Although many musicologists devote a significant amount of time and energy to undergraduate teaching, discussions of teaching at meetings or in print have been relatively rare. The session "Musicology and Undergraduate Teaching" at the combined AMS/CMS meeting in New Orleans in 1987 and its subsequent publication in the 1988 issue of Symposium was a notable and welcome exception, including many valuable insights from thoughtful and expert people. I would like to continue that discussion here with some personal observations gathered over seventeen years of teaching music history to undergraduates.

Like most of my colleagues I began to teach in college without any formal or even much informal training in teaching. In graduate school I led recitation sections in which I reviewed information from the lectures and answered questions as best I could. The classes I observed, with the exception of my advanced musicology seminars, were universally taught by the lecture method. Naturally, when I began to teach myself I fell back upon the model which I had observed, and I uncritically accepted the implications of that model. I was a Professor and my job appeared to be to profess all that I knew to my students. I had the answers and the students' job was to get them. I would keep up with the latest musicological scholarship and then construct a well-organized syllabus reflecting that knowledge.

The first year passed in a blur of class preparation and the excitement of a new job. I was so wrapped up in getting ready for class myself that I really had no time to reflect on what I was doing. But in my second year I began to have some disturbing feelings, which gathered force as the year went by. After all, if I prepared a syllabus which included readings and a text that contained all the answers, then what was class for? My students could read as well as 1. Was I looking forward to forty years of reviewing the answers to students passively seated in rows before me? These doubts were coupled with the realization that I couldn't keep up with even a small percentage of the scholarship being published in my teaching area, i.e. the general music history survey. And if I couldn't keep up, then by definition I was no longer giving the students the right answers. And if I was no-longer going to give the right answers, then what was I doing in the profession? I was pretty depressed. There seemed no way out of becoming that archetypal professor, thumbing through yellow-edged notes and droning on about material with which he had become bored, and as a result boring the students to death. The only alternative might be to find a willing Mephistopheles who would exchange infinite musicological knowledge and bibliographical control for my depressed soul.

Because of this crisis I began examining some of my unstated assumptions about teaching. I realized that I had never really thought deeply about exactly what it was I actually wanted my students to learn. Out of this sometimes painful process of questioning I began to develop a more coherent philosophy of teaching, a philosophy which, I discovered, was at variance with the teaching model, the lecture, that I was using. My philosophy proceeded from the realization that education is a process. My primary goal, therefore, is to help students do something or be something rather than merely to learn about something. Of course, students need to and do learn about something as well, but that is not the final goal. Thus, I want my students to do music history and be music historians rather than just to accumulate facts and observe scholarly controversy about the music of the past.

What do I mean by doing history and being a historian? Basically I mean having students construct their own arguments based on data about music of the past, rather than simply receiving my arguments or those of other scholars. Since the discipline of history fundamentally involves the interpretation of the past, and not just facts about the past, we ought to be teaching that discipline. Students should be trying to answer questions: first, the question, "What is it?" answered through the analysis and description of the music, and second, the question, "What does it mean?" answered through the study of the art from its historical perspective.

This approach might raise some questions, especially since I believe that this process or attitude should prevail right from the beginning of one's study of music history. "After all," I can hear, "how can beginning students who might not even know that Mozart was born before Schubert be expected to construct their own historical arguments?" My answer is that to delay their struggle with historical argumentation is not only to falsify the discipline, but also to take away what is most exciting and interesting about it. Furthermore, I would assert that it is most important at the beginning level to lead students to an understanding of the discipline of music history.

This attitude certainly applies in other disciplines. When well taught, the sciences introduce the scientific method at the beginning level and force students to struggle with the implications of that method. Or, in something closer to home, we certainly teach applied music in this way. I would argue that teaching music history exclusively by observation makes no more sense than trying to teach piano by observation. No matter how inventive my arguments or those of other scholars are, students do not learn much until they attempt to construct their own arguments, any more than piano students can learn interpretation from simply listening to Horowitz. I am not saying that students should not read scholarship or ex perience the workings of a mind more sophisticated, or at least more experienced, than theirs; clearly this is necessary. But observation is only a step in the process of education, perhaps the smallest step. The model for learning must not be that we first learn all the facts and only later (usually in graduate school, if at all) learn what to do with them. To delay our students' experience with speculation and argumentation is, I think, to stifle creative minds, minds that we, like every other discipline, need.

If one accepts the above model as the goal of the music history class, then I believe it follows that the best method of teaching is by discussion. If students are to construct arguments they need to practice. They can practice, and need to, in writing, but the bulk of their practice must come orally. It is only in conversation with their fellow students and teacher that they will have sufficient opportunity to advance, take back, and modify propositions, learn to recognize arguments by assertion or other fallacies, and put together statements made by several colleagues into a new synthesis. In other words, they will learn what a struggle learning really is. It can be a very exciting experience. It is also, to be sure, a difficult and frustrating process for them and, often, for the teacher. My students at times just hate the question "And what is your evidence for that statement?" It is a constant refrain in our class. But it is the defense of one's statements with reasonable evidence (i.e., analytical, biographical, or other kinds of data) that ensures that the class does not degenerate into a mere opinion-giving session, an image that we often have of the discussion class.

