When thinking about the best ways to train doctoral students, historians tend toward conservatism. As methodological innovations percolate through the discipline, we may tinker considerably with the books we assign, and even the broad questions that we pose for our students. On occasion, individual faculty members may react against some aspect of their own experiences as graduate students, or try out novel ideas for a seminar. And over the past decade, largely in response to prodding from graduate students who felt like they had been thrown into the classroom with next to no preparation, many departments have instituted some kind of training for novice teachers. But the basic structure of graduate education—the format of graduate seminars, the nature of assignments and preliminary examinations, the expected scope of dissertation research and form of scholarly prose—hasn't changed very much from generation to generation.

There is no reason, of course, to pursue educational reform for its own sake. When a pedagogical approach works well, no one should celebrate its replacement. And in the case of nurturing academics at the doctoral level, the informal culture of a department surely matters enormously, whatever the formal requirements for coursework or prelims. Still, as the recent AHA report, The Education of Historians for the Twenty-first Century, demonstrates, our discipline has little excuse for complacency in assessing the effectiveness of doctoral training. Like many other disciplines in the humanities, we have significant attrition rates, as well as a painfully long average time-to-degree. Like many other academic disciplines in both the humanities and the social sciences, history also faces enormous challenges in the years ahead, including:

• the imperative of reversing trends toward an increasingly homogenous body of doctoral candidates, as the representation of racial minorities, individuals who attend non-elite universities, and children of the working class have all declined in recent years;
• the dilemmas of sustaining intellectual communities when faculty and students alike are buffeted by the demands of two-career families;
• the continuing oversupply of PhDs for the academic job market;
• the pitfalls of ever-greater specialization, which can encourage narrow research agendas and a growing distance between academic historians and broader public discourse;
• the daunting demands of preparing students for transnational and global research/teaching; and
• the question of how to harness the promise of information technology, both in terms of research and dissemination of knowledge.

Meeting such challenges will surely require fresh thinking about many aspects of PhD programs in history.

Since the fall of 2003, 16 history departments from around the country have been pondering the future of doctoral education as part of the Carnegie Initiative on the Doctorate (CID), a systematic rethinking of doctoral education in six disciplines that has been convened by the Carnegie Foundation for the Advancement of Teaching. The most intensive conversations have occurred within departments, but Carnegie has sponsored a series of meetings among faculty and graduate students from participating departments, where representatives have identified pressing issues and considered the merits of potential responses to them. CID history departments vary in size, location, and focus. Most are part of large public universities, though there are two private institutions represented as well. Some of the departments specialize in a few fields of history; others train doctoral candidates more broadly. Unsurprisingly, this diversity has led to various paths of reform. (For a list of participating departments, and a sense of the reforms underway, see the web-based “snapshots” at http://gallery.carnegiefoundation.org/cid/, which will be available towards the end of March, 2006.) But as we have shared ideas and borrowed initiatives from one another, our collective experience suggests some broader “best practices” for any department embarking on a thorough-going assessment of its doctoral program.

Two crucial starting points involve record-keeping and the determination of a department's pedagogical mission, including the identification of the analytical skills and habits of mind that its students should possess upon attainment of the PhD. Neither of these touchstones are likely to come as a surprise to many historians; but close attention to them can greatly facilitate evaluation of what about a graduate program works well, and what changes, might be worth a try.

One cannot effectively assess a graduate program's strengths and weaknesses without some key figures and some basic qualitative evidence—about placement, about attrition, about the time that it takes students to complete various requirements, about the perspectives that current and former students have on their experiences in graduate school, and so on. As the Duke University history department began the Carnegie process in the fall of 2003, we soon discovered, like almost all of the other participating departments, that we did not systematically keep track of a great deal of relevant information. We now do a better job in this area. We have also learned that our graduate students and former students, including the minority who left without receiving a PhD, constitute a crucial reservoir of insights about our program. Surveys of these individuals furnished extremely helpful reflections on the key elements of a vibrant intellectual community, and on the elements of our graduate program that we should nurture, or that might benefit from reconsideration. (Collecting and analyzing this kind of quantitative and qualitative data, of course, requires the expenditure of resources. But our experience suggests that a department committed to a thorough self-assessment can convince graduate deans to help with those costs.)

