tbid.

COLLINGWOOD ON THE CONSTITUTION OF THE HISTORICAL PAST 327

appear from the language in which he sets his problem that explanation is what he is after. Indeed, just before the passage quoted above about the re-emergence of Celtic art-forms after the departure of the Romans, he uses the expression: “he thing demands explanation.” That, and the frequent use of ‘why’ in the passages quoted all make it seem that I must surely be mistaken in insisting that expla-

nation is not the problem here, and I would not wish anyone to think I was not aware of Collingwood’s language, much less that | hoped to keep attention away from it. he language is plainly there. It is a straightforward language not easily avoided—short of some effort at conscious circumlocution—in expositions of the sort we find in works like ‘Roman Britain’. Yet, for all that, I think that when we look closely at what Collingwood actually does in the attempt to answer his questions, we find historical constitution and not explanation. As in all instances of historical reconstruction and fully in accord with Collingwood’s explicitly stated views, the occasion for re-enact-

ment or rethinking of the relevant past thought is provided by a body of historical evidence. In the present case, this is mainly the remains of artistic work, and Collingwood tells us about what has been discovered and something about its geographical distribution. In addition, of course, he tells us about the sudden disappearance of Celtic art-forms after the Roman conquest and its subsequent return after the Romans depart. With respect to the question of why the seemingly gifted Celtic artists did not continue to produce fine work albeit in a Roman or Romano-British idiom, Collingwood notes that artistic talent is not something biologically transmissable, but, rather, what is handed down over the generations is a tradition of work in a certain way and in a certain field. We have, he thinks, to do with a tradition and not with an artistic school, and he draws this distinction in the following way: The continuity of a school is a conscious continuity: it depends on one person’s teaching another, explicitly, what to do and how to do it. The continuity of a cultural tradition is unconscious: those who live in it need not be explicitly aware of its existence. The continuity of tradition is the continuity of the force by which past experiences affect the

future; and this force does not depend on the conscious memory of those experiences.”

* ‘Roman Bnitain’, p. 252.

328 COLLINGWOOD ON THE CONSTITUTION OF THE HISTORICAL PAST

I think that this conception of the unconscious of tradition, quite apart from its role in understanding the survival of the Celtic manner of art through the centuries of Roman domination of Britain, is very important for the light it sheds on Collingwood’s own conception of thought, as I shall indicate a little futher on. But first, I wish to follow Collingwood’s attempt to pursue his problem to its first major climax. This can best be done in Collingwood’s own words, for while it means I must quote at length, I doubt that I could summarize the points I believe are embedded in his text in any better or clearer fashion. After his remarks about tradition, we find the following: The British people had achieved its first great artistic experience through the intensely abstract curvilinear design of the middle La Téne school.

It was bound, therefore, to respond to contact with a new art in one of two ways. If that new art provided opportunity for the perpetuation and development of this particular experience, it would welcome it, converting the traditional motives of the new art into means for continuing that line of artistic growth to which it had already committed itself. If no such opportunities were offered, 1t would accept the new motives, if forced to accept them, in a dull and uncomprehending man-

ner... and do stupidly and blunderingly what it was told to do, not from stupidity, but because it was preoccupied with its own thoughts betraying this preoccupation from time to time by doing work, crude

and childish, no doubt, quite unlike what it was wanted to do, but expressive of its own desires, and therefore bearing the stamp of con-

viction. The one thing it could never do is to behave as if its own great experience had not happened to it, and learn its new lesson with an open mind. A people, like a single human being, is what the past has made it. Of these two reactions, the first would make for healthy artistic de-

velopment.... The British people would have incorporated the Roman tradition in its own artistic experience, and would be able to go on to the next stage in its development. The second would be in the main a waste of tme...; Roman art teaching the Briton nothing but what he was glad to forget, Britain contributing to Roman art nothing of which it could be proud. After the end of the contact, Britain would be left to take up her own artistic problem where she had been forced

to lay it down.”

While I have been trying to discuss the role of the re-enactment of past thought in historical research, I have not attempted to offer an account of what Collingwood takes thought and thinking to be. And, % Ibid, p. 253.

COLLINGWOOD ON THE CONSTITUTION OF THE HISTORICAL PAST 329

obviously, so difficult a task could not be undertaken as a parenthe-

sis to something else. But I should like to suggest that what Collingwood is doing in the paragraphs just presented is a form of the re-enactment of past thought, though it obviously differs from the examples of re-enactment presented before. In those examples, it order to know what the historical action was which left the evidence which occasioned the historian’s rethinking, we found ourselves considering the thoughts of individual agents whose intentions, plans, and purposes were being carried out on the historical stage. Here we have to do with thought—the ‘inside’—in some other way. I cannot

here work out the details of just what that other way is. But I do think it clear that what is involved is not human behaviour in the sense of what is merely the ‘outside’. As we have seen, Collingwood believed that the ‘continuity of a cultural tradition is unconscious’, but that surely in no way is intended to suggest that the carrying out of the tradition is mere behaviour uninformed by thought. There are meanings—in some relevant sense—which receive expression in the carrying on of a tradition, even if the self-conscious sort of reflection

we find in the working-out of a course of action, say, in military strategy, is not to be found. While, as I say, this is no place to work out the theory of the nature of thought implicit in Collingwood’s work, I feel certain that readers with any degree of familiarity with his thought will recognize that this must be so, that Collingwood would never turn over the ‘continuity of a cultural tradition’ to the tender mercies of the behaviourists. I think it is reasonable to suggest that any effort to work out Collingwood’s conception of thought entirely from his philosophical writings alone without careful exami-

nation of passages such as those just quoted is no more likely to yield satisfactory results than the attempts to construe his sense of re-enactment from The Idea of History alone without attention to the actual examples of that procedure with which his ‘Roman Britain’ abounds. Now let us see what Collingwood has done in the passages quoted.

He has tried to determine the character of the course of RomanoBritish art by attending to the alternatives the situation offered to a tradition which suddenly confronts or is confronted by a new situation. To this extent, there is a resemblance to the sort of rethinking we saw above since in those cases, too, the point of departure for the course of action is the need of a general or an emperor to deal with a situation he confronts or confronts him. Collingwood knows

330 = COLLINGWOOD ON THE CONSTITUTION OF THE HISTORICAL PAST

on the basis of the archeological record what sorts of art followed in what sequence. And so he is able to determine which of the alternatives he specifies in the first of the two quoted paragraphs was followed. He started with archeological remains. But after the course of his re-enactment of past thought, he is able to tell what happened, what sort of human action the history of art in the period discussed was. What happened was what is characterized as the second of the two ways in which an art tradition may respond to a new situation.

