Letting Go

To the Editors:

When I first picked up Mr. Hexter’s “review” [NYR, Feb. 9] of my Foundations of Historical Knowledge and of Arthur Danto’s Analytical Philosophy of History I did so with high expectations, knowing of Hexter’s talent as a historian and of his interest in the philosophical aspects of his discipline. I did not expect agreement but I did think that Hexter would accurately report what I said and fairly criticize my views where he disagreed with them. Instead I found an illtempered, confused set of generalizations about what analytical philosophers believe about history, distortion and misrepresentation of my views, and a peroration on the “wholly denotative” language of science and the “connotative” language of history that should win Hexter some kind of prize for philosophical incompetence.

To support these charges I must first say something that Hexter might have reported about my book if he were not infatuated with the sound of his own harsh noises about analytical philosophers “fiddling with history” and claiming to “have all of knowing, meaning, and truth by the seat of the pants,” about their “pretending” that certain important questions concerning history are irrelevant, and about their being ingenious, “clever” men who are out to dupe their readers.

My book is a study of the interconnected roles of fact-finding, generalization, and value judgment in historical inquiry. Among other things, it contains a detailed analysis of the logical connections between the historian’s causal statements and general truths about human behavior, also an analysis of how these general truths differ from the more speculative generalizations of traditional philosophers of history. After discussing the logic of explanation I distinguish a narrative from a chronicle by saying that a narrative is a conjunction of causal statements that link historical events and conditions with each other. Then, arguing that some true narratives are superior to other true narratives, I try to formulate the basis of this superiority. This effort leads me to a discussion of how the historian selects his facts and how he distinguishes the decisive cause of an event from other contributory causes. In these areas of historical thinking value judgments are implicitly made, and we often regard a true narrative as good because we agree with these value judgments of the historian. It follows that some of the grounds on which we assess a history are different from those on which we assess an empirical statement of science because, in giving high marks to a history, we agree not only with the narrator’s factual assertions about past events but also with his decision to make those assertions rather than others which are equally true of the subject whose story he is trying to tell.

I take great pains in my book to point out that although a narrative is a logical conjunction of causal statements, it is not merely that because it has other features which figure in an assessment of its worth. But so bent is Hexter on saddling me with the idea that narrative histories can “by reduction be assimilated to scientific language” that he is led to distort what I say by means of a polemically convenient excision from a quotation. He reproduces a metaphorical passage in which, in order to explain my concentration at one point on the logical structure of narrative discourse, I say that a roentgenologist may X-ray human beings in order to show that they have a common skeletal structure without denying that they have skins that may vary in texture, color, or beauty. Certain stylistic aspects of a history, I intended to say by means of this metaphor, can be distinguished from its logical structure, but I made perfectly clear that the effort to delineate logical structure should not be construed as denying that some narratives are better written than others, or that we may prefer some narratives to others precisely because they are better written.

Indeed, in that very passage, in order to anticipate a charge like Hexter’s, I said the following: “If upon discovering this [skeletal] structure a roentgenologist were to come to the absurd conclusion that men are nothing but skeletons, we should regard him as mad. And in the same way, if a logician of narrative, upon discovering its structure, were to conclude that narratives are nothing but logical conjunctions of certain kinds of statements, we should regard him as mad too” (p. 220). This Hexter omits from the quotation, replacing it by those little dots that can be so helpful when one is doing a hatchet job. And to make matters worse, Hexter attributes to me the exact opposite of what I assert, namely, that “all that we know about the structure of the human body, all the truth accessible to us about it, is what the roentgenologist finds by X-raying the skeleton.” By analogy I am then assigned another view that I expressly deny, namely, that all we know about a historical narrative is that it is a logical conjunction of causal statements. How much more decent a reviewer Hexter would have been if he had printed those few extra lines of mine! But if he had, of course, the main point of his attack would have vanished. Had he quoted me fairly he could not convince any reader that I believe that the esthetic qualities of a history are non-existent, or unimportant, or foolish, or that history is reducible to “the scientific language model.”

Furthermore, if Hexter were to reread or read page 254 of my book he would find me saying, after I have described certain aspects of historical narrative:

It may be said, of course, that this is completely parallel to what goes on in natural science. We assess a physicist’s theory not only on the basis of its powers to organize his data but also on the basis of certain features of the data. If his measurements are faulty, we may count that against the physicist, and if he fails to take into account certain data with which we are familiar, we may also criticize him. But—and this is a fundamental difference between the situation in physics and the situation in history—a historian may protest against the inclusion of data in the material to be colligated on grounds of interest or value in a way that is foreign to the method of physicists when they criticize each other. A historian may say that a certain datum does not merit his concern simply because it is uninteresting or unattractive by esthetic, moral, or political standards, but a physicist—who also selects his data, of course—does not usually exclude data on such grounds. The beauty or ugliness of the red-shift in the spectrum plays no part in the decision to regard it as a datum that a theory must deal with and link to other data. On the other hand, the data which a historian tries to colligate may be deemed to be data on evaluative grounds that another historian may not accept.

