The Foundation of Historical Knowledge
Analytical Philosophy of History
For twenty-five years, ever since Professor Carl Hempel wrote The Function of General Laws in History, analytical philosophers have been fiddling around with history: At first they proceeded boldly and then with growing nervousness, as if, despite a fairly brave front maintained for public purposes, they had become increasingly conscious that the subject presented unforeseen difficulties. What are these obstacles to analytical philosophy which the ordinary professional work of historians presents?
Orthodox analytical philosophy is committed rather heavily to the notion that there is one sound way to advance knowledge, to provide adequate explanation, to enhance understanding of the real world, and to add to truth, and that the model of that way can be found in the language scientists use in presenting the results of their investigations. The sciences that provide the model are physics and chemistry. For example, Hempel’s model of a historical event was the cracking of an automobile radiator; Morton White in his recent book The Foundations of Historical Knowledge, the lighting of a match. It is worth noting that Hempel selects as a historical event an incident that does not demand the presence of any human actor and that White chooses one that involves no human interaction.
In writing about the past as past, which seems to be what they do when they write history, historians appear to advance knowledge of the past, to provide reasonably satisfactory explanations of it, to add to the truth about it. A few analytical philosophers have come fairly close to denying that this is so, arguing that by their ordinary way of proceeding historians cannot or do not add to existing knowledge, truth, or understanding about the past. For good reasons, however, this view has not commended itself to many people, even to many analytical philosophers. After reading a book like Garrett Mattingly’s The Armada, even analytical philosophers find it awkward to claim that they know no more about the events it deals with than they did before.
Once the analytical philosophers concede that in writing about the past historians do augment our knowledge of it, that historical knowledge, like scientific knowledge in this respect, is incremental and cumulative, they face a dilemma. For historians do not always, or even very frequently, write of what they have discovered about the past in the same way that scientists write of what they have discovered about nature: The language in which they communicate their findings is often manifestly different from the language or languages scientists use. This suggests at least the possibility that to one major discipline which systematically investigates actuality and in which knowledge edge is incremental and cumulative, scientific language does not provide an adequate or even a relevant model for presenting the results of its investigation. If this were so, analytical philosophy would have to reexamine the very model on which it is itself erected, that is, the model that ascribes to the vocabulary and syntax, the rhetoric of the sciences privileged or sole access to the kinds of knowing, understanding, meaning, explanation, and truth which are cumulative in character. For analytical philosophers this is, clearly, upsetting. It poses for them the pressing problem: How to dispose of that part of historical rhetoric that does not conform to the language of the natural sciences?
THE RESPONSE that many of the analytical philosophers have made to this problem can best be described as “assimilationist.” Their strategy has been to maneuver within reasonably safe distance of their base—their commitment to the scientific language model—in an effort to rope in as much historical writing as can be made to appear to fit that model and to leave out the unabsorbed residue of such writing, or to consign it to inconsequence with respect to truth or knowledge, or to sheer foolishness. The initial tactics, as outlined by Professor Hempel, was to set out the structure of scientific explanation, to indicate that in their work historians never fulfilled the conditions of such explanation, but that sometimes they came fairly close to fullfilling them. Thus he held out the hope that if the historians were to reform and systematically try to fulfill these conditions, history could be a science or at least a reasonable facsimile of one. Not surprisingly, this approach met with few takers among the historians. It thus became evident even to the analytical philosophers that unless they adopted somewhat less rigorous assimilationist tactics they were going to end up by assimilating precious little of what historians wrote.
A shift in tactics seemed all the more necessary when it came to the attention of the analytical philosophers that many historians believed that one appropriate response to a demand for historical knowledge was to tell true stories about the past. If analytical philosophy were less Anglo-American in orientation, this hardly startling discovery might have taken place sooner, since in most European languages “history” and “story” are homonyms (istoria, historia, histoire, Geschichte). To tell true stories about the past is not the historians’ only aim, but it is sometimes the aim of some of them; and this fact has troubled the analytical philosophers to such a degree that most of them have ignored the great many other things historians do when confronted with the problem of rendering a faithful account of the past. Since taking the story altogether out of history seemed neither a very hopeful nor a very helpful project, the analytical philosophers had to face its ineradicable presence by developing new tactics which would allow for its presence but nevertheless safeguard the overall strategy of assimilation.
Almost all of White’s Foundations of Historical Knowledge and much of Danto’s Analytical Philosophy of History are devoted to the details of these new tactics. Both books are the works of very clever men and they have elaborated their similar tactics with considerable ingenuity. Although there are interesting differences in detail between them, for present purposes it is more important to keep in mind the strategy and to identify the traits common to both books.