The Roman tradition was not one, he tells us, to which the traditional practices of Celtic artists could be adapted successfully. And so

the Celtic artists adopted the Roman style—apparently there was nothing else to be done—but could never really work well at it. The art that was closest to their hearts would be worked at occasionally, apparently with scant success. But that enabled them to preserve a tradition, to which, when times were again propitious, they could devote their full energies. With that, of course, we get the resurrection of the old Celtic tradition which had neither died out completely nor was something embedded in their genetic constitution. Anyone who wants the details of this development will have to consult the I last half of Collingwood’s chapter. I believe, however, that what have just sketched is a fair account of what he attempted there to do. Thus, in the sense that is usually intended, Collingwood has not offered an explanation of anything in that chapter he would leave as

a monument to his Romano-British studies. Rather, what he tries there to do is present an account of what took place, based upon evidence and using the method of re-enacting past thought. To deny this, one must ask oneself what it is that Collingwood has been explaining. In light of what I have already said when this sort of thing came up earlier in our discussion, the answer is clear enough: if this chapter does in fact contain an explanation it would have to be about the outside of the action taken by itself. The re-enactment could be said to provide an explanation only if outside and inside could stand

alone, each independent of the other. But that is precisely what Collingwood, in his conception of human action, denies. Until the historian has gone through the course of re-enactment, he knows what evidence has survived the depradations of time, but he does not know what human actions have taken place. With re-enactment, we get a conception of what human action must have or might have— I need not choose between these here—taken place. It might be noted

COLLINGWOOD ON THE CONSTITUTION OF THE HISTORICAL PAST 33]

again, that one happy by-product of our discussion 1s a sharpened awareness that Collingwood’s conception of both thought and action

are not so narrow as some might suppose and that a considerably wider swath of history—actually, I think all of history—is quite com-

patible with his view of them. In any event, where we had a per-

plexing body of evidence, we now have a conception of what happened. Why that sort of thing could happen in a situation of the sort that existed, is something else again. IV

The final chapter of ‘Roman Britain’ tells the story of the gradual fading away of the Roman element in Britain. Saxon settlements had been established and were growing in strength. What was an outpost of Roman civilization had retrogressed into a Celto-Roman one, with Celtic elements increasingly prevalent. In the course of his account, Collingwood writes that, This backwash of Celticism over the romanized regions, both attested

and dated by the story of Vortigern,” is traceable by archeological evidence. At Silchester a tombstone was found, with an inscription in Ogams containing the name of a certain Ebicatos and written in the Irish, as distinct from the British, form of Celtic. An Irishman who died in Silchester and left friends able to make him an epitaph in his own language must have been a member of an Irish colony in the town. The Scotic raids, as we saw, had died away; the Silchester inscription shows a state of things in which parties of Scots*’ are settling down peacefully in the lowland zone, and retaining their own language and customs.”®

This is an interesting paragraph for a number of reasons. ‘To begin with, there are still numbers of people who think that the historian discovers what happened by reading old texts and copies out—to be

sure, In a critical spirit—what he finds in them. But there are no texts of that sort which tell about an Irish colony in Silchester in the fifth century. The fact is, the possibility that such a colony existed we owe entirely to the working of the historical imagination. And so °° A local Welsh king of the fifth century who was supposed to have invited Horsa and Hengist to settle in Kent. *” “The Scots... were the inhabitants of Ireland’; ‘Roman Britain’, p. 282. °8 “Roman Britain’, p. 316.

332 | COLLINGWOOD ON THE CONSTITUTION OF THE HISTORICAL PAST

we discover still another kind of passage which ought to be mastered

for the light it may throw on still another of Collingwood’s wellknown conceptions, that of the historical imagination, instead of the more usual practice of trying to tease out its meaning from the text of The Idea of History alone.” Too, a full appreciation of what the quoted paragraph—and many others in ‘Roman Britain’, and the body of historical literature altogether—does would make it evident that historical realism—which I take to be the view that statements about past events mean and refer in the same way that statements about present events do and that, in consequence, the correspondence theory of truth is at least plausible for history—cannot possibly do justice to the epistemological idiosyncracies of history. Given the way in which the account presented in the paragraph is introduced— about which I shall have more to say in the sequel—it clearly does not have the same epistemic status as the report of a witness or the description of something which lies before me in my field of vision.

Of course, some will want to say that if the account is true then anyone who might have been present at the time would see certain things and could describe what he experiences. Be that as it may, and without raising the question of what it is one would see or ex-

perience—being an Irish colony being rather a different kind of being than being a red chair*’—the fact of the matter is that as the event is constituted in historical research it is not something witnessed and the verification or confirmation of the claim that such a colony existed in Silchester is not done in such a way as to implicate the correspondence theory. The claim is historically true if it makes best sense of our evidence. To assert that it is historically true only if in fact the real past corresponds to what it asserts is to assert a criterion of historical truth which can have no application in the practice of history.*' * ‘The historical imagination’ is simply Collingwood’s way of referring to the techniques and autonomy of history in the reconstruction of the historical past. It is surely not a category in any sense; cf. the contrary opinion of Brevard S. Childs, Memory and Tradition in Israel, Studies in Biblical Theology No. 37 (London, 1962), e hs ‘Mos of the philosophical discussion of statements about past events makes the tacit assumption that the two are rather the same. Were the assumption made explicit and then rejected, no small amount of the difficulties we find in such discussions would disappear. *! Gf. my ‘Evidence and Events in History’, Philosophy of Science, xxix (1962), pp. 175-94 (in this volume pp. 3-27); Pt. L.

