A scientist I know, arguing for the importance of practitioners in her discipline learning fundamental theory as well as technical skills, asked me whether there are similar debates in history. That is: are there debates in the historical profession about what kinds of thing everyone learning to be a historian needs to know? I found this a harder question to answer quickly than I would have guessed.

On one level, the answer to a question like that is always, trivially, “yes”: historians pride themselves professionally on rejecting simple answers to almost any question, so it is impossible to imagine a consensus among them on any matter so large as the knowledge required to practice in the discipline. If such a debate did not exist, it would have to be invented. What is more interesting to think about is whether there is a parallel debate over what, or whether, a given body of fundamental theory is understood to be essential to a historian’s education. Without setting time aside to research this question thoroughly, I think the answer to this is much less obvious — but, ultimately, “no” — and that thinking about why that is suggests some of the differences between history and (at least some) scientific disciplines. Inasmuch as the status of history both intellectually and institutionally is a matter of some public concern, not so say legislation, these distinctions may be of interest to people other than practicing historians.

Historians do have a basic common commitment, most broadly, to the study of the past as such. (I almost said, the human past as such, but even that might be needlessly restrictive; animals, forests, oceans, and the earth itself have histories not exclusively determined by human intervention, and historians of these and similar subjects work with sources not exclusively of human provenance.) That commitment does, again very broadly, entail a common reliance on sources from the past, and a common practice of basing claims about the past primarily on the analysis and interpretation of these sources, often if not always in dialogue with past sets of claims constructed on the basis of similar analyses and interpretations. Historians rely on source criticism; they engage in historiographical conversation or argumentation. In doing so, they commonly employ concepts such as context, continuity and change, cause and effect, bias, probability, and so on. The result is written (or otherwise published) history.

As this suggests, handling sources and constructing arguments from them requires methodological training. Most of the time, in programs with which I am familiar, this focuses on documentary sources, written sources — texts of one kind or another. Though other kinds of source — visual media, material culture, oral tradition or testimony — are vital to many kinds of history, they are not invariably or even commonly core parts of historians’ training in fact. (This is an empirical claim, not a normative one.) Probably nobody with so much as a BA in History has managed to avoid at least a few exercises in source interpretation; they may begin in middle or high school. The necessary or ideal scope of such exercises may be debatable, but their importance is agreed. To return to terms of the question with which we began, however, this is a matter of technical skill, not fundamental theory.

Now, I do not think this kind of practice, even as simply as I have presented it here, is free of theoretical assumptions. The nature of the relation between remnants of the past and its lost “reality,” the logic that permits (or forbids) one to infer beliefs, intentions, or even bare facts from sources, the abstraction that conjures coherent processes of change from myriad yet sparse particulars, the erasures that shape the archive and that its authority reinforces — all these matters and more bedevil the most basic research. Not even the sparest historical account is theory-free. But it may be, is even likely to be, theory-inarticulate. (This is often imagined as a distinctive virtue of non-academic, narrative-driven, historical writing; but it is true of much academic and analytical history.) And one reason is that the literature that reflects most explicitly on the theoretical assumptions of historical practice tends not to be written either by or for historians. Even where courses in “theory and history” are taught, the former — if it is engaged with directly — appears largely exogenous to the latter.

There is such a thing as “philosophy of history,” which may connote theories of historical change (from Hesiod to Ibn Khaldun to Hegel to Marx to Weber, and so on), or reflections on method (Nietszche, Bloch, Benjamin, Foucault, Farge), or criticisms of narrative or archival practice (White, Scott, Davis, Trouillot, Chakrabarty, etc.). Some of these might seem like candidates for fundamental theory: the gravitation, or relativity, or quantum mechanics of history. But, setting aside for the moment the messy problem of dividing questions of theory from questions of method in many of them — and that many of the more “theoretical” names I’ve listed come from disciplines outside history — it is almost a matter of chance how much of any of them or their like a history student will encounter directly, as an essential part of training or induction into the discipline. I read Ibn Khaldun’s Muqaddimah (it now seems quaintly) as preparation for grad school conversations that never transpired, but Trouillot only long after getting a tenure-track job. Most of the Marx, Weber, and Durkheim I read in graduate school came from field preparation in sociology, not history. And despite brushes with big names of the 1870s-1970s or so, my grad school education included very few contemporary “philosophers of history.”

