From History to Theory

The Rise and Fall of Historiography

One way to get at a big question like "What is history?" is to pose smaller, related questions. How do historians survey the property lines of the profession? How do we train our students? How do we trace our own genealogy? According to Hans Kellner, history, like other disciplines, works largely by imprinting students with particular sets of anxieties. If so, these questions are roundabout ways of asking, What makes us anxious? Kellner's argument has special resonance for those members of the guild with the greatest cause for anxiety: graduate students, junior faculty, and the growing numbers of nontenured temporary lecturers and staff. Anxieties seldom run higher than when money is on the line. My own fading memories of my time in the job market include a talk I gave at a big school. The talk went well enough, as best I can recall, but afterward, at a meeting with the graduate students, one of them declared, with a bit of hostility, that it took some nerve to present a "historiography" paper as a job talk. I had just finished a dissertation on historical imagination in American academia and had learned not to tell anyone who asked that I was interested in "historiography." Instead, depending on the question, I would describe my research interests in "philosophy of history" or "history and theory" or mumble something about "cultural studies." Like many other graduate students, as part of my socialization into the profession I had learned that historiography was not something to which "real" historians aspired. In letters, interviews, and even hall talk, I treated the term cautiously.

Historiography may have loomed especially large in my particular landscape, but it is a surprisingly ambiguous and highly charged term. It is also a word that historians cannot do without. It is one of those drab bits of professional furniture that has become so familiar, it is virtually invisible, at least until someone bumps into it. In his 1938 review essay "What Is Historiography?" Carl Becker measured the common meanings of this "unlovely word." His assessment remains recognizable today:

What precisely is Historiography? It may be, and until recently has for the most part been, little more than the notation of the historical works since the time of the Greeks, with some indication of the purposes and points of view of the authors, the sources used by them, and the accuracy and readability of the works themselves. The chief object of such enterprises in historiography is to assess, in terms of modern standards, the value of historical works for us. At this level historiography gives us manuals of information about histories and historians, provides us, so to speak, with a neat balance sheet of the "contributions" which each historian has made to the sum total of verified historical knowledge now on hand. Such manuals have a high practical value. To the candidate for the Ph.D. they are indeed indispensable, since they provide him at second hand with the most up-to-date information. From them he learns what were the defects and limitations of his predecessors, even the most illustrious, without the trouble of reading their works.

Becker hoped that historiography could mature into a more analytical and thoughtful enterprise ("a phase of intellectual history"). The mercenary approach to scholarship that disturbed Becker was scarcely unique to history. He spent much of his career decrying the decline of the liberal arts. But what was common in Becker's day has become dominant in ours, partly because we have so many more monographs and specialties. Half a century after Becker, the historiographic essay had become a boilerplate exercise in which students reduce each monograph to a summary paragraph; criticize the book as reductionist; string together enough summaries to fill out the requisite number of pages; and conclude with a call for a new paradigm. For most graduate students, "historiography" means the synoptic reduction of a variety of monographs on a given topic to a narrative march out of the darkness and into the intellectual light of their own dissertation.

Becker's account suggests the ways that the meanings of historiography have narrowed. In 1938 historiography had at least three discrete meanings. First, and most commonly, the word meant the writing of history in the abstract. In this broadest category, it simply meant written historical discourse of any sort. Second, historiography could refer to the critical study of the writing of history. This was the meaning Becker had in mind, and it included what other writers called "the history of historical writing" as well as philosophical discussions of historical method and exercises in source criticism. Finally, historiography could refer to a specific body of historical scholarship, such as "the historiography of Progressivism." In recent years, this last meaning has predominated, and few historians who specialize in the study of historical discourse refer to the practice as historiography or identify themselves as historiographers. Along the way, historiography has acquired the odor of beginner work, appropriate to entry-level graduate students preparing for exams rather than to a serious intellectual inquiry.