COLLINGWOOD ON THE CONSTITUTION OF THE HISTORICAL PAST 333

But there is still a third way in which the quoted paragraph is of interest, and I should like to devote some space to that, namely, to what it indicates about the relation between the historical event there described and the historical evidence upon which it is based. In the present case, it is rather easy to deal with this because Collingwood specifies just what his evidence is—a single tombstone in Silchester. Of course, a single tombstone without any sort of context might not

get an historian too far, and a really full account of the relevant evidence would presumably have to include all of the evidence on the basis of which the general circumstances within which the tombstone is located was established. But we need not bother about that

here. It does seem plain enough that Collingwood’s view that an Irish colony existed in Silchester in the fifth century is based entirely upon the discovery of the tombstone. An entire Irish colony seems

like rather a weighty structure to erect upon so frail a foundation, and if there were any truth at all to the widely held opinion that the historical past is inferred from historical evidence it is hard to see how Collingwood’s Irish colony could at all be taken seriously. But clearly,

there is never a relationship of logical entailment—and what else could ‘inferred’ mean?—between historical evidence and _ historical events. Rather, the event is postulated by the historian as most likely

given what evidence he has and what general information he has about the period. After writing about King Arthur and his military campaigns, Collingwood says: These are conjectures. But they are based on the facts of fifth-century warfare; and the probability that they may be correct is at least slightly increased by the fact that the traditions which first appear at a much later date, and are embodied so far as we are concerned in the twelfthcentury romance of Geoffrey of Monmouth, revolve around the conception of Arthur as the creator of a band of knights.*

Why that should be so is of no interest here, and anyone who 1s curious about it may look it up for himself. The point to be made in the present context has to do with the conjectural character of the

*” Yet it is surprising how far the skilful application of the historical imagination can go with very little evidence. See, for example, Hjalmar R. Holand, Norse Discoveres and Explorations in Amenca, 982-1362 (originally published as Westward from Vinland: on Mecount of Norse Explorations and Discoveries in America, 982-1362) (New York, 1969),

* ‘Roman Britain’, p. 323.

334 | COLLINGWOOD ON THE CONSTITUTION OF THE HISTORICAL PAST

event Collingwood had been proposing or postulating before the passage quoted, an event based upon some evidence and rendered somewhat more plausible than the evidence itself might suggest by

its conformity with what is otherwise known about fifth-century warfare. Although the interest that brings both scholars and readers to history

is the human past, a curiosity about what happened in days gone by, in terms of its strict logical function in the context of historical research, the purpose of the historical event is to explain historical evidence. Collingwood tells us that an Insh colony existed in fifthcentury Silchester because in his view only that renders intelligible the presence there of a tombstone inscribed in the Irish form of Celtic.

Even if one must admit that interest in the period or some such thing led Collingwood to his studies of the Roman period of British history, surely no one can say that he was particularly interested in that Irish colony and for that reason paid some special attention to the remains of Silchester. Quite the contrary! He had no interest in

such a colony at all until he had virtually to call it into existence in order to explain the tombstone. No serious reflection on the genesis and function of the statement, ‘An Irishman who died in Silchester

and left friends able to make him an epitaph in his own language must have been a member of an Irish colony in the town’, in the context in which it appears in ‘Roman Britain’, leaves room for any alternative view of its epistemic status. It is all well and good to say that Collingwood’s statement is true only if there really was such a colony, but that is to say something that has no consequence for historical inquiry; it simply expresses the hope that the historical past is identical with the real past. But Collingwood himself had no illusions about the relevance of that hope to the methodology of historical practice: “The game is won not by the player who can reconstitute what really happened, but by the player who can show that his view

of what happened is the one which the evidence accessible to all players, when criticized up to the hilt, supports.” The point, of course, is entirely general. The logical function of any constructed historical event, understood strictly in terms of its ** Here the reader may add any qualifications he deems necessary to avoid having an apoplectic sputter. * Collingwood, Essays in the Philosophy of History, ed. W. Debbins (Austin, Tex., 1965), p. 97.

COLLINGWOOD ON THE CONSTITUTION OF THE HISTORICAL PAST 339

function in historical inquiry and without reference to the hopes and interests of both historians and those who read what they produce, is to explain the body of evidence deemed relevant to the investiga-

tion being undertaken. In particular, all of the examples we have considered earlier in this paper do the same thing. If Collingwood is right about Celtic art, he has explained the remains that archeologists have uncovered. That is, he has explained to us why it is that the art remains of the various periods of British history involved have the character they have, why excellent work in one style should have

been succeeded by inferior work in another style, and why, when there seemed every reason to think the first style had vanished completely, instances of it should suddenly appear once again. Likewise, to mention only one more, if his account of the Antonine Wall 1s

sound, Collingwood has explained why it is the wall has certain characteristics, that is, why the surviving evidence has the features it

does. But in all of the examples we considered above, the central technique used by Collingwood in his attempt to constitute the historical past was the re-enactment of past thought. Since the re-enactment of thought results in historical constitution, and historical constitution has the function of explaining evidence, it would seem that my insistence that the re-enactment of past thought is not done for the purpose of offering historical explanations was mistaken after all. Not really. It is clear enough that to talk about explaining histori-

cal events and explaining historical evidence 1s to talk about two rather different things. I do not wish to review the literature here, but I think few will argue with the assertion that those critics of the covering-law conception of historical explanation who have claimed to be working in the spint of Collingwood—taken as loosely as you

like—have usually meant to say that Collingwood’s theory of reenactment or rethinking was a way of taking seriously the relevance of rational intentions in the explanation of historical events. Such critics generally contend that the covering-law conception is incapable of doing justice to explanation by rational intention. In addition, since the covering-law position is virtually always presented as

a theory about the nature of the explanation of events—the pecularities of a queen’s signature or the migrations of farm workers from blighted to better land—one expects that a critical stance in opposi-

tion to it is directed at the same sort of problem. It is hard to avoid the conclusion that it is the general lack of interest among philosophers of history in the intellectual procedures whereby the historical

336 COLLINGWOOD ON THE CONSTITUTION OF THE HISTORICAL PAST

past is constituted in historical research, and their virtually exclusive preoccupation with the nature of the explanation of historical events, that have made it so difficult for the true purport of Collingwood’s idea of the reenactment of past thought to be understood. But, surely, it is by now obvious that the constitution of the historical past and the explanation of past events are two different concerns of historical inquiry. And while each involves some kind of explanation, they are markedly different in kind. While the re-enactment of past thought is central to the constitution of the historical past with its function of explaining historical evidence, that it is to be used in the explanation of past events is something else again. So far as Collingwood is concerned, we have discovered three bases for claiming that this last is not the position to which he subscribed: (1) it is not compatible with his conception of human action; (2) it is not the way he actually uses the technique in ‘Roman Britain’; and (3) we have found a perfectly

good instance of the explanation of an historical event in ‘Roman Britain’ that does not make use of re-enactment.