Since I explicitly say that a history is assessed on grounds which are different from those we use in assessing a scientific theory, and since I explicitly say that a history can be true without being good, I am astonished to find Hexter writing the following passage as if he had originated its basic idea:

What do good historians actually do when they set out to tell a historical story? The answer is that from available surviving records they try to tell the best historical story they can about the subject they have chosen. Obviously, to the extent that they have made narrative statements which by reference to the historical record can be shown to be extremely doubtful, they have told a bad historical story. But this leaves a great many true narrative statements that they can make. Among these, how does one distinguish those that constitute a bad true historical story from those that constitute a good one? This is a very difficult question, and one that we can only raise here.

Then, not content with parading the basic idea of this passage as his own, Hexter goes on to say that I couldn’t possibly subscribe to it, couldn’t possibly raise the difficult question he raises. Though I devote more than a sixth of my book to a chapter on just this question, Hexter claims that “it is one that orthodox analytical philosophers cannot raise at all, not only because they have never thought to ask it, but because their tool kit contains no instrument that enables them to deal with it. This is why they pretend it is irrelevant.” Now there is a point in Hexter’s review in which he seems to say that I am not an orthodox analytical philosopher, whatever that is, and so, strictly speaking, he may not be condemning me in this passage, but it comes right after an attack on me which is calculated to show that I “reduce” history to science and therefore can’t possibly raise the “difficult question” that Hexter raises. So the suggestion is unmistakable: I, like all the rest of these awful analytical philosophers, pretend that a profound and difficult question about narrative is irrelevant. Talk about cleverness and ingenuity and flimflam! No wonder that Hexter is determined to defend the historian’s right to use “imprecise, connotative, evocative, non-scientific rhetoric.”

In his pompous peroration Hexter says that a superior true narrative can be identified by making a distinction between what he calls “the wholly denotative” language of science and the language of history, which he says is both denotative and connotative. The truth about the past which a historian is committed to tell, Hexter says, “cannot be adequately communicated within the narrow confines of narrative explanatory sentences and the language of science. Therefore historians must break through such bounds in order to gain access to the whole range of rhetoric, denotative and connotative, that their command of language affords them.” What the devil does this mean? According to the only comparatively clear distinction between denotation and connotation that I know, a general term like “bachelor” denotes individual persons but connotes, for example, the attribute of being unmarried. But surely Hexter can’t say that in this sense all of the terms of science are wholly denotative whereas all of the terms of history are both denotative and connotative. Clearly, scientific terms like “lithium” denote and connote, and historical terms like “revolution” also denote and connote. If Hexter were more philosophically literate than he appears to be, he might say that when he uses the word “connotative” he has in mind what some philosophers mean by the word “emotive.” There is no doubt that we sometimes prefer one history to another partly because it emotively expresses approval of persons or actions of which we approve. But this fact, which is quite compatible with my view that value judgment enters into the assessment of a narrative, does not provide Hexter with a basis for holding as he does that “truth,” “knowledge,” and “explanation” mean something different in history from what they mean in other domains of discourse.

Who, I ask therefore, has been fiddling around with history? Who, in that other snappy expression of Hexter’s, is producing flimflam? Philosophers of history who try to explain what they mean and conscientiously defend what they say, or a reviewer who in quoting omits a passage that would clearly count against his interpretation of an author if it were allowed to appear in the quotation, a reviewer who implies that an author “pretends” that a problem is irrelevant when in fact that author explicitly formulates the problem, a reviewer who then proposes an allegedly original solution of that problem by using technical philosophical terms that he evidently does not understand?

If anybody is engaged in producing flimflam it is Mr. Hexter, who has such a fancy view of historical truth that he cannot meet one of the elementary obligations of historical scholarship, which is to report accurately what an author has said.