Although Becker imagined historiography as an analytical enterprise, by the 1960s historians had begun to use other words, especially theory, to refer to the reflective study of historical discourse. Word substitution may seem a trivial matter, but since key words of historical method are at issue, our linguistic habits deserve a closer look. The politics of word choice can tell us a good deal about other sorts of politics, and historiography's semantic shifts may help to sketch a history of the way we imagine history as a discipline. Here I would like to follow Carl Becker's lead by chasing semantics into pedagogy. A history of the use and abuse of historiography will lead us into the situations where the word has most resonance-places where we write about and teach the history of historical discourse. That focus, though, creates peculiar obstacles. The historian's reference to the profession as a "guild" is not purely a bit of nostalgic hope. Graduate students typically learn teaching by observing mentors, and that process has not been well represented in the reflective work on the discipline. Simply put, try as we might to collect old syllabi or survey old course catalogues, we have no guarantee that these documents give us a truly reliable guide to classroom practice. Despite the recent emergence of the "scholarship of teaching and learning," we know remarkably little about the history of historical training. Since historiography has probably circulated more frequently in seminars and hall talk than in published documents, my semantic history here is necessarily brief and impressionistic. But while the particulars remain murky, we can trace out a narrative arc.

Historiography is indeed an unlovely word, and its homeliness is reflected in its absence from a wide array of dictionaries and encyclopedias. Historiography appeared only briefly in the 1933 and 1989 editions of the OED. There, it served chiefly as a synonym for history in the sense of a written narrative, mainly to distinguish history as a linguistic artifact from history as actual events. Another related word, historiology, "the knowledge or study of history," denoted the actual study of historical discourse, rather than the general body of written history. The most recent usage dated to 1813: "Erudition has been divided by a German professor into glossology, bibliology, and historiology." In a footnote to his 1988 social history of the American Historical Association, That Noble Dream, Peter Novick lamented the disappearance of this "once respectable word." And a recent attempt by Hayden White to bestow the kiss of life on the term has apparently failed. Still, historiology at least has its mourners, unlike the other archaic alternatives, historiognomer, historionomer, and historionomal. For good and ill, historians ended up with historiography, a word that survived its early years as a synonym for historical writing and, having cannibalized historiology, expanded to encompass metahistorical reflection.

Most accounts of the rise of historiography as the study of historical discourse focus upon the German origins of "scientific" history in the eighteenth and nineteenth centuries. German historians could use the term Historiographie, but the language generated a host of other words that one might translate as "historiography": Geschichteswissenschaft, Geschichteschreibung, Geschichtesphilosophie, Geschichtestheorie, and Johann G. Droysen's preferred term, Historik. What these terms connoted ranged from exercises in source criticism to critical readings of monographs to philosophical hermeneutics. English and American historiographers tamed this linguistic wilderness by telling the story of "scientific history," a narrative in which modern methods of source criticism, as exemplified by the writing of Leopold von Ranke, created professional history and, with it, historiography. Herbert Butterfield's 1955 classic, Man on His Past, noted that Renaissance humanists had written both histories of historiography and treatises on historical method, but that such works amounted to little more than "ropes of sand," lists of titles with no analytical content. The real "history of historiography" began at Goettingen in the late eighteenth century. As early as 1760, J.C. Gatterer had called for a "History of History." According to Butterfield, the papers of the youthful Lord Acton suggest that, by the end of the 1850s, the subject "was already a familiar thing."

By the end of the nineteenth century, professional historians in England and the United States had a substantial list of titles to consult for direction on historical method, narrative tradition, and philosophy of history, and the possibility, if not yet the habit, of referring to such works as "historiography." How-to books on source criticism mingled promiscuously with evolutionary stories of the history of historical discourse and philosophical speculation upon historical substance. German works decorated these early bibliographies. One of the more suggestive bibliographies appeared as an appendix to Frederick Jackson Turner's 1891 essay "The Significance of History." Turner listed Lord Acton's 1886 essay "German Schools of History," but also detailed a long line of German works ranging from methods texts, such as Ernst Bernheim's Lehrbuch der historischen Methode und der Geschichtsphilosophie (1889), to surveys of philosophy of history, such as Rudolf Rocholl's Aufbau einer Philosophie der Geschichte (1878). Turner's own essay narrated the transition in historical discourse from romantic narratives of the lives of great men to more socioeconomic approaches to history that he embraced as his own. And he drew heavily upon one of his featured texts: J.G. Droysen's Grundriss der Historik. Turner's essay was the one of the first important historiographic works by an American, and we should mark its form: it was at once a philosophical reflection upon historical practice, a narrative account of the evolution of historical discourse, and a manifesto for a new history.