THE IDEA OF HISTORY AS A SCALE OF FORMS'!

I

Since the view [intend to put forward of Collingwood’s concept of history requires that the structure of The Idea of History be taken seniously—that is, had Collingwood lived to bring to conclusion the work he was trying to produce on the principles of history, he would have

emerged with a work essentially like what we have as far as its organization is concerned—I must begin by commenting on the statement of its editor, the late Sir Malcolm Knox: “I think it nght to say that although the layout of the book and some of its forms are due to the editor, the content is everywhere Collingwood’s” (WH, v). The main point of possible contention would concern the relation of Part V, “Epilogomena,” to the earlier four parts. hose four parts purport to be an historical account of the idea of history, and it is hard to see how Knox could have done anything about their order. Surely, the Greeks have to come first, the renaissance historians before the emergence of scientific approaches in modern times, and so on. Part IV, “Scientific History,” is presented in the volume we actually have, not strictly chronologically, but according to national cultures, beginning with England and ending with Italy, and perhaps there was room for Knox’s judgment in dealing with that. Inasmuch as that will have no role to play in my presentation, I shall not bother about it. ‘The materials collected in the “Epilogomena,” however, are not part—

or so it seems—of an historical account of the concept of history. They, rather, contain discussions of the elements which enter into the practice of history as Collingwood took that to be in his own time. We have discussions of evidence, of rethinking past thought, of the historical imagination, and so on. Perhaps that material was not seen by Collingwood as belonging with Parts I through IV, and it is only because of Knox that they are included in the same volume. Since we are told by Knox that the “content” of the book is everywhere Collingwood’s, I presume that we may take seriously what ' Page references given in the text preceded by “EPM” are to R. G. Collingwood, An Essay on Philosophical Method (Oxford, 1933); those preceded by “JH” are to The Idea of Fitstory (Oxford, 1946).

338 THE IDEA OF HISTORY AS A SCALE OF FORMS

Collingwood himself tells us, that he intends to present us with an account of the modern idea of history but first he intends to “cast light” on that idea by presenting us with an account of its history (LH, 10). Thus, it would seem that whatever the volume we have owes to

the labors of Sir Malcolm Knox, the structure of it is what we might have expected had Collingwood been able to put it together himself. Is there a principle which guides the construction of The Idea of FAistory? If 1 am night in thinking that it is an attempt to trace the stages through which the concept of history expresses itself as a scale of forms, what we must expect to find in it is the variable expressions of the generic essence of history. In his account of the scale of forms in his Essay on Philosophical Method, Collingwood argues that the

generic essence of the sort of concept he is dealing with is the variable which changes. In his opinion, he is attempting there to characterize philosophical concepts and to point to ways in which they differ from concepts which are not philosophical. Sometimes, he allows, the same concept may have a philosophical and a non-philosophical use, and in such cases he is characterizing the concept in its philosophical use. In all that, as I shall suggest, he is mistaken. Rather, he is characteriz-

ing certain concepts which carry their histories with them, which undergo—for whatever reason—determinant changes during the course of which they are nevertheless recognizable as the same concept though markedly different at different stages of its history. Collingwood makes

it a matter of principle that if we can discern in a concept that it has the features he characterizes in chapters two through four of The Essay on Philosophical Method it is—by definition—a philosophical con-

cept; but, I fear, that is merely dogmatic. I could not be persuaded that the concepts of freedom, parliamentary government, romantic poetry, and baroque music are philosophical concepts at all, yet it seems to me that every one of them is a scale of forms in which the generic essence undergoes changes and in which difference in degree

is accompanied by difference in kind (LPM, 54-57). And while Collingwood wants particularly to distinguish philosophical concepts from those of the natural sciences (EPM, 26-31), I suspect that one knowledgeable in the history of those sciences would have little difficulty in finding scientific concepts which could satisfy Collingwood’s characterization of philosophical concepts so-called. I suspect—though I cannot prove—that Collingwood’s attempt to present his analysis as a distinguishing of philosophical from non-philosophical concepts, particularly scientific concepts, 1s owing to the time he was working

THE IDEA OF HISTORY AS A SCALE OF FORMS 339

on it. It was a time during which the spirit of tough-mindedness in philosophy seemed ascendant, when claims were made that only concepts able to satisfy certain rigorous sensationistic criteria could be accepted as having meaning. In the face of that, Collingwood sought

to protect the standing of philosophical concepts by insisting that they were different from others and need not be subject to positivist strictures. What is the generic essence of history? If there is anything to my claim that The Idea of History presents the idea of history as a scale of forms, one who reads the book ought to be able to discern such a generic essence and trace the course of its change, and, indeed, see how its change of degree leads to a change of kind. The reader would have to have extraordinary powers to find such a thing. It is not there. In one sense, it must be said that Collingwood does not succeed in carrying out the project. But, in another, it would have

been too much to have expected him to do so since I rather doubt that it was possible for anyone to have done so in the 1930s; it would not be all that easy today—half a century later—though it would be considerably easier. Yet the question might easily be raised:

how is it that I claim that the idea of history is a scale of forms in the sense that Collingwood characterizes in The Essay on Philosophical

Method, given that Collingwood makes no such claim; that I admit that the project is not to be found carried out in The Idea of History; and that I say that the concept of history is not a philosophical one, the only kind of concept Collingwood took to be a scale of forms? Although Collingwood does not, for reasons I shall soon indicate, carry out what I take to be the intended project of The Idea of History,

he does tell us what its generic essence is. At one point he tells us that he wants to produce “a philosophical inquiry into the nature of history regarded as a special type or form of knowledge with a special

type of object” (ZH, 7), and he will go on to say “that history is a kind of research or inquiry” which “consists of fastening upon some-

thing we do not know, and trying to discover it” (JH, 9). Thus, the generic essence of history is research, and an account of the idea of history as a scale of forms should proceed by attending to the history of history-as-inquiry showing at each step of the way how change-— may I presume to say progress?—in the history of the techniques of historiography results in the constant transformation of the discipline

and its character, that changes in degree result in changes in kind. Although part I of The Idea of History begins with mythic and theo-

