Revisions to the Graduate Program May 15, 2004 INDEX Introduction 1:  The Size of the Ph.D. Program 2:  The M.A.-only Program 3:  The M.A. and the first year of the M.Phil. A. Course requirements and writing requirements in courses B. Course advising C. Registration and admission to courses D. Course evaluations E. Grades F. The M.A. seminar and lecture series G. The M.A. essay H. Introduction to the Discipline I. 4000-level courses J. Course distribution and planning K. Language requirements 4:  Doctoral orals A. Definition and structure B. Guidance on forming lists C. Timing D. Orals advising and preparation 5:  The dissertation A. Drafting and advising B. Sponsorship C. Dissertation seminars 6:  Fellowships and funding A. The cost of living B. Conference travel C. Outside fellowships 7:  Graduate Student Teaching A. Second-year TAships B. University Writing C: Upper-year teaching 8:  Going on the Market In Conclusion Introduction In conjunction with our department's participation in the Carnegie Foundation's Initiative on the Doctorate, our graduate Committee on Guidance and Evaluation (CGE) has looked closely this fall and winter at our program's structure and its requirements. As at many schools, the present program is a patchwork based on decades of evolution and tinkering, and requirements may persist long after the reasons for them have disappeared. We have studied the results of the Carnegie committee's extensive survey of our graduate students last spring, which highlighted a number of concerns we want to address. We have surveyed the requirements of several peer institutions (Berkeley, Chicago, Harvard, Penn, Princeton, the University of Virginia, and Yale), and have adopted or adapted a number of ideas from those programs. Finally, we have gathered statistics on enrollments, attrition, time to degree, and job placement. Taking all of this into consideration, we have developed a range of proposals, which have been refined in discussion with the Carnegie Committee and the Policy Committee and in a series of meetings with the full faculty and with our students. This report embodies the resulting proposals in the form approved by the faculty on March 2, 2004. We plan substantial changes in some areas, and in others we find things working well and needing fine-tuning at most; we also propose further work on various issues in the coming year. Having looked hard at all aspects of our program for the first time in years, we hope that the result will be a genuinely coherent program that embodies our collective sense of the best way to prepare the future scholars and teachers in our charge today. We have found it useful to begin from an awareness of some leading strengths of the program; we want to try and ensure that none of our changes would have the unintended result of harming its real strengths. We find general enthusiasm for the following: 1. We have as lively and engaged a group of graduate students as can be found anywhere in the country, and they do excellent work. We want to build on the flexibility, and the mix of collegiality and independence, that mark our program. 2. Over the past decade, we've gradually achieved full funding for all our sequential M.A./M.Phil./Ph.D. students, all of whom are funded at the same level (tuition plus $17,000 in stipend for six years, or five years for those who come in with an M.A. and enter the M.Phil. program directly as "second-year" students). Recruitment and student morale alike have benefited enormously by this full and equal funding, and we must be sure we don't do anything to jeopardize this. 3. We are more careful now than in the past to match incoming students to faculty resources. We want to ensure that any revisions (such as in the responsibilities of TAs) will not cause a skewing of admissions and an imbalance from field to field. Further, none of us wants to return to the old days of entering classes of 90 M.A. students and 30 M.Phil. students. Keeping these strengths in mind, we also had some serious concerns going into the process, and these have only been reinforced as we've studied our various sources of information. Ideally, our program revisions will help to mitigate the following problematic features of our program: 1. Time-to-degree remains high here as at most programs, and it hasn't dropped with full funding (though attrition has certainly dropped). Few students finish within the six years of funding. Can we improve this situation? 2. Advising has never been our strong suit, and continues to be a very active concern to our graduate students. Within realistic limits of time and attention, can we find ways to give our students better guidance and more substantive and timely feedback on their work? 3. Our students do a lot of composition teaching and relatively little teaching of literature courses; apart from the sought-after Lit Hum and CC preceptorships, employing a third of our students, opportunities to teach literature courses have actually decreased in recent years. Can we find ways to vary their teaching experience and give them greater opportunities to teach in their field? With these general points in mind, we proceed to the shape and structure of our program. Back to index 1: The Size of the Ph.D. Program A fundamental determinant of our program is its size. Just how many students should we have? Numbers vary widely from one English department to another, and even individual programs have often fluctuated dramatically in size in the postwar era. According to the best contemporary study of graduate programs, Bowen and Rudenstine's In Pursuit of the PhD (1992), the leading (Tier 1) programs in English granted an average of 9.4 doctorates in 1958, ballooned to 24.1 in 1972, and contracted to 10.3 