We should pause, briefly, to remark the text that became Turner's preferred exposition of method. Droysen's Grundrisse differed dramatically from other period choices, such as Charles Victor Langlois and Charles Seignobos, Introduction aux études historiques, in that it offered far more philosophical ambition. Droysen's text appeared in English language translation as Outline of the Principles of History (1897), and it mixed a broadly Hegelian epistemology and universal history with an infinitely practical account of the investigative and imaginative tasks of historical research. The book was at once a philosophy of history; a practical account of the mechanics of source criticism; an analysis of the various modes of historical interpretation; and a treatise upon the formal differences among various types of historical exposition, from essay to narrative to monograph. Droysen's work was also arguably the last important study to imagine all of these topics in a synthetic manner, as parts of an organic whole, each of which necessarily informed the others in a dialectical fashion. Much subsequent writing on historiography would systematically dismantle that architecture, pulling out particular pieces and discarding others. That the Outline appeared at all in North American practice is surprising. And in Turner's hands it had influence, if only because it invited reaction. Carl Becker and Merle Curti were only two of the future historiographers to pass through Turner's seminar.

Thanks to Turner's diligent biographers and many students, we probably know more about his seminars at the University of Wisconsin than we do about other training situations of that period. Most of Turner's students appear to have encountered their mentor's preferred historiographic texts, Droysen in particular, in one of the famous research seminars. Here, Turner worked in the fashion he had learned at Johns Hopkins, immersing himself and his students in a shared body of archival and primary materials, with each student producing a paper on an assigned topic relevant to Turner's own research. For instance, while he worked on his Rise of the New West, the seminar would focus on a few counties in a particular decade or two matched to census data. Turner would assign each student a topic-mining or immigration or agriculture-and at seminar's end, the collective project could produce something like a total history for a specific locale during a specific period. On the one hand, the seminar was a residue of German idealism as it materialized in Wisconsin public education, elevating the enthusiasms for local antiquities to something like universal history. On the other hand, it was also the beginning of a peculiarly North American approach to history as a social science. We do not know if historiography was even a key word, but the easy blend of philosophy of history, source criticism, and historical research was a hallmark of early Turnerian practice. The idea that source criticism and historical method had philosophical as well as practical significance proved attractive to an early generation of historians and helped to underwrite a growing body of technical literature.

By the early twentieth century, historiography could refer to a wide variety of commentaries upon historical discourse. Most professional historians were introduced to research methods through such texts as Droysen's Outline, Langlois and Seignobos's Introduction aux études historiques, and Bernheim's Lehrbuch. They could consult a growing number of annotated bibliographies. They could piece together the manifestos of predecessors ranging from von Ranke to Charles Beard. They could read and write critical reviews of contemporary scholarly works. More reflective thinkers could read Droysen or Croce. And new monographs narrated the march of historical science. In 1913 George P. Gooch's History and Historians in the Nineteenth Century became the first book-length history of historical discourse published in English. In 1922 James Shotwell's Introduction to the History of History appeared. In 1938 Harry Elmer Barnes's History of Historical Writing turned up on the book review desk at the American Historical Review. The editors sent it to Carl Becker, and his lengthy review became a classic in its own right partly because its reckoning of historiography as a form of intellectual history proved timely. James Westfall Thompson's survey of historical writing from antiquity to the twentieth century appeared in 1942, quickly followed by a long list of titles, including Butterfield's lectures and the Social Science Research Council's Committee on Historiography publications of 1946, 1954, and 1962. Three years later, in 1961, John Higham's History: Professional Scholarship in America, noted that, where the 1931 AHA Guide to Historical Literature listed only ten works on the philosophy of history (only one of which had appeared after 1875), the 1961 Guide listed fifty-nine titles. Historiography had arrived, or so we might presume.