cratic ideas of the course of human events, if there is any substance

340 THE IDEA OF HISTORY AS A SCALE OF FORMS

to what I have been saying, the real beginning of Collingwood’s account is his treatment of Herodotus and Thucydides, because only with those founders of history-as-discipline does Collingwood begin to tell about historical method. There were, Collingwood tells us, tendencies of Greek thought which were inimical to the possibility of history. Epistémé was possible only with respect to the unchanging,

but history deals with the particular that changes. That begins to change with Herodotus and ‘Thucydides. ‘They begin the development

of methods whereby the human past may be known. Not all that much of the human past: their method was limited to the questioning of witnesses, which means that an historical account of their grandfathers’ generation was inconceivable. Collingwood writes: “Scientific

history has been invented. Its field is still narrow; but within that field it is secure” WH, 26). I do not wish to get involved with the details of the early history; I want only to note something he says about Hellenistic historiography and draw from it one observation. By the time we reach the Hellenistic epoch, those who practice at being historians have at hand the work of their predecessors. They need not limit themselves to the questioning of witnesses—though I suppose they may still do that—but, in addition, may extract material from the writings of those predecessors which they may incorpo-

rate into their own account. The method of scissors-and-paste has been invented (JH, 33).* That is not only a new method to be added to the old method; it, rather, effects a fundamental transformation of the discipline which is no longer restricted to the limits of memory. A new method has been added—a difference of degree; and the discipline has been transformed—a difference in kind. Thus, Collingwood tells us that history is research, and very early on in his account of zs history we get to see what he means. But from his treatment of Greek and Hellenistic historiography until we are presented with his idea of the contemporary practice of history, in the “Epilogomena,” there is very little of this kind of thing. For ? The question has been raised with respect to what I say here, “does not his treatment of ‘scissors and paste’ history wind up with the conclusion that it is not

really history at all, but rather a perversion of it leading only to a dead end?” Should anyone presume to practice that sort of history in our time, that would certainly be the case. Yet when it first appeared in the history of historiography it represented progress, an improvement over history that could only cross-question witnesses. If we look at Collingwood’s treatment of moral ideas in his Essay on Philosophial Method we find the same thing. Something which 1s progressive when it appears seems retrogressive from a later perspective.

THE IDEA OF HISTORY AS A SCALE OF FORMS 34]

the most part, we are presented with sketches of conceptions of history as other than research. For Collingwood to refer to the German poet Schiller and Michael Oakeshott as historians 1s, surely, an expression of despair, and he certainly has nothing to tell us about the methods whereby they come to know anything at all about the human past. As I read The Idea of History, Collingwood would hike to offer an

account of the development of the idea of history-as-research but does not. And he does not because he cannot. Why? Because the scholarly work which would have to have been done before an account of the sort Collingwood would have liked to have written was simply not in existence. It would have been entirely different if Collingwood

could have had available to him work such as Pocock’s work on seventeenth-century English historiography,’ Huppert’s* and Kelley’s’ treatments of sixteenth-century French historiography, Iggers’s work on modern German historiography,° Breisach’s general overview of the subject,’ and all manner of other scholarly studies on the history of historiography which have been published over the decades since

Collingwood’s death.’ I would say that Collingwood had a sense of the idea of history, and knew what was necessary to show that its generic essence took new shape over the course of its history, but lived before the material needed for it to be possible was available.” Ii If the above discussion makes it seem reasonable to believe that the idea of history is a scale of forms the generic essence of which is mani-

fested differently in each of the different shapes the idea comes to * J. G. A. Pocock, The Ancient Constitution and the Feudal Law (New York, 1967). * George Huppert, The Idea of Perfect History (Urbana, Chicago, and London, 1970). ° Donald R. Kelley, Foundations of Modern Historical Scholarship (New York, 1970). ° Georg G. Iggers, The German Conception of History (Middletown, Conn., 1969). ’ Ernst Breisach, Historiography: Ancient, Medieval, GS Modern (Chicago, 1983). ® It has been suggested that in Collingwood’s time there had already been important writing on the history of history and that I might be exaggerating the claim that the material was not available to produce an account of the sort I say Collingwood could not

have produced. Yet the kind of detailed study of the history of history-as-inquiry that we find in the studies I cite in the preceding notes, and the quantity of it, did not exist, and I continue to believe that what I claim in the text is essentially correct. * Of course, I make no mention of studies of the Annales school’s methodology, cliometrics, and other recent forms of historical scholarship, inasmuch as there is no way that Collingwood could have been concerned about them.

342 THE IDEA OF HISTORY AS A SCALE OF FORMS

have, one must conclude, so Collingwood would have it, that the concept in question is a philosophical one inasmuch as he takes it that it is that which distinguishes a philosophical concept from a nonphilosophical concept or the philosophical use of a concept from its non-philosophical use. While Collingwood is most insistent about that,

I cannot say that I find him persuasive about it. Indeed, he seems more to assert it than to demonstrate it, and I rather doubt that he could have demonstrated it if he wanted to. I said earlier on that I thought that what really interested Collingwood was a desire to defend philosophy—in some traditional sense—from its hard-nosed tough-minded critics in what was surely a period of ascendancy of positivist criticism; but that ascendancy is long past, and we need not concern ourselves all that much about it. I think that in his account of concepts in An Essay on Philosophical Method, Collingwood has im-

portant things to say about concepts of certain sorts, but they are not merely philosophical concepts. But before I go on to specify what

I have in mind, I should like to make some brief remarks about other concepts which I take to resemble the idea of history in the respect dealt with above, yet, in my view, cannot be considered to be philosophical concepts. I want to begin with the concept of freedom. I can imagine that some will quickly assert that freedom is surely a philosophical concept, a concept that belongs to political philosophy, and if I can show that

it resembles the idea of history in the specified way that will surely underscore the point. But while freedom belongs to the sphere of politics,

it is not obvious that it is a philosophical concept. We can think philosophically about all manner of things without those things becoming philosophical. Freedom has to do with the possibilities afforded people to act in ways they choose, and the possibilities of acting and

the sphere of actions are human and social, not inherently philosophical. And I shall stand by that view, even though I intend to discuss freedom from the perspective of Hegel’s history of freedom in a work misnamed Lectures on the Philosophy of History. 1 shall do so

because I find it convenient, not because I have any commitment to its correctness. A history of freedom by an author who died in 1831 can hardly be the last word and, anyway, I have my doubts that the swath of space and time dealt with in Hegel’s lectures constitute a unit which has a history. One occasionally comes across a wooden formula which attempts to sum up Hegel’s account of freedom by noting that in China one