Morton White

Department of Philosophy

Harvard University

J. H Hexter replies:

I do not have the time (nor, I am sure, the New York Review of Books the space or the inclination) for an adequately detailed reply to each of the interesting points that Mr. White makes against my review of his Foundations of Historical Knowledge. I shall, therefore, have to confine myself to a very few that seem to me important:

1. In writing a review of two books in a limited space, one may be faced by a dilemma—either fully to summarize both of them or to focus on a central issue, common to both, which seems to have major implications. No matter which choice one makes, one is committed in advance to something less than perfect justice: to scant what one judges to be the central issue in the interest of detailed summary or to scant the latter in the interest of dealing with what one judges to be the heart of the matter. I chose the second option because I hoped thereby to render intelligible the long-standing failure of a considerable number of talented philosophers writing about history to say anything of much interest to historians.

In all fairness, I must add that the rather acrid tone of my review reflects not only my disagreement with White on a central issue, but twenty-five years of patient impatience with the radically imperceptive and stultifying views on history with which he has associated himself. That this accumulated impatience should find its target in his book is not, however, wholly unjust or unjustified. Of the group of philosophers who share these views, he alone has also been an accomplished practicing historian, so that one might have expected him to know better.

1. White quite explicitly puts his finger on the central issue in the very title of his book The Foundations of Historical Knowledge. The foundations that he offers for critical inspection in his book are reasonably solid for the fraction of historical knowledge he intends them to support. They provide little or no support for a great deal of historical knowledge, because White recognizes that great deal as knowledge only dimly, if at all. Therefore naturally he does not try to find out how customary historical discourse provides foundations for it. The title of his book suggests that he deems the issue as important as I do, and his letter convinces me I have identified it correctly, even if I have stated it in language that he is not fully attuned to. Let me try once more to state it. White’s position both in his book and his letter seems to be:

a) that in the books they write about the past, historians use many words, sometimes in the ways that scientists use them and sometimes not;

b) that insofar as what they write adds to knowledge and understanding, what they do is offer explanations of the past, often (always?) in the form of narratives;

c) that the narrative explanations they offer are causal;

d) that if well used, a very considerable part of the language that historians use improves history as “a literary art,” but does not add anything to the knowledge of the past, to “ordering truth” about it, which is achieved through operations b) and c).

In contrast my position is:

a¹) that historians sometimes offer causal explanations when they tell true stories about the past;

b¹) that, beyond this, they sometimes “explain” the past in ways that do not involve causal ascription;

c¹) that, beyond this, what they tell about the past is sometimes not explanatory in the ordinary sense of the word (and not, I think, in any sense that Mr. White would be happy with);

d¹) that in operation a¹), b¹), c¹), when successfully performed, historians extend our knowledge of the past;

e¹) that in order to do so, they draw on rhetorical resources not appropriate for use in scientific discourse, and that they do so not merely as a matter of “literary art” and “esthetic standards” but because they find resort to such resources indispensable for “ordering truth” about the past.

In all this, I do not suggest, did not suggest, and do not wish to suggest that White believes that “the esthetic qualities of a history are non-existent or unimportant or foolish.”

1. Mr. White seems to feel that I have said nothing about history (except what he has already said better) that a rational being could understand, or at any rate, make sense of; yet on the main point White understands me well enough, although there might be a shade of difference in our emphasis. He says that I am “determined to defend the historian’s right to use ‘imprecise, connotative, evocative, non-scientific rhetoric.”‘ I should put it that I am determined to assert the historian’s need and obligation in “ordering truth” about the past often to use just such rhetoric. So White did not misunderstand the crux of the issue that divides us.

2. Nor did I misunderstand him. The quotations from his book that he supplies in his letter strengthen my conviction on this point. In the review I omitted the quotation from page 220, not in order to do White in, as he appears to believe, but because it seemed neither to add to nor subtract from his argument, as I understood it.

White’s second quotation is intended to show that in my review I raised a question he had already raised, and not only paraded it as my own but argued that his philosophical commitments prevented him from even considering it. This is a misunderstanding of the intent of my question, but it is possible that the misunderstanding is my fault, the result of exposition easily misunderstood by a philosopher of White’s persuasion. The misunderstanding is the result of White’s construction of what I intended by “the best historical story.” My conception of the best historical stories includes the belief that they “instruct…better, they tell…more of the truth about the past.” [NYR, Feb. 9, p. 27]. But White’s account of what pertains to ordering truth about the past is exhausted by the skeletal structure of narrative sentences. Therefore he cannot raise the question, “How does the traditional discourse of historians tell more of the truth about the past than is exhausted by the skeletal structure of narrative sentences?” because for him it is meaningless; it asks how historians do what he seems quite sure they can not do and what I am equally sure they do. It is possible that had White looked close, he would have seen this distinction; but I should have made it so clear in the first place as to prevent misunderstanding, thus the fault is mine.