in 1988; given attrition rates of perhaps a third, those figures would mean entering classes of about 13 in 1958, 36 in 1972, and 15 in 1988. Today no major programs have the very large sizes of 30 or more entering Ph.D. students that Columbia, Berkeley, and others had in the 1980s; most are either small (around 10-12 entering students) or medium-sized (15-20); a few large programs still have 20-25 students, and there are also a few highly regarded micro-programs, such as the Johns Hopkins Humanities Center, that take just a handful of students per year, though most micro-programs are at third- and fourth-tier schools with only nominal Ph.D. programs. In our major program revision of 1990, we achieved full funding for all M.Phil. students with a class size of 30, a number driven by staffing needs of the writing program. Finding ourselves overloaded, and with our students facing persistently tight job markets, we voted in 1997 to decrease the entering sequential class to 20. These cohorts were fully funded at the M.Phil. level, but about half the sequential students still had to pay full tuition for the M.A. With very welcome support by Henry Pinkham and his colleagues in the Graduate School, we've managed gradually to increase M.A. sequential funding over the past several years, and have reduced the class size to 18 to help us achieve a fully-funded sequential program. We have had this in place for two years now. As a result of these decreases, our graduate student/faculty ratio has declined steadily from 7.5 students per faculty member in 1995-96 to 5.5 per faculty member last year and about 5.1 this year. Our number of undergraduate majors has stayed steady during this time. We averaged 265 majors per year (juniors and seniors together) in 1994-96, and 266.5 in 2001-03. The question is whether the graduate program is now at an optimal size. As our first reduced cohort is now in its sixth year, this is a good time to assess our numbers. Recalling the stresses of our formerly much larger and less selective program, we feel strongly that an increase over our present class size would be harmful intellectually. It would also jeopardize our hard-won full funding, unless we had major infusions of new money from the Graduate School. There is no likelihood that we could be given funds to expand the program, even if we wanted to. The real question, then, is whether we should stay at 18 or decrease to the 10-12 students typical of small programs or an in-between number like 15. A decrease would have both advantages and disadvantages, which largely mirror each other. They can be summarized as follows: Advantages of a decrease: more time for close attention to individual students less competition among students for professors' time and then for jobs on finishing greater selectivity, perhaps a higher ultimate placement ratio time freed up for more work with undergraduates: we'd offer fewer graduate and more undergraduate seminars we could have most or all of our majors write senior essays, as is the pattern at Yale and at Princeton Disadvantages of a decrease: fewer opportunities to teach graduate seminars potentially claustrophobic atmosphere for graduate students fewer graduate students available to serve as readers/TAs and research assistants faculty may prefer the present mix of graduate/undergraduate work the administration might expect us to shrink our faculty if we shrink the graduate program further, especially if we don't undertake major new undergraduate commitments The CGE's discussion began with the relative advantages for our graduate students under each model. The faculty on the CGE were particularly impressed by the strong preference of our three Graduate Student Council representatives for the present size. Our GSC reps believe that our medium-sized program offers them wider choices of courses, and greater opportunities for friendship and intellectual exchange with peers, than are found in the small programs they passed up in order to accept our offers. Their views are supported by our survey of our graduate students as a whole, in which 89% of respondents describe their "intellectual relationships with fellow graduate students" as good, very good, or excellent, while only 11% rate them as either fair or poor. Social relationships lag just a little behind the intellectual ones, with 81% rating them as good, very good, or excellent, and 19% rating them as fair (15%) or poor (4%). Graduate student life is famously isolated and competitive, but these numbers indicate much more camaraderie than we could expect to find at most other programs, not to mention among the faculty of many departments. Our program size does mean that students are likely to be going on the job market at the same time as one or two friends in the same field, but at least they will have friends with whom they share common interests. Rather than breed competitiveness, our students say, the size of our program eases claustrophobia. Competitiveness was a major problem when we had up to 90 M.A. students per year and cohorts of 30 or more M.Phil. students, but it isn't a problem at our present size. Selectivity is also not an issue. As recently as the mid-eighties, we were accepting two-thirds of our applicants, an unconscionable proportion. Selectivity improved dramatically in 1990 with our reduced M.A. and improved funding, and has steadily increased ever since. We are currently accepting fewer than 5% of our applicants: we made 33 offers last year from a pool of 704 sequential applicants. None of us on the CGE believes that there is any discernible difference in quality within the ranks of our recent entering sequential classes. Then too, it is very difficult to predict which new B.A.s will prove