THE IDEA OF HISTORY AS A SCALE OF FORMS 343

was free, in Greece some were free, and in the modern world all are free (see JH, 125). That would seem to suggest that in the course of its history freedom is manifested by differences in degree, that the difference between the expression of its generic essence in one time and place and another time and place is a difference of degree. But this is clearly not the case. As Hegel moves from one historical stage to the next—and, in spite of the formula, there are more than three— he indicates not only that increasing numbers of people are free, but the way in which the institutional and conceptual changes he describes

make it possible for increasing numbers to be free. This makes it clear that from one stage to the next the very character of freedom has undergone change, which means that the change in degree is accompanied—indeed made possible—by change in kind. An account of the history of parliamentary government should show the same pattern. Indeed, I rather suspect that such a history would be part—a central part—of the history of the growth of political freedom

in Britain, which would, I think, lend force to my claim that freedom is a political idea, not a philosophical one. In any event, let us sketch some of the main features of the modern history of the idea without going back to the origin of it where the idea of parliament is not even visible at all. In recent centuries one notices two major tendencies. One is the long-term struggle for supremacy between the monarchy and the parliament, particularly the House of Commons, during the course of which more and more power devolves upon the latter. [he other is the slow broadening of the franchise so that in-

creasing numbers of citizens may participate in the governmental process, culminating in the enfranchisement first of Roman Catholics and then of Jews, the movement towards universal manhood suffrage

during the course of the nineteenth and early twentieth centuries and, finally, the extension of the franchise to women. It will be noted that in both of these movements I have stressed the quantitative aspects—more power, more voters—yet it 1s clear that with these advances there is a transformation of the very character of parliamentary government; the generic essence of it is manifested in different ways and is, in Collingwood’s terms, the variable that changes. Essays on the nature of parliamentary government—on the concept of parliament—written in the middle of the seventeenth century, in the

middle of the eighteenth century, in the middle of the nineteenth century, and in the middle of the twentieth century would each characterize the generic essence of the thing in terms of numbers

344 THE IDEA OF HISTORY AS A SCALE OF FORMS

and quality, precisely what we would accept given Collingwood’s

characterization of concepts which are not sharply defined. | Although I feel less comfortable speaking about it, I would venture

to suggest that we would find a somewhat similar situation should we turn to the world of the arts. According to John and Elizabeth Lowe, “typically a cultural pattern will follow much the same trajectory: a period of accelerating growth, followed by saturation, and finally exhaustion and eventual dissolution as most of the possibilities

inherent in the pattern are used up.”'° Think of a new development in the history of one of the arts, the beginning of romantic poetry or baroque music. Some one or more creative people begin to work in the new way, although it is not likely that that could be observed at the tme and the beginning is more likely to be discerned only retrospectively. In the course of time more and more of the generic essence of this new mode of creativity is expressed in the work of those who come to create in that way, and as this happens that way of working is transformed: as the generic essence of romantic poetry or baroque music is more dominantly expressed, those ways of working undergo change. In tme, as the Lowes suggest, the possibilities offered

by these ways of creating are exhausted and artists go on to seek new means of expression. And as in the case of the concepts Collingwood discussed in the Essay on Philosophical Method, the variable

which changes is the generic essence. ITI

In the discussion of the scale of forms that Collingwood presents in the Essay on Philosophical Method, the point being developed is made

clearer to the reader through the presentation of examples. Thus, he talks about good (LPM, 78). Virtue, knowledge, and pleasure are all taken to be good, but each is good in its own way. Thus, when we

ascribe goodness to each of them we cannot mean that they are good in the same way. Pleasure is a good, but a lesser good than, say, virtue, and in being a lesser good it is different with respect to good (EPM, 79). I think that what Collingwood says about “good” is 10 John W. G. Lowe and Elizabeth D. Lowe, “Cultural Pattern and Process: A Study of Stylistic Change in Women’s Dress,” American Anthropologist 84 (September, 1982), 521.

THE IDEA OF HISTORY AS A SCALE OF FORMS 345

typical of the way he thinks about philosophical concepts. The concepts that I have been discussing, however, including what I take to be Collingwood’s own idea of history, differ from good as he construes it in one, to my mind, major characteristic. Those concepts, one and all, are concepts that carry their histories with them; their generic essences develop in the course of time, and one cannot understand what the concept means without understanding its history. A proper explication of the concept requires not a timeless analysis in terms of conditions necessary and sufficient for the concept’s application, but, rather, an awareness of how the concept has come to

have the shape it has. I do not want to say that this must be so for all concepts of which the variable is the generic essence. I do not want to challenge Collingwood on the idea of good or the idea of thought (EPM, 36ff.) or any other concept he deals with in the Essay on Philosophical Method. But 1 do want to say that there are interesting

and important concepts of which the generic essence is the variable in Collingwood’s sense which are virtually incomprehensible unless we construe them historically. Given both Collingwood’s extraordinary

sensitivity to matters historical and his interest in the arts, one may well wonder why Collingwood seemed not to notice the character of the kinds of concept I have been talking about and why, especially, he did not notice that the idea of history is a scale of forms in his sense of that idea, particularly since his account of it comes as near to illustrating it as a scale of forms as was possible given the virtual nonexistence of the required scholarly foundation. Of course, as we have seen, what seems to have motivated the Essay on Philosophical Method was the desire to protect philosophy and

its concepts from positivist attack. Whatever one might say about other concepts, there is something different—unusual—about philosophical concepts in consequence of which expectations with respect to other concepts are not to be looked for in philosophy, and failure of philosophical concepts to meet those expectations are not really failures. In addition, implicitly or explicitly, the tradition of science and philosophy has tended to agree with Frege that “a concept that is not sharply defined is wrongly termed a concept,”'! and so while

concerned to defend philosophy against its positivist detractors, Collingwood may have been entirely open to thinking that they may be right about other concepts. The interest that has been developing '! Quoted in Morris Weitz, The Opening Mind (Chicago and London, 1977), x1. ,

