We were talking about memorable courses in college. "I remember calculus," someone recalled. "The professor began classes standing on the edge of a stage. With his thumbnail, he'd pierce the cellophane on a pack of Camels, draw out a cigarette and pass it under his nose. Then, with a click! snap! from his Zippo, he'd light up and inhale, rocking back and forth as if he were about to fall off the stage. This performance held us spellbound. And that's what I learned about calculus."

What do students really learn from us? "If I mention it," a colleague assures me, "that means they learned it." He's kidding, I think. But consider a recent "round table discussion" in the Journal of American History on best practices for the U.S. history survey.¹ Eleven gifted historians talk about goals, content, and pedagogy. But what does it mean that this round table on teaching has 26 footnotes while a second round table in the issue (on "Federal Power and Southern Resistance During WWI") has 115? Of the 26 footnotes, what does it mean that none refer to research on educational theory, historical pedagogy, or student learning? The scrutiny we give to our teaching often sidesteps elements we prize in scholarly work. Much of what we believe about our effectiveness comes from anecdotal testimony (such as, "The bright students tell me…") or evidence that begs important questions, such as course evaluations and papers. Who is not tempted to define "good" teaching by almost any criteria except the one that truly matters: what do students learn?

"But I'm a historian, not an educationist!" That's what I was thinking when, in 1999, I accepted a fellowship from the Carnegie Academy for the Scholarship of Teaching and Learning (CASTL) to study a course I had designed as an alternative to the traditional lectures-plus-textbook U.S. history survey. What follows is a brief report from the field on a new type of scholarship that puts issues of learning at the heart of its concerns. Taking a scholarly approach to the problems we encounter as teachers may not be for everybody or for anybody all of the time. Nevertheless, it offers exciting possibilities for making history teaching more of a scholarly endeavor, and here is why.

If ever a course demanded inquiry into learning, surely it would be the introductory history survey. Courses vary from teacher to teacher, but the dominant pedagogical model is "coverage," a metaphor suggesting professors in biplanes broadcasting pages of knowledge over acres and acres of students. But when history is covered, something important gets covered up. "What is history?" I ask on the first day of class. "History is the story of the past, "writes a typical student. "Historians gather the facts and ideas that were going on during important periods of our past and write it down so the future generations can learn from it." The naïveté of this response is dismaying—The past has only one story? Historical inquiry is just gathering the facts and writing them down?—but students who think this way have drawn a plausible deduction from the way they were taught history. "Covering" historical knowledge covers up the epistemological linchpins of our discipline. Lectures can be brilliant and textbooks loaded with features, but they serve novices poorly when students are led to believe that history is no different from biology, except for the facts one is asked to read and remember.

To teach against this misconception, I designed a new model for instruction in my U.S. history surveys. The new model replaces coverage with what course-design theorist Grant Wiggins calls "uncoverage:" a deliberate attempt to lay bare for students the central assumptions, forms of inquiry, and cognitive habits that transform data into knowledge for practitioners of our discipline.² My courses retain a standard periodization, but in place of textbooks, students read a pair of best-selling histories arguing dramatically divergent arguments: Howard Zinn's A People's History of the United States, and Paul Johnson's A History of the American People. Each week, before reading the historians and considering the nature of their disagreements, students write their own histories of given topics based on questions they devise and apply to primary materials. "Make historical sense of these documents," I charge them, which they do in the form of written historical arguments that become the basis of class discussion and the students' graded work. This weekly assignment allows students to demonstrate a deepening understanding of basic elements of historical cognition, which I teach one by one over the length of the course.

When I took this model to CASTL, I wanted to write a thoughtful description of my course for "show & tell" in a teaching journal. But a mathematician's question brought me up short. "How do you know your students are learning what you say they're learning?" Cornered, I told the truth: "By the intelligent looks on their faces at the end of the term." She laughed—and I changed the subject. This was my first lesson in how to think about teaching: first and last, it's really about learning.

The next thing I learned took some of the sting out of that first lesson. Research on teaching, emphasized Carnegie's Lee Schulman, will look different when historians do it than when natural scientists or engineers do it, because scholarship on teaching and learning should be grounded in disciplinary domains of thinking. This was a liberating idea—it saves historians from having to become educationists—but it came with the reminder that scholars, like monks, answer to a demanding rule of conduct, which in our case requires carefully crafted questions, sophisticated theoretical approaches, multiple sources of evidence, and accountability to others through rigorous peer review of our work.³

Historians call enthusiasts who show little interest in scholarly concerns "buffs." If I was going to be more than a teaching buff, I needed a question more sophisticated than "Is my course cool, or what?" The question I settled on put me to work examining to what extent novices in my surveys become "historically literate." I defined "historical literacy" in terms of: (1) Attitudes and understandings (for example, "history is valuable," "history is an argument without end"); (2) Cognitive habits (questioning, connecting, sourcing, inferring, considering multiple perspectives, and recognizing limits to knowledge); and (3) subject-specific information (such as the who, what, when, why of "black power"). With my question defined, I went looking for evidence.

Measuring learning entails stepping through a minefield of theoretical and practical problems. One way of disarming these problems is to employ multiple measurements, so I planned an array of exams, course evaluations, student self-reports, and comparison of first and last papers to determine what students in two survey sections had learned.