I do not mean to say that the lecture format is unsuitable for all classes. Lecture works well for music literature classes; I teach a jazz history class for non-majors with lectures. I am saying, however, that you cannot teach the discipline of music history in a large lecture class. Large lecture classes deny students the opportunity to practice their skills in reasoning and arguing, skills they will need to master in order to be something more than mere observers of musical facts. I am not unaware of the budget implications of such small classes in music history. But if we believe that we are actually teaching a discipline rather than a collection of facts, we ought to be able to argue for appropriately sized classes, as have our colleagues in music theory. Certainly no one would argue that music theory should primarily be taught in large lecture classes. It is obvious that students in music theory need time in class to practice and demonstrate their mastery of the skills of that discipline. I believe that students in music history require no less opportunity.

Now, how might this actually work? What does an actual class do? How would the syllabus differ from that of the lecture course? Emphasis on process and teaching by discussion does not change what will be covered in a music history class as much as how it will be covered. The course contents must provide interesting opportunities for in-class inquiry. As always, tough choices have to be made as to what will be covered, and these decisions are even more difficult when teaching by discussion, since discussions are less "efficient" than lectures in simply covering material. Music analysis remains the most important basic activity of the class, but students are asked to describe the works rather than having the works described to them by the teacher. This activity focuses on answering the question "What is it?" Often, however, analysis can be done in such a way that it might help in answering the question "What does it mean? or, at least, "What might it mean?" and provide opportunities for stimulating inquiry in class.

Here are three examples. The class can be asked to analyze the Act H finale of The Marriage of Figaro from the point of view of the relation of musical and dramatic structure. In the following class they might be asked to analyze a Mozart piano concerto in the same way; they could even be asked to write a dramatic plot suggested by the musical structure. Setting up the assignments in this way provides the opportunity for both "covering the material" (that is eighteenth-century opera and concerto form) and learning to interpret the material critically. Of course, a teacher could explain all this to a class in a lecture, but in that case the students would merely be observing how someone else might construct a model for dealing with the relationship between music and drama in Mozart's music, not endeavoring to construct their own model. Similarly, a study of Schumann's Dichterliebe might ask the question "Did Schumann understand Heine?" First the students must read the poems carefully, then analyze the music, and then, finally, try to figure out the very sophisticated relationship between the two. Since this question has been answered by scholars in the field in diametrically opposed ways, it is hardly surprising that the class reaches several possible, though by no means absolutely provable, conclusions. In fact, the class usually begins with everyone agreeing that the answer is "Yes." As we study each song, however, problems with that simple acceptance arise and the class usually gravitates toward "No." When more evidence is examined, especially the postlude of the final song, most individuals in the class end up with the answer "Yes -- probably." Finally, for a general introduction to Romanticism I often show a group of slides of gardens from the eighteenth and nineteenth centuries. Starting with broad comparisons, the class explores concepts such as "closed" and "open" forms, "natural" and "artificial" landscape, "process" and "structure," "truth" and "order." There are no correct answers here, but there certainly are differences. I usually take the position that the perfectly pruned conical shrubs of the earlier garden are actually more "natural" than those found in the constructed disorder of the later garden, since their conical shape reveals the truest essence of the shrub. The class, in arguing against my position (they never agree), raises issues directly related to the perception of truth and art in Romantic music. They also learn that different times can have very different notions of truth and reality d our view of those notions is strongly conditioned by the values of our own age. I hope the exercise ends up as a cautionary tale.

In all of these examples the "correct" answer is much less important than the process of using evidence from different sources in an attempt to construct a model for understanding music and its evolution. Students learn that there might be a way of looking at music and interpreting the evidence in a manner quite different from what they had first assumed. They learn that models that explain some music might not work for other types of music. They learn to question assumptions, both stated and unstated, and to put together their own arguments. This, to me, is the essence of studying music history.

I am sure everyone reading this can come up with teaching strategies as interesting as the ones I have offered. My purpose here has not been to provide class assignments, but rather to make a plea that we do not intellectually shortchange our students, even at beginning levels. I do not believe that students need a lot of factual information before they can learn to think and interpret. I would argue that, on the contrary, if we are to produce better thinkers about music, they need to see that "facts" and interpretation are intertwined right from the beginning.

I am reminded of a quote from the novelist John Fowles which has given me some comfort in my search for a more effective way of teaching. He said, "Education is what is left over after you forget all the facts."