346 THE IDEA OF HISTORY AS A SCALE OF FORMS

in the openness of certain concepts is quite recent,'? and part of what I have been exhibiting in the paper is that Collingwood himself was a pioneer in the exploration of their logical nature—at least of the logical nature of some of them—even if he was not aware of it.!° But I have one more reason, one that may seem rather surprising both because it is said about Collingwood and said by someone known to be sympathetic to his thinking about history. As I was rereading the account of Hegel contained in The Idea of History, I came slowly to the conclusion that Collingwood’s concept of the history of ideas is rather defective. It will be remembered that what distinguishes the concepts that I offered as exemplifying the conceptual manifestation of the scale of forms from the examples discussed by Collingwood in An Essay on Philosophical Method is that they are historical in nature, that the variation in their expressions of their respective generic essences emerges in the course of their histories. If Collingwood does

not have an appropriate sense of the history of ideas, then it is not surprising that he should have missed the nature of concepts such as these. ‘here are other passages that might be cited, but let me quote one short paragraph from Collingwood’s discussion of Hegel: [Since all history is the history of thought and exhibits the self-development of reason, the historical process is at bottom a logical process. Historical transitions are, so to speak, logical transitions set out on a time-scale. History is nothing but a kind of logic where the relations of logical priority and posteriority is not so much replaced as enriched or consolidated by becoming a relation of temporal priority and posteri-

ority. Hence the developments that take place in history are never accidental, they are necessary; and our knowledge of an historical process 1s not merely empirical, it 1s a prori, we can see the necessity of it. IH, 117).

That this is simply wrong about Hegel need not detain us inasmuch

as we are not interested here in Collingwood on Hegel. What is more to the point is that it is wrong about the history of ideas and wrong in a way which could result in one who holds it simply miss-

ing the character of the concepts I have been talking about. And 2 See my “Reflections on Conceptual Openness and Conceptual Tension,” in Freedom and Ratonaltty: Essays in Honor of John Watkins, ed. I. C. Jarvie and Fred D’Agostino (Dordrecht, Boston, and London, 1989), 87-110. '° Just as his account of absolute presuppositions in his Essay on Metaphysics pioneered the reconception of the nature of science which becomes explicit with the appearance of Thomas 8. Kuhn, The Structure of Scientific Revolutions (Chicago, 1962),

and is reflected in work by Stephen Toulmin, Imre Lakatos, and others.

THE IDEA OF HISTORY AS A SCALE OF FORMS 347

while Collingwood is presumably discussing Hegel, he clearly subscribes to the view he is presenting (cf. ZH, 118). If Collingwood really means that history is a logical development,

that its stages are necessary, and that “the thoughts behind the events—not the events themselves—formed a chain of logically connected concepts” (/H, 118),'* it is no wonder that he was not sensi-

tive to the character of genuinely historical concepts, because it 1s simply not the case that history—political, institutional, or conceptual—develops as a series of deductively necessary stages. For deduc-

tion to be possible we require to have premises formulated in terms of sharply defined concepts, and the concepts we have been talking about, the generic essences of which emerge in the course of their histories, are not sharply defined. Each step is brought into being by the impact upon the previous stage of factors which are logically

contingent to it. The stages of a course of Hegelian dialectic are precisely of that sort and are not a prom or logically necessary at all, but it serves no purpose of the present discussion to deal with Hegel

here.'? Not even the history of history, to which Collingwood devotes The Idea of History, is like that. Each stage of the history 1s affected by the development of new techniques for constituting the historical past from the evidence at hand, and it makes no sense to say that each stage of the history is a deductively necessary consequence of the stage which preceded it. I suspect, though, of course, I cannot prove, that had Collingwood been more sensitive to the way in which the contingent pushes the development of concepts along and leads to the reshaping of their generic essence, he might have come to see that the sort of concept he actually discusses in An Essay on Philosophical Method is not the only kind of concept in which

the variable changes, and might have recognized that the idea of history 1s a scale of forms.

'* Ascribed to Hegel, but from the discussion it is clear that it is Collingwood’s Own VI1ew.

'5 But see my “Dialectic and Necessity in Hegel’s Philosophy of History,” in Substance

and Form in History, ed. L. Pompa and W. H. Dray (Edinburgh, 1981), 42-57; and “Force and the Inverted World in Dialectical Retrospection,” International Studies in Philosophy 20 (1988), 13-28.

BLANK PAGE

INDEX OF NAMES

Adorno, T.W., 144n2, 159 Eckstein, Harry, 45n6

Ankersmit, F.R., 235n6, 259 Elkins, Stanley M., 46-48, 50, 56

Ardant, Gabriel, 267 Elton, G.R., 66

Austin J.L., 174 Engerman, Stanley L., 251n16 and 17, Ayer, AJ., 174 270 Banton, Michael, 26n22 Feigl, Herbert, 257

Barker, S.F., 189 Feyerabend, Paul K., 257

Bayley, David H., 267 Finer, Samuel, 267 .

Beard, Charles A., 288n26 Flint, John T., 63n17, 65, 67n29, 70, Becker, Carl L., 106-118, 158n11, 256 72, 73n45, 74 and n47, 75 and

Beer, Samuel H., 39-41, 43, 49-52 nn48 and 49, 78, 264

Bendix, Reinhard, 262f Fogel, Robert W., 198, 251n16 and

Benson, Lee, 198, 251 17, 270

Berkhofer, Robert F., Jr., 198-199 Frege, Gottlob, 345 Bernal, Martin, 249f

Blegen, Theodore C., 186n17 Gale, Richard M., 102-105, 290n27

Bousma, William J., 65, 71, 73-75, Gallie, W.B., 33, 35, 42n4, 56, 86n5,

77, 78 235n6, 259, 299n38

Braaten, Carl E., 199-203, 205, 206 Gardiner, Patrick, ix, 4nl, 17 and

Bradley, F.H., x, 200 nl5, 35-38, 87n6, 209, 284n20, Breisach, Ernst, 341 285n21, 291n28, 306n48, 308n53

Brodbeck, May, 36n]1 Gaster, Theodor H., 95, 96n7, 179n12 Buchdahl, Gerd, 309n53 Giesebrecht, Wilhelm von, [1 Bultmann, Rudolf, 201 Goldstein, Leon J., xi, 114n4, 130n1, 140n17, 162, 164n20, 176n6,

Carnap, Rudolf, 100n16, 257 178n10, 190n23, 204n49, 216n22,

Carr, David, 247 230n31, 236n7 and 8, 238n11,

Childs, Bernard S8., 332n39 239n14, 254n2, 259n18, 266n37,

Cohen, L. Jonathan, 306n48 269f, 346n12, 347n15

Collingwood, R.G., xn, xi, 3, 133, Goodnough, Ward H., 23 137, 158, 178-179, 180-181, 190, Gorman, J.L., 150n5, 165n21, 166n22,