most successful on the job market years later as new Ph.D.s. A recent winner of the MLA prize for the best first book, for instance, now teaching at a leading institution, was almost cut from our program at the end of the M.A. and was admitted to the M.Phil. only on probation, then went on to do spectacular work. Even with the final large classes coming on the market, our general placement rate has been consistently strong in recent years, with about 15 students placed annually in tenure-track jobs and in postdocs. The following table shows these placements (excluding one-year positions) for the past eight years. In order to calculate the percentage placed, we compare placements to the total number of degrees granted each year. A larger number go on the market each year, as people often make a trial run before they're really ready, and often take two further years on the market before they either get a job or turn to other careers. If we went by the total number of dossiers sent out each year, someone who gets a good job on her third try would rather oddly count as a 33% placement rate. The ultimate test of placement is what percentage of a given cohort gets placed, but as people in any cohort spread across a span of years to degree, that picture too is variable. Over time, the degrees granted gives a good basis from which to assess the numbers who complete and who get jobs or not. back to index Table 1: Degrees granted and job placement, 1997-2004   Degrees Placements % placed   1997 16 15 94%   1998 29 21 72%   1999 21 14 67%   2000 17 13 76%   2001 18 15 83%   2002 23 17 74%   2003 26 9 35%   2004 20 17 85% Averages 21 15 73% These numbers are for our final 30-student entering classes; we can expect our much more selective current group to achieve the rates of the more successful earlier cohorts. Realistically, we should probably anticipate that a few candidates each year will always, unpredictably, fail to find the right match: imponderables of student interests and geographical preferences, field fluctuations, and random chance will likely keep us at around a 75%-85% rate for as long as the overall job market remains tight. Anecdotally, we understand that in several recent years, we've done better than most of our peers, in percentage terms as well as in overall numbers, and we don't believe that during the past decade we've ever lost a prospective student to programs elsewhere on the basis of a better placement record reported by any of our rivals. A smaller sequential class would reduce faculty workload on orals and dissertations, but we don't find that on average this load is high, and it will decrease in the near future as we finish working with the remaining students from the old large cohorts. Our current sixth-year cohort began with 20 and now has 14; our fifth-year cohort began with 21 and now has 16. With our present M.Phil. class of 18, we are enrolling slightly fewer than one new sequential student per year for every two faculty members. This proportion seems satisfactory to the CGE, faculty and student members alike. We typically lose one student every year or two from each cohort, for around a 25% rate of attrition over 6 years, which is quite low by national standards (programs generally range anywhere from 35-65% attrition). With normal attrition in the middle years of the program, we are likely to average around 15 students beginning dissertations each year, and perhaps 13 or 14 completing them. If our "carrying capacity" of job-seeking students allows us to place 15 students most years, that number was certainly problematic for an M.Phil. class of 30, but would more than provide a 100% placement for a cohort of 13-14 dissertation students. Any failure in placement at this class size will result from the vagaries of the market and of student interests and success in interviews, rather than from numbers as such. If the overall number is reasonable, though, it may still be the case that individual fields are overburdened. This has certainly been true in the past, when some entire areas (notably American) attracted a disproportionate number of students, and even in other fields individual faculty members such as Steven Marcus and George Stade were directing as many as 30 dissertations at a time. This is a serious issue, and in preparing this report we have tried to develop a full picture of these variations, both by field and individually. The following table illustrates current practice, as seen in this past spring's offers and acceptances, grouped by the general field divisions we used in last year's admissions. Overall, this division reflects the CGE's concerted effort in the past decade to make offers in proportion to the number of faculty in each field, and not flood the faculty in any one area or strand those in any other. Table 2: Admissions for fall 2003   Faculty Offers Acceptances   Medieval 4 4 3   Renaissance 5 4 3   18-19c British 6 4 3   20c British/CL 4 4 3   Theater & Drama 3 3 3   Post-colonial 6 6 2   American pre-1900 4 3 2   20c American 6 5 2   TOTALS 38 33 21 back to index Though admissions are proportional to resources these days, students sometimes change fields after arrival, and it's important to see where they end up. Table 3 measures dissertation sponsorships and second-readerships for the current fifth- through eighth-year cohorts: Table 3: Dissertations by field, 1996, 1997, 1998, and 1999 cohorts   Faculty Dissertations Total     96 97 98 99   Medieval Crane, Ferrante.5, Hanning.5, Strohm, Yerkes   4   2   1   2   1   6 Renaissance Crawford, Eden, Howard, Kastan, Murray, Shapiro, Stewart   7   1   2.5   2.5   2   8 18c British Davidson, Seidel.5, Siskin   2.5   0   0   0   0   0 19c British