Of course, sustained, direct reading of theorists’ works in their original form is not a requirement of education in theory in every discipline. Probably few physicists today read Newton’s Principia cover to cover, even in vernacular translation, just as few chemists struggle through Robert Boyle’s A Defence of the Doctrine Touching the Spring and Weight of the Air, and few economists undertake The Wealth of Nations in more than excerpts. Modern formulas, systematic restatements, and textbook summaries do the work instead, by abstracting from the original (or from the last-but-one reconstruction of its abstracted content) just that set of claims understood to be true or at least relevant for current purposes. But while there may be nothing inherently wrong with this pragmatic abstraction from most disciplinary points of view, it is itself a deeply ahistorical procedure. So while there are books that survey and summarize theories of history, few historians would regard reading them as serious theoretical study; by the standards of the discipline itself, that would require firsthand engagement with the sources of the theories in question.

And here I think is a second distinction between history and some other disciplines — because, although many past theories (or, better, theoretical traditions) may have more or less critical adherents in the present, none of them is generally current for the discipline as a whole. So, for one thing, instead of being trained to work within theories of history, we tend to encounter more or less partial histories of theory. This is a very different thing. To the extent that one studies a theory historically, one understands it as a product of particular historical contexts, conditions, and processes, rather than as a free-floating set of truths about the past or rules for its study. And it is hard to treat a set of ideas as both the historically contingent object and the privileged analytical framework of investigation — at least, at the same time. (Incidentally, this is one reason why disciplinary practitioners in the sciences sometimes regard historians with suspicion or hostility, as if understanding the theories or laws they work within as historical phenomena were tantamount to judging, debunking, or denying their “reality.”)

Most of the time, however, historians-in-training encounter “theory” neither in textbook-style formalizations nor through deep study of sociologists’ or economists’ or literary critics’ or philosophers’ theories in their original form, but by reading putatively paradigmatic examples of its use in historiography. You may read bits of The German Ideology, the Grundrisse, or Capital, but your introduction to Marxist history (say) is more likely to come via E. P. Thompson’s Making of the English Working Class or Perry Anderson’s Passages from Antiquity to Feudalism or Eugene Genovese’s Roll, Jordan, Roll. Such, at least, has been my experience, both as a student and as an educator. I think this tendency of historians’ education to include (mostly non-historians’) theories not solely but chiefly as mediated through the example of other historians’ use of them in substantive research is distinctive. I also think it has one or two implications for the shape of the discipline that are worth tracing, by way of conclusion.

One is that theory tends to appear as ancillary to method. (Ask not what your investigation can do for theory, but what theory can do for your investigation.) There may be histories prompted by the desire to prove or disprove the hypothetical implications of a fundamental theory, but these are, in my experience, rare. More often, a body of theory is brought to bear because it enables a new set of questions to be asked about a set of sources or a subject of interest to the historian. There are Marxist historians, but even their work need not, and usually does not, centre on establishing the accuracy of Marx’s predictions; it can simply explore problems to which Marx drew attention, using concepts he forged to make sense of sources with which all historians must deal. Nor does this preclude engagement with other bodies of theory. (This is one reason charges of “Marxism,” “postmodernism,” etc., that equate methodology with advocacy often miss the mark; though, to be fair, hitting it is probably not important to the accusers.)

A second implication is that how (or whether) historians encounter a given body of theory depends a great deal on the subfield of history they happen to be in. Even the canonical works of Marxist historiography — to run with our example — are different for historians of the United States than for medievalists, Caribbeanists or specialists in early modern Europe. And while some theoretical traditions have manifested themselves in various ways across the discipline, other bodies of theory developed or were borrowed into a specific subfield, and have remained relevant there without influencing the practice of “history” as such. History of science is rich in theoretical discussions that barely register even in seemingly adjacent subfields, such as intellectual history. Art history is, for most purposes, a separate discipline from history altogether, notwithstanding its pertinence to the study of almost any place or time. Perhaps if there is no obvious set of fundamental theories for budding historians to study, it is because historians are trained to be historians of something, and they learn principally by reading and emulating older historians of that thing; there is no (or no longer any) substrate of “history in general” to serve as the matrix for basic theorizing. There is no history, and so no theory of history, that all historians need to know.

I have meant these comments so far to be empirical rather than normative. But — for what it’s worth — I am not sure that this lack of a common theoretical framework in history writ large is a bad thing, given what the dominance of any one of them would seem to mean. In any case, it seems inevitable. For one thing, historical research at the level of working with primary sources is still largely a matter of individual investigation, even where this generates data or interpretations for collaborative efforts or syntheses. For another, expertise grounded in knowledge of particular times and places rather than general themes or problems is well suited to idiosyncratic or ad hoc engagements with different intellectual traditions and resources. This is sometimes deplored; but the larger-scale, more capital-intensive work that characterizes at least some of the sciences is not something many history programs will now or perhaps ever be in a position to contemplate, for fiscal and institutional as well as intellectual reasons. And in a discipline whose future depends on diversifying not only its personnel but also its questions, methods, and approaches, a certain incoherence might well be for the best.