The essence of thinking historically involves a consistent focus on changes and the cause of change, with a concomitant ability to recognize continuities even amid change. Using historical data to define what significant changes occurred in the past or to note developments that do not constitute significant changes lies at the core of historical interpretation. Explaining why—that is, identifying the major causes of change or continuity—completes the chief tasks of historical analysis. At the intersections of history and many other disciplines, sorting out the areas of potential change from those fixed by the nature of the human species or by the inevitable qualities of social organization constitutes the principal challenge of interdisciplinary discussion.

Historical thinking is simple enough to define, but it is not easily taught or learned. References to change over time or to the balance between change and continuity come so readily to most professionals that it is chastening to step back and realize how many students never quite get the point. These students seem far more comfortable with fixed patterns of human or social behavior, or with stark before-and-after contrasts, than with the real subtleties of historical thinking about change. Others find it difficult to move from a kind of events parade—which appeals to "concrete operational" rather than analytical thinking—to the assessment of change for which the events serve as evidence. At an extreme, their habits lead to the "data dump" response to exam or essay questions, in which facts are meant to serve as their own justification.

One of the problems with historical thinking, as I define it, is that it depends on a number of other skills. No big surprise here, but it is a problem, first because in hoping to involve college students in handling change, continuity, and causation, I expect them to have assimilated some of the ingredients already (or, in the case of first year students, I try to help train them to do so), and second, because many history teachers bog down in some of the subordinate components and never quite get to the main point.

Historical thinking presumes: (1) An ability to form an argument, using data for a purpose. I work on this a good bit with first year students in the first part of a course, preaching about stating a problem, working toward a conclusion, excluding material that does not pertain to the argument, and so on. (2) A capacity to glean data from primary sources and to weigh the reliability of such sources. This is an important historical skill (which is also vital in other social sciences), but it should not obscure the ultimate historical thinking goals. Too many good history teachers currently see document assessment as the pinnacle of "thinking like a historian," rather than as a building block. Documents may as a result take up too much time, or refer to a single point in the past rather than focus attention on the broader processes of change. Source assessment is, however, a genuine ingredient, and no introductory college course should omit some training in it and (an important additional step) in using data to discuss change. (3) Some capacity for handling diverse interpretations and testing theories about change. Because historical thinking, for all its importance, does not generate laboratory replicable results, some experience in juggling diverse analyses is vital, though an analytical agenda that ended with a "great debates" approach would be doing students a disservice.

Finally the constituents of historical thinking listed here should, time permitting, be expanded to include comparative analysis, which links history and anthropology but which can also be applied to an understanding of change, as in, quite simply, asking students to compare "then" with "now."

The steps to historical thinking all have merits in and of themselves—a capacity to sort out conflicting interpretations, for example, helps students relate historical dispute to contemporary arguments about what kinds of changes are occurring now—but they must lead toward the greater result. Even with able high school students, perhaps not fully trained in some of the steps, exposure to analytical problems associated with change and causation is vital, lest they fail to uncover the larger goals of historical methods. Certainly the introductory college course, while still working on constituent habits of mind, must bring students to a recurrent engagement with the problems of assessing change.

Obviously, the chief means of inculcating historical thinking about change is explicit exposure. We must not assume that students will understand how to deal with change analytically unless we directly and recurrently talk about the process. Readings that deal with significant changes and attempt to isolate the principal causes of these changes are vital, which usually argues against a textbook diet alone. Lectures in which an instructor works through a problem of change—from outright revolutions, with their inevitable if unacknowledged concomitant continuities, to more subtle shifts—provide vivid illustrations of historical thinking at work. Deliberate discussion of periodization is to my mind essential, since this is how historians normally seek to manage change analytically and designate those points at which causation analysis is particularly essential. Examples drawn from different scales also help: a large social change like industrialization provides one opportunity for historical thinking, but so do smaller developments, such as significant shifts in leisure or voting. Teaching historical thinking, beyond providing illustration and opportunity, proceeds to some extent by trial and error. Some students catch on relatively quickly, but pending needed research on the cognitive constituents of historical thinking their facility is not easily predictable.

It is essential not to impede the process of inculcating historical thinking by lesser, distracting goals. In particular, the ambition to cover a certain amount of material should always take second place to the provision of adequate opportunities to develop thinking about change and causation. Testing mechanisms must obviously reward historical thinking, rather than lesser skills like memorization. Even in routine quizzes, for example, it is far better to ask students to identify change or continuity than to seek recall for single episodes. Thus, early in a first-year course, "How did the India of 1900 differ from the India of 1700" and "Identify three similarities between the Russia of 1900 with that of 1700," are preferable to "What were the policies of the British East India Company," or "Define tsarist autocracy."

Above all, history classes must provide recurrent sets of problems that require historical thought. This is where the goal of explicit guidance in dealing with change shines through. Early in my first-year course, I talk with students about some frankly mechanical guidelines in dealing with change. I advise them to beware of the temptation to assume that change in a society completely overrides continuity, to distinguish between change in preexisting directions and a shift in the directions themselves (which is what real periodization is about); and to test monocausal determinisms by looking at other possible variables. As the course proceeds, I provide examples and exercises that put students to work on historical instances where these guidelines are brought into play. I develop a list of analytical questions—questions that involve historical thinking—as a guideline to studying for the final examination, and then I draw the bulk of the examination from this list. And when papers are called for—particularly research papers—I insist that students build on issues that call for historical thinking, rather than elicit data, however recondite, for the descriptive record alone. At Carnegie Mellon University we have developed twin courses for our history majors, one on research methods, the other on issues in conceptualization, including hypothesis-building about novel topics on historical change. Both of these courses deliberately work toward furthering historical thinking. It is especially important that historical thinking be worked on sequentially from one course to the next: students should be able to learn habits of historical thinking concerning one subject area that they can apply to another area; they should be encouraged to identify the processes of their thought clearly enough to work on this transferability.

Thinking historically sometimes eludes even professionals. We receive too many monographs still that dip into a past point "for its own sake," without grappling with change or continuity. Yet the thinking process also receives encouragement from some of the newly explicit attention to the "constructedness" of institutions and behaviors. (Cultural construction is one of the concepts I now try to get first-year students to work on, precisely because it involves historical thinking about change and causation.) Perhaps, however, the final ingredient in teaching historical thinking is no single approach, but rather the excitement that its grasp can generate. When students can apply historical thinking not only to the past but also to recent developments—when they can approach these developments by worrying about real versus merely apparent change, or about causation—we as history teachers can be satisfied that we've passed the torch.

—Peter N. Stearns is dean of the College of Humanities and Social Sciences and Heinz Professor of History at Carnegie Mellon University. He regularly teaches in the first-year world history course. Stearns's most recent book is Meaning over Memory: Recasting the Teaching of Culture and History. He is vice president of the AHA's Teaching Division.