In retrospect, we can imagine Carl Becker distilling a relatively new and fragile usage. Where Gatterer had spoken of a "History of History" and Benedetto Croce of the "history of historiography," Barnes and Becker simply said "historiography." That usage could appear when it did because of the growing number of monographs and even teaching texts that treated the history of historical discourse. And these changes in usage tracked changes in professional training. As Becker noted, the word tended to get used most often in teaching situations, and as the profession grew, so did the need for historiography. In the United States, the postwar expansion of higher education facilitated the development of methods courses and subsidized the market for texts. The increase in students and in available monographs also changed the way the subject could be taught. Where Frederick Jackson Turner had learned historiography in Herbert Baxter Adams's famous seminar at Johns Hopkins University, imported directly from Droysen's Germany, Turner's grandstudents-his students' students-frequently encountered general historiography in a lecture hall in courses taught by American scholars or European émigrés who assigned English translations of classic works of continental philosophy and history.

We cannot speak with too much authority upon the precise content of these historiography courses, but we can reasonably conjecture that a significant percentage drew upon that postwar publishing innovation the teaching anthology. If the nineteenth century introduced treatises upon source criticism and histories of historical writing, the teaching market provided a rich context for publications that sampled representative texts useful for the undergraduate and graduate classroom. By the early 1960s, Fritz Stern's Varieties of History (1956), Hans Meyerhoff's Philosophy of History in Our Time (1959), and Patrick Gardiner's Theories of History (1959) were readily available. Of those three works, the Meyerhoff anthology presented the clearest argument in its editorial comment and textual selection. For Meyerhoff, a professional philosopher, the story of the last century was the story of the collapse of historicism as a philosophical project and the rise of antihistorical challenges from positivism, existentialism, and Christianity: Meyerhoff began with Wilhelm Dilthey and ended with Karl Jaspers and the neoorthodox theologian Reinhold Niebuhr.

The Stern and Gardiner anthologies followed a trajectory very different from that traced by Meyerhoff. Each made his selections so as to tell a story of the increasing disciplinization and secularization of historical thinking from Voltaire or Vico to the cold war. Stern, a historian of modern Europe, emphasized the writings of professional historians. Gardiner, a philosopher, anthologized mostly philosophical texts. There was virtually no overlap in selection. Only one text (and author) appeared in both books-an excerpt from Karl Marx's German Ideology, although that convergence scarcely tilted either text toward radicalism. Each book, rather, complemented the dominant narrative of historiography as told by Butterfield and other historiographers.

Fritz Stern's book proved popular with historians, and by 1963 The Varieties of History had reached its ninth printing, an impressive record for a teaching anthology. Stern organized his text as a history of the evolution of source criticism. His very first author, Voltaire, appeared as a "pioneer of a new type of history," based upon the careful examination of primary sources, and Stern's second selection was Barthold Niebuhr, exemplar of the "critical method." That editorial decision helped to impart a certain conservative cast to such courses, as students learned basically Whiggish stories about the evolution of source criticism from either the ancient Greeks or the Enlightenment, right on up to their own sorties into the archives. Stern was not, of course, a vulgar Rankean. "History springs from live concern, serves life, deals with life," he declared, warning readers that future historians might have to "struggle to preserve a sense of the freedom of man." Stern's colleague Jacques Barzun, in his concluding essay on cultural history, agreed. Historical writing demands an "esprit de finesse." "The rest," said Barzun, "is footnotes."

That footnotes and esprit might go together was one of the morals students were likely to draw from another popular teaching text, Marc Bloch's elegant defense of historical studies, posthumously published and translated into English in 1953 as The Historian's Craft. Bloch's work had served as a defense of historical discourse against the demands of the sciences and social sciences. Learned, casual, and stylish, the book described the "critical method" as the true legacy of historical studies, the genuine contribution that historians had made to the creation and survival of a human civic order. For students, that claim must have had a special resonance, for one of the most memorable features of The Historian's Craft was that its author had died at the hands of the Nazis while fighting for the Resistance. Here, historiography set itself in a global context, although by the time North American readers encountered Bloch's treatise, the curriculum had become more cold war than world war. But craft still served as an instrument of freedom.