Evidence from these sources was very encouraging. Students reported they learned a lot. They liked "uncoverage" and said their appreciation for history had grown. Performance on exams reached the same mediocrity as in my old lecture courses. But I will pass over such evidence here. It is one thing to quote students who say they have learned to think historically, and quite another to provide evidence from performances indicating deep understandings beyond mere imitation.

To find out whether students' competencies had improved at all, I borrowed from Samuel Wineburg's important work on historical cognition a technique called "think alouds." The idea behind a think aloud is that if a subject can be trained to verbalize thought patterns while completing a certain task, then researchers can record and analyze these introspections to determine the cognitive processes displayed therein.⁴ Wineburg used think alouds to determine "the epistemological pillars" that distinguish experts' historical thinking from that exhibited by bright high schoolers. It occurred to me that think alouds might be powerful tools for longitudinal studies aiming to measure changes in thinking over time.

So, in week one of my course, I selected six subjects to meet with me individually for the purpose of conducting a think aloud. After training subjects to think out loud, they were then presented with primary and secondary documents concerning the Battle of the Little Bighorn, a subject outside the period of our course. After directing subjects to "make sense of the documents" in any way they wished, a tape recorder was turned on and subjects read their way through the documents, verbalizing any and all thoughts that came to mind as they struggled to make sense of the past before them.

Later, the tapes were transcribed and analyzed by a second evaluator and myself.⁵ Analysis involved coding each discrete verbalization to indicate the type of thinking it represented. Fifteen different thinking processes were identified, ranging from the uncategorizable ("it sure is hot in here") to comprehension monitoring ("I don't understand why Benteen would have done that") to various types of historical thinking, such as sourcing ("I can't trust Whittaker; he wasn't there.") After coding each think aloud independently, my partner and I then rated each subject's proficiency on six thinking skills emphasized in the course, using a scale where "1" indicated the naïve ability of an average high school student and "5" indicated the sophisticated ability of a trained historian. We then compared our coded transcripts until reaching consensus on a rating for the students' abilities in the six key areas.

At the end of the course, the subjects completed a second think aloud, this time making sense of documents related to the Haymarket Bombing. When these sessions had been transcribed and coded, we compared the first and second think alouds in order to determine what gains students had made in historical thinking.

I can only offer a sliver of my findings here, but the think alouds opened a fascinating window into the "on-line cognition" of students before and after the course. In her first think aloud, "Michelle" suspected bias in a few sources but assumed this made them unusable for any purpose whatsoever. On our 5-point scale, we scored her a "2" for sourcing. Ten weeks later, Michelle's thinking had grown considerably more sophisticated. Overviewing the documents for her second think aloud, she commented: "I have kind of a totally different approach than I did to the last one. I want to try to look for the change, the progression of what happened over the time period. And, uh, look closely at the biases there might be, where the document's coming from." Now Michelle analyzed sources not only to determine credibility but also to establish what questions the source might and might not be useful for. This was impressive improvement for a beginning student.

Overall, the think alouds revealed cognitive enhancements that were not as dramatic as claimed in student self-reports nor as sensational as I hoped to see, but significant enough to show that "uncoverage" is a fruitful approach for teaching introductory surveys. The largest average gains came in the ability to construct good historical questions (+1.6) and make supportable inferences (+1.3); the smallest change was noted for recognizing limits to historical knowledge (+0.4).

Teaching historical thinking turned out to be far more difficult than I had realized. "Teaching by mentioning" has no impact at all here, nor is it likely that workbooks and "critical thinking" readers offer an easy way to success. If we want to help beginning students get started with historical thinking, the critical elements seem to be well-designed exercises that uncover the routines of historical thought, and recurring performances (e.g., writing the same assignment once a week) that allow a sharpening of skills.

When talking about effectiveness, modesty becomes a teacher. My research seems to show that "uncoverage," at the very least, does students no harm. Can those teaching with traditional methods say the same? I have put this question to many audiences. Professors get quiet, while students smile knowingly. What are they learning in our surveys?

—Lendol Calder is Chair of the Department of History at Augustana College and the author of Financing the American Dream: A Cultural History of Consumer Credit (1999).

Notes

1. "Teaching the American History Survey at the Opening of the Twenty-First Century: A Round Table Discussion," Journal of American History 87 (March 2001), 1409–41.

2. Grant Wiggins and Jay McTighe, Understanding by Design (Association for Supervision and Curriculum Development, 1998), 98–114.

3. Lee S. Schulman, "Teaching as Community Property: Putting an End to Pedagogical Solitude," Change (November/December 1993), 6–7.

4. Samuel S. Wineburg, "Probing the Depths of Students' Historical Knowledge," Perspectives 30:3 (March 1992), 1–24. See also Maarten W. van Someren, Yvonne F. Barnard, and Jacobijn A. C. Sandberg, The Think Aloud Method: A Practical Guide to Modelling Cognitive Processes (Academic Press, 1994).

5. I thank Professor Tim Hall of Central Michigan University for his help with my project, and Sarah-Eva Carlson of Augustana College.