204, 209, 216, 219n27, 233, 209, 222-225, 227 237-238, 249, 273-347 Gutman, Herbert G., 252n17 Croce, Benedetto, 106, 107

Hall, John W., 197

Danto, Arthur C., 90n7, 98, 101, 161, Hanson, N.R., 43, 44, 192

209, 232, 247 Harris, Errol, 273n3, 276

De Grazia, Sebastian, 22 Hegel, G.F.W., 232, 342-343, 346-347 Dewey, John, 102-105 Heidegger, Martin, 201, 231, 232, 247 Dilthey, Wilhelm, 176, 295n36, Hempel, Carl G., x, x, 4nl, 15,

296n37 35-37, 40, 60n9, 98, 100n15, 257,

Donagan, Alan, 15n13, 17n15, 28-32, 260 33, 35, 278n11, 280n15, 324-325 Herodotus, 196, 197

Dray, William H., 4nl, 15, 16, 33, 35, Hill, Christopher, 84n2, 126n4

36, 55, 87n6, 209, 235n6 Hulgruber, Andreas, 251

Ducasse, C.J., 206n50 Hodges, H.A., 273n3

Durkheim, Emile, 22, 68, 69 Holand, Hjalmar R., 183-187

350 INDEX OF NAMES Hoschander, Jacob, 96n4 Murphey, Murray G., 119-128,

Hume, David, xi, 15, 200 133-135, 224, 229 Huppert, George, 174n2, 341 Husserl, Edmund, 247 Nagel, Ernest, 36, 38, 54 Needham, Rodney, 137n15 Iggers, Georg G., 341 Newman, F.D., 306n48 Nielsen, Kai, 36n1

Johnson, Allen, 6n4, 10, 11, 136n12, Nolte, Ernst, 251

177n8 Nowell-Smith, P.H., 143-167

Johnston, William M., 281n17

Oakeshott, Michael, 209, 341

Kant, Immanuel, 232-234, 238, 242, Oesterley, W.O.E., 8n6

244-245, 252 Olafson, Frederick A., 90n7

Kelley, Donald R., 174n2, 218n25, Oppenheim, Paul, 100n15 34]

Kepler, Johannes, 43 Pannenberg, Wolfhart, 201

Keynes, John M., 100n16 Peirce, Charles C., 128, 190 and

Kingdon, Robert M., 67n29, 71 n23, 191 and n24, 192, 216, 221, Knox, Sir Malcolm, 289, 337-338 225 Kuhn, Thomas S., 58n2, 59, 233, 246 Pfeiffer, Robert H., 7n5, 8, 96n6 Pirenne, Henri, 45

Langlois, Ch. V., 173, 174, 177, Pocock, J.G.A., 341

227-228 Pompa, Leon, 214-218

Lavine, Thelma Z., 233n5 Popper, Karl R., 4n1, 15, 17 and Lewis, Clarence I., 156n10, 220n30 nl4, 25, 38-40, 257

Lewy, Julius, 96n95 Porter, Dale H., 208 Louch, A.R., 315n6

Lowe, Elizabeth D., 344 Reeves, Richard, 248 Lowe, John W.G., 344 Renier, G.J., 306n48

Lundberg, George A., 59,60 andn1l0,61 Rhine, J.B., 206n50 Richardson, Alan, 202

Maclver, A.M., 5n3, 193-194, Rickman, H.P., 295n36

291-292 Robinson, James M., 201n43

Maier, Charles S., 79n59 Rotenstreich, Nathan, 63n15

Maitland, F.W., 38 and n2, 54, 55 Rousseau, Jean-Jacques, 254n3 Mandelbaum, Maurice, x, 15n11, 99, Rubinoff, Lionel, 324n26

100 and nl6, 101, 115n5, 132, Ryle, Gilbert, 28 150n5, 187n18, 210-213, 237-238,

243-245, 250, 252, 273, 307n50 Salmon, Wesley C., 189

Mannheim, Karl, 17n14 Schiller, Friedrich von, 341 Marrou, Henri-Irenee, 174n3, 176, Schneider, J., 194-195

177, 179, 194-195 Schonfield, Hugh J., 306n47 189n22, 218-222 Scriven, Michael, 14, 16, 17, 33, 35,

McCullough, C. Behan, 130n2, Schutz, Alfred, 60n8, 175

Mead, George H., 5n2, 104, 245f 37, 38, 87n6, 257 Mink, Louis O., 235n6, 259 Seignobos, Charles, 37, 173, 174, 177,

Mischel, Theodore, 80n60 227-228

Mogilnitsky, B.G., 254-257, 258, 265 Smelser, Neil J., 262f

Momigliano, Arnaldo, 279n13 Sokolowski, Robert, 174n4

Morgan, Edmund, 122 Stinchcombe, Arthur L., 261-263 Morgan, Lewis Henry, 264 Straus, Erwin, 62n14

Muller, Gert, 79n59 Strauss, Leo, 255f

Murdock, George P., 20, 21 and nl8, Swanson, Guy E., 65, 67-75, 79

22, 70n34 Syme, Ronald, 162

INDEX OF NAMES 35] Teggart, Fredernck J., 4nl, 18 and Watkins, J.W.N., 24n20 nl6, 19, 20, 56, 95 and n3, 96 Weber, Max, 45, 46, 48, 59, 60 and

Tilly Charles, 49-52, 56, 198, n8, 61, 78

254-270 Wedgwood, C.V., 305n46

Tocqueville, Alexis de, 261f Wertenbaker, T.J., 123-127, 229 Torrey, Charles C., 7 and n5, 8 White, Hayden V., 235n6, 259

Toulmin, Stephen, 58, 59, 233, 246 White, Morton G., 41, 63n16, 86,

Trotsky, Leo, 261f 209, 229, 235n6, 239 Williamson, J.A., 28-32

Vaughan, Frederick, 256 Winchester, Ian, 154n9 Woodham-Smith, Cecil, 45, 52, 53 Walsh, W.H., 306n48, 308n51 and n7, 260 Washburn, Wilcomb, 123-127, 229 Wrigley, E.A., 154n9

BLANK PAGE