By the end of the 1950s, many departments of history in the United States expected graduates and sometimes undergraduates to suffer through at least one course in general historiography. A 1962 American Historical Association survey, The Education of Historians in the United States, reported that, of 376 four-year colleges, 35 percent offered coursework in "methodology"; 33 percent in "historiography"; and another 15 percent in "philosophies of history." It is not clear if these represented distinctly different courses or if the three fields could be compressed into a single course. Intriguingly, "Catholic colleges" and "Colleges for Negroes" offered methodology courses at a higher rate than did other institutions. Of the 77 "Ph.D.-training Institutions" surveyed, all offered "historiography" courses, although 66 percent offered them only to graduate students. Seventy-seven percent offered "methodology" courses," and 36 percent courses in "philosophies of history." Still, the authors of the survey doubted that this instruction was sufficiently broad. They recommended that undergraduates, as well as graduate students, receive a better grounding in the "classics of historical literature," in "philosophies of history," and in historical method. Recent graduates seemed sympathetic to such suggestions; 90 percent of recent history PhDs told the committee that a course in "historiography" or "philosophies of history" should be required of all doctoral candidates.

Although most PhD-granting universities offered historiography courses, it is difficult to divine what was actually taught. The three rubrics used by the committee-historiography, methodology, and philosophies of history-suggested the topical variety. The course could focus strictly upon methods: how to assemble a bibliography, distinguish primary and secondary sources, subject sources to textual criticism, abstract the relevant information on index cards, and piece together the story. Alternatively, a historiography course might introduce students to classics of historical literature. Or the historiography course could explore the philosophical analysis of historical discourse. In actual teaching situations, the blend of topics depended heavily upon departmental traditions and even more upon the whims of the teacher. The same course at Columbia University produced texts as different as Shotwell's narrative of the evolution of historical discourse, History of History; Barzun and Graff's Modern Researcher, with its chapters on bibliographies, source criticism, and outlining; and Fritz Stern's Varieties of History, which anthologized classics of historical literature from Voltaire to Huizinga.

Even those texts that attempted to synthesize historiography's many tendencies suggested how impossible the task had become. Houghton Mifflin found enough buyers for Wood Gray's 1956 primer, Historian's Handbook: A Key to the Study and Writing of History, that the publisher released a second edition in 1964. In eighty-eight pages the text introduced the nature of history, catalogued bibliographical resources for various periods and all the major continents, introduced source criticism, and offered a tutorial on prose. No page was wasted, for the back cover included a graphic guide to the arcane symbols of proofreading. Even the authors recognized that the subject had been so condensed that students would need to read each and every word: "Don't skim!" That the warning was needed said something about the evolving reading habits of young historians. Meanwhile, Norman F. Cantor and Richard I. Schneider's How to Study History (1967) attempted to give undergraduates the basics they needed to survive lecture courses and undertake their own research. Chapter 6, "A Practical Lesson in How to Read a History Book," offered an early step-by-step guide to "gutting" a history monograph. With "creative" technique, one could acquire a thorough grasp of a secondary scholarly study in half-an-hour's time, primarily by concentrating upon introductions, conclusions, and topic sentences while avoiding the bulk of the prose. Following chapters on evidence and historical prose, the authors closed with a quick survey of "historiography and philosophy of history." The radical compression of topics hinted at the difficulties that could appear in the classroom.

For those attempting to integrate all three of historiography's possible applications-methods, history of discourse, and philosophical analysis-in a single semester, the strain could be severe. In the late sixties, Robert D. Cross described Columbia's compulsory course as "an uneasy balance between methodology, historiography, and the philosophy of history." By 1970, when Walter Rundell Jr. published In Pursuit of American History, a survey of graduate training that doubled as a jeremiad, the courses had grown increasingly segregated and specialized. Methods and historical literature increasingly were taught as an aspect of a specific field. One learned research method in the research seminar on American Progressivism; one learned historiography in a colloquium that surveyed important writings on American slavery; one learned philosophy of history not at all or else in a specialized course on great thinkers in sociology or literature. According to the unhappy Rundell, most graduate programs had turned to specialized topical seminars rather than general historiography courses that stressed source criticism: "When professors hold widely diverging views on methodological training, the danger exists ... that some students may escape this training altogether."