The common strategy of White and Danto is to defend the analytical base of operation, the science-language model, but to rope in enough of the story-telling aspect of history to persuade themselves and their uncommitted readers that using scientific language to subsume a particular event under general laws—so that the former conforms to the latter—is really no different from telling a story in the sort of language historians ordinarily use. In pursuit of this end, both writers, it is true, follow slightly divergent courses. White gives the back of his hand to critics of the analytical approach to history, such as the late Professor Collingwood, and to recent heretics such as Professor William Dray, while Danto tries to dragoon critics, heretics, and analytical philosophers alike into one big happy family, embracing them all in a great ecumenical bear hug. The difference between these writers, however, can be attributed mainly to personal idiosyncrasy, and the ways in which they differ, though technically of interest, need not detain us, since they do not lie at the heart of the matter, their general procedures, if not absolutely identical, being at least very similar.
ROUGHLY THE ARGUMENT of both books goes like this: 1) Stories are narrative explanations. 2) Narrative explanations consist of a sequence of narrative sentences. 3) These sentences of narrative explanation express or imply causal linkages between and within them. 4) Each such causal linkage derives its efficacy from a possible general law or laws which guarantee the connection of the expressed or implied cause with its expressed or implied effect; and these general laws are somehow present in narrative explanations, even though historians rarely state them. Thus when historians tell stories about the past, they are really doing very much the sort of thing that natural scientists often do when they elicit general laws from or apply general laws to nature. Meaning, explanation, understanding, knowing, and truth, are much the same for history as they are for the sciences; the model of scientific language which sustains analytical philosophy applies equally to history. So historians should stop being such nuisances to analytical philosophy because there is really nothing for them to be nuisances about…QED.
But I am afraid it does not work. What was to have been demonstrated—that the stories about the past which historians tell can by reduction be assimilated to scientific language—has not, in fact, been demonstrated. What has been demonstrated is how to appear to win games without actually cheating—and also without actually winning. For the hard fact is that the stories historians tell about the past are not merely a string of causally connected sentences of narrative explanation—although it is true that any story about the past will contain such sentences. Danto and White have managed to rope in a constant trait of the stories historians tell; they have not been able to rope in the stories as they stand in the books historians write. Both are aware of this, but White, himself a sound historian, finds the awareness more painful. Somewhat ruefully, in a sentence whose muddy rhetoric may be a symptom of his uneasiness, White remarks:
The historical narrative, the extended story, is so large and rambling by contrast to the single sentence treated by logicians that any effort to treat it as a repeatable and identifiable pattern of language may give an impression of remoteness and distortion well beyond what might be felt by the historian who finds his causal statements cast in single syntactical mold. (p. 220)
Here the oracle has a somewhat Delphic ring. White seems to suspect that historians may not find wholly accepable the notion that in telling stories all they are doing is setting out strings of words whose meaning and truth can be adequately dealt with by treating them as examples of “a repeatable and identifiable language pattern of narrative explanation.” He seems to imply that, as a historian himself, he is not quite satisfied with this either. Fortunately, however, White the Analytical Philosopher has at hand a pill to ease the discomfort of White the Historian. The philosopher soothingly observes:
The vast differences that human beings exhibit do not prevent us from X-raying them in an effort to discern the skeletal structure that each of them possesses…. History is a literary art as well as a ordering truth and if we neglect ordering truth and if we neglect some of the narrative’s literary qualities in order to clarify certain epistemological problems connected with it, our procedure is like that of the sane roentgenologist who searches for the skull without denying that the skin…may vary enormously in color, texture, and beauty [italics mine].
This unctuous ukase seems to put to rest the qualms of White the Historian, who thereafter ignores all the literary qualities of historical narrative. (It is hard to avoid the suspicion that Historian White is something of a patsy when Philosopher White lays down the law.) The ukase would not quiet the qualms of Professor Krieger, a historian who once observed: “What philosophers seem to be interested in are the…remains of historians.”
INDEED, Professor Krieger’s suspicions would approach alarm at the following passage, from Danto’s book, where some of the necrological flimflam is nearly identical with White’s.
It is the job of history to reveal…changes…and to explain these changes at the same time as they [sic] tell what happened…The skeleton of a narrative has this form:
/ ../ ../ ../ …./
but the skeleton may be fleshed out with extra descriptions, anecdotes, moral judgments and the like. But these…are philosophically…of secondary interest. (p. 255)
The primary philosophical interest of history to Danto appears to be / ./ ./ ./ …./. It is, moreover, no accident that both writers resort to the same dubious metaphor at the same place in their reasoning. For it is here that Clio seems to threaten their argument with disaster, and they are impelled to try to rescue it from that ruthless muse by distracting her with a display of sepulchral figures of speech and rackety nonsequiturs, like the ancient Chinese scaring off evil spirits with firecrackers. Their anxiety is at any rate understandable since here history threatens to undercut the models of scientific language on which analytical philosophy depends. But flimflam is flimflam nevertheless; however highclass and however decked out it still collapses from structural feebleness.
The sophistical device concealed in the metaphors of White and Danto operates through a set of implicit equations and subsumptions:
The historical story equals the whole body.
The series of statements of causal connections or narrative explanations derivable from the story equals the skeleton.