As to the rest of his long quotation, (p. 254) I think that it does add to his argument, but that what it adds is wrong. There may be a historian who has said “that a certain datum does not merit his concern simply because it is uninteresting or unattractive by esthetic, moral, or political standards.” but I have never heard one do so. Indeed, a historian who does should be ashamed of himself, because if such a datum has serious relevance to the problem that he is working on, he is simply ratting on his vocation when he fails to concern himself with it. Above all else that vocation requires him to do justice to the record of the past, and “esthetic, moral, or political standards” be damned. In any case, White’s argument still remains that historians are concerned with “literary art”—“the color, texture and beauty” of what they write, with its “esthetic value” and that they are also concerned with “ordering truth” about the past, but that the second interest is separable from the first. My contention is that the language of historical discourse is frequently evocative, affective, and emotive, not merely for the sake of literary art or esthetic value, but because that language helps historians to order and communicate truth about the past not otherwise communicable.

1. Since White does indeed seem to find this notion incredible or absurd, perhaps an illustration will help him. In my review I offered Garrett Mattingly’s The Armada as a sort of global refutation of the views on knowledge and truth in history to which White holds. More specifically, one of the persons who moves through that book is the Duke of Medina Sidonia. Mattingly introduces him at the moment when, because of the death of the Marquis of Santa Cruz, Spain’s ablest naval commander, King Philip, orders him, a landlubber, to take charge of the Armada in the greatest and most difficult combined operation ever to have been attempted. We follow Medina Sidonia from that moment to the time when he brought whatever he could save of his disaster-ridden expedition limping back to Spain. In the interval we have traced him along the foredoomed course of the Armada. We have watched him, as the catastrophes that others had unwittingly prepared came crashing down on the enterprise, doing all that he could do, perhaps all that any man could have done, to cope with them and stave them off—all in vain. After following Mattingly’s story of Medina Sidonia’s fate, we know that we have read the tragedy of a heroic man. And so we know something about the past, something that few if any knew before Mattingly wrote The Armada—we know that the Duke of Medina Sidonia, who had often been written off as a fool or an incompetent, was a tragic hero.

The point does not seem difficult. The language by means of which Mattingly enabled us to know that Medina Sidonia was a tragic hero was not the rhetoric of science but the rhetoric of history, in part connotative, evocative, and affective, rather than denotative, univocal, and free of affect. To know of Medina Sidonia what Mattingly enabled us to know is not of course to know an explanation in any ordinary sense of the word; much less is it to know a causal explanation. But to know that a man who lived long ago and who we thought was a figure of fun was anything but that is to know more of the truth about the past than we knew before. White, of course, could try to bring us to an equivalent awareness about Medina Sidonia in a rhetoric that is denotative, univocal, and free of affect; but it is not a very promising project, and I would not advise it. Anyhow it does not happen to be the way Mattingly did it. That leaves White free to assert that knowing a man we thought of as an incompetent fool was actually a tragic hero does not increase our knowledge of the past; but to do so is to be a little peremptory and preemptive about what counts as knowing. And this brings us to our last point.

1. White devotes a considerable section of his reply to straightening me out on what philosophers mean by “denotative” and “connotative,” for failure to heed this, he suggests, I should “win…some kind of prize for philosophical incompetence.” Even more generously, he helps me (and with a helper like White I need no hinderers) to say what he wants me to mean by “connotative,” but what I did not say and did not mean when I used the term. Yet the problem he raises is worth considering. I wrote the review as a historian, addressing a literate group of readers who are not philosophers, on the subject of historical discourse. In so doing, I used the words “connotation” and “denotation” and their cognates as they are commonly used.^* These terms are also used by philosophers in a more restricted and technical sense in their discourse. In context my meaning was probably clear enough for most readers (in one place it was clear enough to White for him accurately to repudiate it), so I do not quite see why I should have used the words in a way other than I did. Under the circumstances, it does seem to me that White’s directive on the use of “denotation” and “connotation” is both peremptory and pre-emptive. This would indeed be a small matter, but as White’s letter indicates, and as I rather suggested in advance, he is ready to be equally peremptory, pre-emptive, and prescriptive about “meaning,” “knowing,” “understanding,” “explanation,” and “truth.” And I really do not think that will quite do, which was what I tried to make clear in my review of his book.

That is all I have time or space to say here, but I hope to say more elsewhere soon (1) in an article shortly to appear in History and Theory, (2) in the entry HISTORIOGRAPHY in the International Encyclopedia of the Social Sciences, and (3), in a book on which I am currently engaged. From an occasional asperity of tone in White’s letter I infer that he would not be sanguine about the gain to the intellectual community likely to accrue from these enterprises. Still it will provide us, if he wishes, further opportunities for an exchange of ideas.