By the same token, explicit articulation of a department's aspirations for its PhD recipients, both in general terms and at every stage of the program, can greatly assist the determination of whether a particular curricular structure or particular academic requirement is well tailored to achieve its ostensible purpose. Doctoral programs that embrace a well-defined thematic niche or that focus on readying students for careers in public history will surely have a different look from those that maintain broader coverage, but admit relatively small cohorts of graduate students; and all of these programs will similarly diverge in important ways from larger programs that have more chronological depth and greater geographic reach.

For far too long, moreover, the rationales underpinning key elements of doctoral education in history have remained unstated, whatever a given department's profile. Thus if one can take the material posted on departmental web sites as an accurate guide, very few history departments go to the trouble of explaining the point of their preliminary examinations, or of describing, in even the roughest terms, what students should ideally get out of the process of studying for and taking those examinations. Without a clear delineation of what qualifying exams are supposed to accomplish, a department has little chance of figuring out whether they are working effectively.

In Duke's case, a series of departmental conversations about the purpose of the prelims eventually led us to replace traditional “examinations” altogether. Beginning with the cohort that arrived on campus in fall 2005, doctoral students will spend their first two and a half years in the program creating a portfolio of historiographic essays, research papers, and teaching materials in their various fields, along with their dissertation prospectus, and will then have an oral defense of the entire portfolio. At Duke, we have decided that this approach will likely do a better job of cultivating the skills that we want our students to have—the capacity to map a given terrain of knowledge, to make analytical connections across geographic and thematic fields, and to frame historical questions and arguments that speak to broad academic and even nonacademic audiences.

We, along with other CID departments, have made many other curricular reforms in the past three years. The University of Pittsburgh has moved away from traditional geographic fields, opting instead for a series of comparative, thematic areas, all supported by newly conceived, team-taught seminars. At the University of Kansas and Ohio State University, history departments have instituted ambitious programs to prepare graduate students to become effective teachers, relying on both formal teaching courses and informal mentoring. The departments at the Universities of Minnesota and Illinois at Urbana-Champaign have formulated detailed mentoring guidelines, while the University of Connecticut has launched a series of efforts to build intellectual community across the often divided subfields of the discipline.

At Duke, we have either inaugurated or are in the midst of planning initiatives in most of these areas, while also creating a new curriculum for the first two years of coursework, anchored by three required courses on historiography, research methods, and teaching, each of which brings students together across all chronological, geographic, and thematic fields. Our hope is that the required seminars will build cohesive cohorts of graduate students who can think, speak, and write across the boundaries of space and time, and who thus have the capacity to build the kind of intellectual community required for historical study in a global age.

In addition to reliance on sufficient information and careful attention to programmatic goals, the process of reform among CID history departments has depended on a few other key features. Broad participation of stakeholders, including both faculty and graduate students, has been absolutely essential, both in the initial stages of formulating and refining policy proposals, and in the ongoing labor of implementing them. Awareness of the changes being made by other departments has frequently spurred creative adaptation. In part because of the discussions across departments and with the staff members of the Carnegie Foundation, CID departments have further paid a great deal of attention to issues of communication. First and foremost, several departments have strived to make the all-too-often implicit assumptions underpinning our programs explicit, leavening the usual bureaucratic rules with articulated rationales for various elements of our curricula, and clearly delineated expectations for graduate students at different stages in their studies. Finally, the reforms that have emerged as part of the CID process are avowedly experimental in character. We are not simply assuming that they will necessarily bring the results that we hope for, and we are committed to assessing their impact and adjusting them as circumstances and evidence dictate.

There may be still more radical innovations in doctoral education worth a trial. At CID meetings among participating departments, there have been discussions about experimentation with research models more akin to those of the sciences, or of a broader array of dissertation formats, particularly in the area of public history, or of internship programs that would give a greater proportion of graduate students a taste of nonacademic work done by individuals with high-level historical training. (Such internship experiences would likely serve the dual aims of cultivating effective communications skills and better preparing students for the real possibility that they will not be able to find tenure-track academic jobs.) One might further imagine novel efforts to foster a larger stream of applicants from currently underrepresented groups, perhaps involving summer programs for talented undergraduates.

Whether or not other doctorate-granting history departments choose to venture into such largely uncharted waters, this is a propitious moment for rethinking graduate education in our discipline. Generational transitions are afoot in almost every department across the country, and the everyday work of historians continues to be transformed by the emergence of web-based archives, databases, and teaching platforms. For those departments that decide that the time is ripe for a reconsideration of their approach to training the next generation of historians, the experiences of CID departments should provide some useful points of departure for their own deliberations.

—Edward J. Balleisen is associate professor of history at Duke University.