In analyzing narrative explanation the philosophers deal with the skeleton of history, which is all of history, insofar as history is concerned with truth or what pertains to knowing (see the italicized phrases in the above quotation from White).
The elements of the historian’s story left unanalyzed by the philosophers have to do with history as “a literary art” (White) and are merely “color, texture, and beauty” (White), mere “fleshing out” (Danto), which is to say that they have nothing to do with truth or knowledge.
If White and Danto can slip these four propositions by, they have rescued analytical philosophy from the threat of history. They have done it by assigning to the area of truth and knowing only those elements of the historical story that are most nearly reducible or assimilable to scientific language. But the least we can demand of their four anatomical propositions is that they be carried another and final step:
All that we know about the structure of the human body, all the truth accessible to us about it, is that the roentgenologist finds by X-raying the skeleton; all the rest has to do with its color, texture, and beauty, and is mere fleshing out.
This is a very curious view of the matter indeed. Surely a man suffering acute pain in his lower right abdomen might feel some alarm if a roentgenologist told him either that the only available treatment was a pelvic X-ray or that treatment was unnecessary because the veriform appendix had to do only with the color, texture, and beauty of the body—that scientific physicians, seeking necessary truth and useful knowledge about the body, were concerned only with its bony structure. Indeed he might reflect quite unfavorably on the sanity of White’s “roentgenologist.” The trouble here is not the consequence of literary ineptitude merely, but of the inherent foolishness of Danto’s and White’s analysis.
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. But 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. But if the aim of historical story-telling is to tell the best true story, not merely a true story, about the past, then the question cannot be shunted aside. For it is evident that none of the best true historical stories has ever consisted entirely of sequences of explanatory narrative sentences in the wholly denotative vocabulary of the sciences.
Moreover, they are the best historical stories, not merely because they please the consumer of history more than others do (although they certainly do that), but also because they instruct him better, they tell him more of the truth about the past. If this is so, then all the connotative rhetoric of historians which White and Danto so summarily segregate as “color, texture, and beauty,” mere “literary art,” comes tumbling back into the area reserved for meaning, explanation, and truth, where narrative explanation in the denotative vocabulary of the natural sciences alone is supposed to reside. And, indeed, why not? Why should we assume that in the writing of history the best rhetoric for arriving at the truth about the past is the one to which the best historians, committed to telling the best true historical stories about the past, will never wholly commit themselves? The notion that they ought to do so is wholly arbitrary; it has no logical foundation, and all the empirical evidence runs counter to it. In fact its sole foundation is the oldest sophistical dodge of all—the petitio principi, a collector’s item of rococo questionbegging. To demonstrate its inappropriateness one need only imagine the consequences of actually using it.
Suppose, for example, one were to reduce all the sentences involving narrative explanation in Garrett Mattingly’s The Armada to the denotative language of science and to present them so as to exhibit their causal connections. Then one would have a story of the Armada that would contain as many true narrative statements and causal ascriptions as The Armada contains. Would anyone seriously claim that the dreary result was the best true historical story about the Armada? Would he even dare claim that from such an account he knew and understood as much about that story as he could learn from Mattingly’s book, that he had been confronted with as much of the truth about it? And if he did so claim, would not any sensible man suggest that he had better rethink his notions of knowing, understanding, and truth? For such a comparison would make it painfully clear that narrative, explanatory sentences couched in a wholly denotative scientific vocabulary cannot do historical justice to the story that culminated in the fight in the Channel between July 30 and August 12, 1588. And the reason for this is plain: The truth about the past which a historian is committed to tell when he sets out to tell a historical story 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.
HISTORY UNDERCUTS the claim of the orthodox analytical philosophers that all the elements of truth and knowledge can be contained within or assimilated to the vocabulary and syntax of the language family of the sciences. This is history’s threat to the philosophers. Yet it only threatens them with the loss of what they are better off without. It requires them to surrender their claim that they have all of knowing, meaning, and truth by the seat of the pants. The claim has always been so preposterous anyway that, except for the philosophers themselves, few can ever have taken it seriously; and to judge by those actions that are said to speak louder than words, even they may not really believe in it. The claim seems to be as much the result of an intellectual imperialism generated by the sin of intellectual pride as of zeal for the discovery of truth, and as such it is but a foolish paper world well lost. An enterprise which for twenty-five years has investigated history without ever once subjecting to close and detailed scrutiny a single complete piece of historical writing by a competent historian in order to discover how it works, how it actually communicates the truth it contains, is doomed not by the malevolence of its critics but by its own fatuousness. If they can accept the actual conditions of historical writing, the facts about it that every historian must learn to live with, the analytical philosophers will find themselves in a strange but perhaps ultimately congenial company which includes historians and rhetoricians and perhaps a good many others who are as fully committed to the search for knowledge and truth as the philosophers are. There they would be cherished for the keen minds and magnificent technical equipment they could bring to the task of finding out how historians actually do succeed in communicating increments of knowledge, understanding, and truth with their sometimes imprecise, connotative, evocative, non-scientific rhetoric. It is a thing worth finding out.