Whatever vocabularies Turner and his students had used to describe those early seminar sessions, the days had gone in which a single course would smoothly blend philosophy of history, older monographs, and serious investigation of archival materials. Those three different historiographic practices had once fit into a single linguistic and practical space, but by the 1960s, discipline had taken its toll. The sort of breezy generalist work common in Victorian scholarship in the United States had come to look like amateurism. The time had past in which one would write a twenty-eight-page dissertation, in longhand, as Turner had done, or confidently set out Droysen's Historik as a text for explicating and legitimating the analysis of census data. Scholars truly engaged with Hegel or similar authors found new specialties for their work and began to speak with greater frequency of "philosophy of history" or "history and theory." And the accumulation of specialties, collections, and monographs meant that even the most energetic scholar could no longer keep abreast of the scholarly literature within a single field, let alone across fields. Was it really possible to offer a "general" introduction to source criticism when different fields could use sources so far apart? Mastering the National Archives was of little instrumental value to the medievalist who needed to polish her Latin. Reading Dilthey in translation offered little utility to the social historian anxious to collate data from early modern tax rolls. The physical separation of source criticism and research from serious reflection on historical discourse was more nearly a logistical necessity than a philosophical demand, but the separation of source criticism from critical historiographic work could have more subtle consequences.

Historiography spent a century gradually expanding its usage to encompass historiology; now its usage began to constrict. Historiography courses grew more specialized, and the number of courses actually introducing students to classics of historical literature declined, but the profession held on to a vague expectation that students would learn a simple creation tale that stretched from the origins of historical thinking (located anywhere from Herodotus to Herder) to the emergence of history as a discipline in nineteenth-century Germany and culminated in one's favorite (and usually American) predecessors. Insofar as the remaining courses on the history of historical discourse placed source criticism at the center of historiographic progress, they complemented the more numerous courses in which graduate students learned to do their own specialized research. At worst, the historiography course was an empty formal exercise allowing its professor to rant at length upon a topic in which he (seldom she) had little or no special competence. At best, it introduced students to a wide range of important texts and encouraged them to think historically-and critically-even of history.

In the late 1980s I was an undergraduate student at the University of California, Riverside, and one of the last for whom History 102: Historiography was a required course. My experience there was probably fairly representative. It was a lecture course taught by an older British historian, who organized the syllabus along fairly traditional lines by mixing treatises on method with overarching narratives of discursive evolution. Ernst Breisach's textbook, Historiography, was the core text, and the lectures followed that book's organization fairly closely. I can also remember reading E.H. Carr's What Is History? And our essay assignment was also a holdover: we were to read the entire works of a single historian. (My professor was not stodgy-he allowed me to write on Bernard DeVoto.) The course was not a popular one. The materials were not sexy, and few lecturers kindle the flames of undergraduate excitement with such topics as the Magdeburg Centuries. But once the quarter had ended, even those students who hated the course grudgingly confessed that the experience and reading had been useful. That did not prevent the department from removing His102 from the list of courses required for the major. With that move, the University of California, Riverside, joined the mainstream of American departments of history, most of which had done away with the course many years earlier.

The reasons for the disappearance of Introduction to Historiography are legion. Given the traditional resistance of most professional historians to anything approaching philosophical reflection, it is a surprise that the course was ever offered anywhere; its disappearance is less surprising than the fact of its existence. Nor can we safely imagine that there existed a hard, reliable core of "historiographic contents" in such courses-the label was sufficiently general that virtually anything could happen. In the sixties, some of those who demanded an end to the historiography requirement were actually student radicals who believed the course too frequently served as a site of ideological production. By the late seventies, many historians had a somewhat different reason for objecting to historiography courses-in such places students might be exposed to theory, especially of the feminist and French post-structural varieties, and thereby lose their enthusiasm or fitness for doing "real" history.

Something was clearly cutting into history. In 1960, history had been one of the three most popular undergraduate majors in the United States. Between 1971 and 1980, the number of American undergraduates majoring in history declined 57 percent. During the restructuring of the university system in the 1970s and {apos}80s, job lines were hard to come by and historiography could appear to both radicals and conservatives alike as a politically suspect luxury. By the end of the 1980s, historiography as an undergraduate requirement and a job specialty verged on extinction. A survey in fall 2000 of course catalogues from the top twenty-five departments of history in the United States found only two that required its majors to take a historiography course. No major department had advertised a tenure-track position in historiography since the early seventies.

