In response to:

Star-Spangled History from the August 12, 1982 issue

To the Editors:

Gordon Wood’s review of Robert Middle-kauff, The Glorious Cause: The American Revolution, 1763-1789 [NYR, August 12] is interesting but frustrating—interesting because he perceptively discusses the split in the American historical profession between traditional narrative and quantitative social science, frustrating because he doesn’t explore alternatives to that sharply-defined choice. He mentions the work of such unclassifiable thinkers as Michel Foucault, Hayden White, and Frank Kermode, but doesn’t suggest that it presents a way out of the stale debate between “old” narrative historians and “new” social historians.

Both sides in that debate are hampered by the epistemology of nineteenth century positivism. Both assume implicitly that the facts about the past are “out there,” that the historian’s primary task is to collect them, and that assiduous data-gathering can bring us closer to knowing the truth about the past “as it really was.” The only disagreement involves whether to order those facts with an eye toward decoration and inspiration or toward precise measurement.

The debate is pointless because it never raises fundamental epistemological questions. Neither side acknowledges that historical facts are not simply “out there” but rather embedded in the questions historians frame. To keep framing intelligent questions, historians need to construct forms that give meaning to the tangle of human experience. Whether those forms are implicit or explicit and whatever labels are attached to them (theories, models, concepts, interpretations), all constitute “regulative fictions”—to use Kermode’s term, which Professor Wood quotes.

The word “fiction” does not imply falsity; it does not mean that historians should play fast and loose with evidence; it does not preclude exhaustive research and rigorous accuracy. It does underscore the essential point that historical explanations are crafted forms—mental constructions historians use to make sense of an inchoate mass of data.

The most illuminating works of history are those governed by the most imaginative and capacious regulative fictions. (One thinks of Perry Miller on New England Puritanism, David Brion Davis on slavery, Philippe Ariés on childhood and death, E.P. Thompson on the English working class, and Wood himself on The Creation of the American Republic.) These works are not simply characterized either by the narration or the quantification of facts. Rather, their chief distinguishing feature is that they use regulative fictions flexibly to explain changes in human experience without flattening its variety and complexity. The current debate within the historical profession tends to overlook this analytic, synthetic tradition.

The positivist heritage is alive and well among American historians, narrowing their methodological debates and de-sensitizing them to some of the most interesting developments in modern historical thought. The blurring of lines between history and fiction ought to humble historians, reminding them how fragmentary and oblique their view of the past must always be; it ought also to alert them to new possibilities. Giving up a positivist epistemology, they might explore a wider variety of regulative fictions and reveal a broader range of historical truths. They might even acknowledge the truth-telling power of literary fictions. Gabriel Garcia Marquez’s One Hundred Years of Solitude devastates positivist assumptions about linear causality and historical truth, as Professor Wood notes, but it also tells some profound historical truths about the “modernization” of a colonial society. That subject has inspired numerous valuable monographs but few syntheses as compelling as Garcia Marquez’s extraordinary novel. The point is not that historians should become novelists but that they might well ponder the infinite variety of paths to the past.

Historians do not need to be fashion-mongers to recognize that “the intellectual fashions of the late twentieth century” could reveal a great deal about the problematic nature of their craft.

Jackson Lears

University of Missouri-Columbia

Columbia, Missouri

To the Editors:

In his review of Robert Middlekauff’s The Glorious Cause, Gordon S. Wood observes that such eminent contemporary scholars as Bernard Bailyn, Eric Foner, Lawrence Stone, and Middlekauff are calling for a return to narrative history in the “story-telling” tradition of nineteenth century historians like Henry Adams and George Bancroft. Wood believes that any return to a methodology that partakes of the “narrative form” is innocent of the problems of dealing with “causality,” “motivation,” and “human intention.” These problems have been brought to our awareness, Wood states, by a whole tribe of French luminaries, including Sartre, Foucault, the “structuralists,” and others who are promoting “forms of non-linear thought.” Wood seems to feel that the return to narrative history represents some kind of failure of nerve.

It might be noted that John Dos Passos, whose USA Sartre hailed precisely for its “structural indeterminancy,” dropped that genre and went on to write narrative history, albeit pretty mawkish stuff. Perhaps the novelist realized that a “structuralist” sensibility, not to say a “deconstructionist,” cannot be used to write about the Founding Fathers since that sensibility had not entered their consciousness and hence could not have been a basis for their thoughts and actions. Jefferson assumed that the Revolution had a historical, linear explanation. “When in the course of human events it becomes necessary….” It is difficult to see how the eighteenth century mentality can be reenacted with the conceptual knowledge of the twentieth century mind. Imagine Jefferson composing the Declaration of Independence. “We hold these truths to be socially conditioned and dependent for their meaning on the ‘structures,’ ‘paradigms,’ and the ‘episteme’ of the epoch.” If narrative history risks being antiquarian, projecting back into the past a “non-linear” mode of explanation risks being anachronistic.