Historiography's disappearance was untimely. Even if students, lacking a background in the history of historical discourse, still could learn the rudiments of source criticism in occasional exercises in lecture courses or specialized seminars, such exercises threatened to become ahistorical. And the disappearance of classics of historical literature and philosophy of history-or rather their removal to specialized corners-could incline history toward anti-intellectualism. The intellectual upheavals of the late 1960s and the 1970s-structuralism, post-structuralism, feminism, postcolonialism, and multiculturalism-hit campus just as departments of history eliminated the one place in the curriculum that seemed especially adaptable to engagement with such a variety of topics. If history students were to learn anything about any of these innovations, they would have to do so as individuals and specialists, picking up a piece here or there from other departments, depending upon their course schedules and personal inclinations.

Engagement with classic historical literature and philosophy of history did not end with historiography courses, much less with the decline in the popularity of the actual word. But it did divide itself into two distinct traditions. In one, reflection or historicization of historical practice became a sort of beginner's task, to be carried out as briefly and unself-consciously as possible, separated off from real historical work in a few prefatory and invariably Whiggish paragraphs. In the other, reflection became a specialty all its own. Or more accurately, rigorous historiographic discussion threatened to become a subspecialty of a subspecialty, namely, modern European intellectual history or, more recently, South Asian history.

The lexical shift has implications aside from questions of curriculum. In the usage of most professional historians, the shift away from historiography to history and theory implies that history is one thing and theory quite another. Where the single term historiography subsumed a wide range of meanings, history and theory may put events and writing on one side and reflection on another-history and theory easily slips into history or theory. In everyday practice, the separation of history and theory frequently translates into divisions of intellectual labor and narrative form. Many who identify as "working historians" treat theory as a mysterious black box filled with occult instruments. From time to time, we might run over and pull out some specific theoretical tool and then scurry back to history to see if it can be applied. In such applications, theory typically shows up in an introduction or an occasional present-tense sentence or paragraph stuck amid the larger narrative flow. For historians who think of themselves as theorists, the separation of history from theory is equally useful, since it allows them to imagine what they do as a practice more rarefied than the dismal fact-grubbing of their colleagues. For them, theory is different from-and somehow better than-simple intellectual history. The creation and maintenance of theory thus serves both philic and phobic tendencies even as it reflects and facilitates the bureaucratization of our intellectual life. We might expect all historians to do historiography, but we cannot have the same expectation about theory.

The shift from historiography to theory has also made it easier for both historians and theorists to avoid historicizing their own preferred practice. Those who identify as "working" or "real" historians routinely excoriate "theorists" for writing obscure, ahistorical treatises in the present tense, but "real" history has not been routinely rigorous about imagining its own pasts. Against the ahistorical senses of theory stand equally ahistorical varieties of practice. As Walter Rundell noted, historians have long responded to demands for theoretical reflection with a simple but effective Deweyan rejoinder: one learns history through practice rather than through speculation. But as Dewey would have reminded us, we cannot responsibly beg the question of where "practice" comes from. Historical discourse is the one subject we refrain from historicizing; instead we teach modern research method as if it were the stuff of timeless revelation. Despite an eternal round of "new histories," our historical sense of both theory and practice remains fairly rudimentary. Lacking a mediating third term, we suspend history between theory on the one hand, and practice on the other, and then leave each of these terms thoroughly unhistoricized. We may practice new historicisms, but our sense of scholarly selfhood remains, as Lowie once said of civilization, a "thing of shreds and patches." We piece together our intellectual genealogies from bits of hall talk, salacious gossip, random grad school memories, oral legend, and chronicles culled from oral exam lists, the entirety wrapped in decorative but transient jargon.

The story of historiography, then, is partly the story of the inability of scholarly discourse to develop a third term or space that mediates theory and praxis. Although I do believe that the shift from historiography to history and theory facilitated intellectual bureaucratization and segregation, I do not believe that a simple return would restore a golden age of intellectual community, or even that a "return" to historiography is possible. As far as the word goes, we may doubt that historiography, a term with a thoroughly checkered past, has much of a future. As far as curriculum, there are signs, if only in the form of a number of new teaching texts, of new interest in teaching the history of historical practice, if only under some other rubric. We should hope for a more engaging and dialogical course than the cold war curriculum that dominated our postwar historiography. As a logistical matter, however, it is much easier to reduce than to increase requirements for graduate and undergraduate study, and bureaucratic inertia alone makes it unlikely that we will see a dramatic rise in the number of departments of history requiring historiography. As for that unlovely word, when colleagues ask about my research, I tell them I am hard at work on a book about history and theory.