Wood’s own explanation for the theoretical problems confronting the historical profession today is both geographical and chronological. “Most historians, especially in the English-speaking world, are by no means ready to accept such Gallic logic,” and in particular “most American historians do not have much appreciation for the intellectual fashions of the late twentieth century.” The issues that trouble Wood, and rightly so, are neither peculiar to the twentieth century nor attributable to the superior consciousness of the Parisian intellect. Actually Henry Adams was the first great historian to grasp the impossibility of writing history. In his nine volume study of the Jefferson and Madison administrations, Adams came to the conclusion that he could not explain what he had set out to explain—the War of 1812. Neither his narrative form nor even his “scientific” attempt to establish sequential connections yielded causal understanding. History, like power, remained impervious to intelligence. Adams’ growing sense of “chaos” and “entropy” is as “decentering” as anything found in twentieth century “Gallic logic” (logic?). Long before Foucault, Derrida, and other contemporary movers and shakers, Adams saw that without knowledge there is only power (“authority is police”) and that without narration there is no rational explanation of events and thus “silence is best.” But Adams, who had the courage of his confusions, was honest enough to resign from teaching history at Harvard. What ought we to do?

John Patrick Diggins

University of California

Irvine, California

To the Editors:

As Gordon S. Wood rightly says, the “revival of narrative” in historical writing “will not be easy,” given the current preference for quantifiable or highly localized monographs and the skeptical philosophy that “historical narrative is just another form of fiction.” But any revival will be stillborn is so far as the issue is posed, as he puts it, between “problem-solving” scientific history and the traditional historian’s “Newtonian belief” in causality with a slothful inclination to remain “mentally in the nineteenth century.” To define the alternatives in these positivistic terms inevitably makes narrative form archaic, simplistic, or fashionably subjective (as in Doctorow’s Ragtime).

There is, after all, a modern philosophical line of development out of R.G. Collingwood (including William Dray and Louis Mink, for example) that finds in complex stories an inescapable way of answering certain kinds of questions. Science itself has its own history and therefore its own stories. Narrative is not opposed to analysis and problem-solving; it entails them both. Whether in histories or novels to narrate is also to describe, explain, and evoke, as the historian J.H. Hexter has shown. One might say the same thing about Freud’s Dora, which retells its story in a spiralling fashion that is not linear or Newtonian. Similarly, Doctorow’s novel about the Rosenberg spy case, The Book of Daniel, uses modernist techniques of flash-back and interrupted interior monologue in the service of a historical interest in the mentality of the Popular Front and the Cold War eras. Doctorow and some other contemporaries may have claimed to have dissolved the distinction between history and fiction, but his Book of Daniel depends upon appealing to it in order to give his account pertinence and meaning.

The root fallacy is to contrast imagination with truth, as if historians (or scientists for that matter) had no stake in imagination and novelists no stake in truth. If narrative historical writing is to revive in prestige, it is essential not to constrict it with a reductive notion of narrative and a positivist idea of truth, as if there were no such thing as what I have called, in George Eliot’s phrase, “the veracious imagination.” If this be treason to “Gallic logic,” let us make the most of it.

Cushing Strout

Cornell University

Ithaca, New York

Gordon S Wood replies:

Professor Lears is wrong when he suggests that historians need humbling. Historians are already very humble people. They know only too well “the problematical nature of their craft” and “how fragmentary and oblique their view of the past must always be.” To be sure, most of them have not gone as far as Henry Adams and Professor Diggins to admit “the impossibility of writing history.” But few historians nowadays are so naïve as to believe that their task is simply to collect facts about the past. Historians realize that understanding the past requires imagination and the use of what might be called “regulative fictions” in order to make sense of the collected data. Nor are they opposed to telling what Professor Strout calls “complex stories” that integrate narrative with problem-solving and that rely on coherence as much as on correspondence theories of truth. Many historians have blended storytelling with analysis very nicely, and, it is hoped, will continue to do so.


Yet when all is said and done, when all the concessions to subjectivity, imaginative reenactment, and the use of “regulative fictions” have been made, historians still remain necessarily tied to what Professor Lears calls “the epistemology of nineteenth-century positivism”—to the view that the past “out there” really existed and that they can through the collection and ordering of evidence bring us closer to knowing the truth about that past “as it really was,” even if the full and complete truth about the past will always remain beyond their grasp. It is precisely because ever-widening circles of our culture are casting doubt on this traditional epistemology that historians feel more humble about what they do. Some of the most eminent working historians such as G.R. Elton and Oscar Handlin know that ultimately there can be no alternative for their craft than this old-fashioned epistemology. Historians, warns Elton, “require not the new humility preached in the wake of Heisenberg, but some return to the assurance of the nineteenth century that the work they are doing deals with reality.” “The historian’s vocation,” writes Handlin, “depends on this minimal operational article of faith: Truth is absolute; it is as absolute as the world is real.” This faith may be philosophically naïve, may even be philosophically absurd in this skeptical and relativist-minded age; nevertheless, it is what makes history writing possible. Historians who cut loose from this faith do so at the peril of their discipline.

This Issue

December 16, 1982