Claybaugh.5, Dames, Kroeber, Marcus, Robbins.5, Rosenberg, Spiegel.5   5.5   3   2   1   1   7 20c British Cole, Damrosch.5, Golston.5, Heise.5, Mendelson, Rosenthal, Seidel.5   5   1.5   4   2.5   0.5   8.5 Postcolonial/CL Negron.5, Damrosch.5, Heise.5, Robbins.5, Slaughter, Spivak, Viswanathan   5   1.5   1.5   2   2.5   7.5 17-19c American Arac.5, Blount.5, Claybaugh.5, Delbanco, Tawil   3.5   1   0   1   0   2 20c American Adams, Arac.5, Blount.5, Douglas, Golston.5, Griffin.5, O'Meally   5   2   3   3.5   0.5   9 Theater & Drama Peters, Puchner, Quigley   3   0   1   1   1   3                      TOTALS   39.5         51 With each dissertation needing three departmental readers, these dissertations involve a total of 153 readers, or an average of c. 38 per year during these four years, i.e. one new dissertation committee membership per faculty member per year. (An additional 11 defended dissertations aren't shown on this table, but given our larger earlier cohorts, the listing of current dissertations gives a closer approximation of numbers in future.) The 20th-century fields tend to be somewhat above the norm, each carrying an added total of about two dissertations among these four cohorts, beyond what would be typical for their faculty numbers. It appears that both the British and the American 20th century fields took on students from the 1997 and 1998 cohorts during the period of high turnover in the 18th-19th century fields. As can be seen, 6.5 of the 20th-century British total of 8.5 dissertations have come from those two cohorts, as have 6.5 of the 20th-century American total of 9. Most recently, though, both fields show a below-average 0.5 figure for dissertations among the 1999 cohort. We can expect the 18th and 19th century fields to both attract and also retain more students with our renewed faculty numbers in these fields. Overall, these numbers mark a welcome contrast to the situation that prevailed well into the 1990s, as no fields are dramatically overburdened these days relative to other fields. Most notably, the Americanists, who previously carried a third or more of all our dissertations, now have a number closely proportional to faculty numbers: 21.6% of the current dissertations shown above, and 21.5% of current faculty. In sum, the present size of the program works well for us and should retained. Recommendation 1: Our sequential program should stay at its present size of 18 incoming students per year, i.e. an average of 2 students for each of the 9 fields listed above. The CGE should continue to admit in overall proportion to faculty numbers by field, with modest adjustments up or down in light of the strengths of a given year's pool and the previous year's yield, and making offers conservatively in the 20th century fields. We should monitor placements over the next two years, and revisit the question of program size if placements decline for any length of time. As important as overall numbers is the workload of individual faculty members, which often varies widely within fields. Variations in workload are natural, given that some faculty thrive with a large number of advisees, while others work best with fewer, and some devote themselves primarily to our undergraduates. It's neither feasible nor desirable to mandate any common load. Yet there is no reason to carry on reflexively with unwelcome imbalances, or to have only a vague idea that they must exist but no hard information. We are still in the process of gathering full information on faculty workload, but on the basis of our dissertation list, it's possible to group people broadly by a low, medium, or high involvement in dissertation sponsorship/second-readership. According to our full listing of current dissertations, no faculty member is now sponsor or second reader to ten or more active dissertations, so the huge imbalances of past decades have reduced considerably. Even so, there remains a significant range. Of our current total of 42 departmentally-based faculty, about half (22) are serving as sponsor or second reader for no more than two current dissertations, another 10 are advising three to five students, and 8 people are advising six to nine. These numbers can be adjusted once we have a listing of third-readerships and a full listing of out-of-department sponsorships, but the current totals give a general idea of the range we have. Even with a good balance in admissions decisions, imbalances within fields are likely to persist. At any given time, a majority of students in a field may be drawn to certain approaches and issues, not always in ways that can be predicted at admissions time. When there are imbalances of this kind, how should they be handled? Decades ago, there used to be a "point system" that calculated these loads, as well as teaching and administrative burdens, and gave some symbolic or financial rewards. From the perspective of the graduate program, we have a suggestion that could be of substantive value for heavily involved faculty advisers and also of benefit to upper-year graduate students. One real loss that occurred when we shifted from a second-year class of 30 to one of 18 was the ability to assign graduate-student research assistants to most faculty. Some work-study students are available, but they are usually undergraduates who don't have developed knowledge of a scholarly field and they may not be able to do more than basic photocopying and library searches. It would be good if this situation could be ameliorated. Recommendation 1.2: Faculty with unusually heavy loads of dissertation advising should be given research assistants hired from post-orals students. We expect that a good number of upper-year students would welcome both the added income and the ongoing connection with a faculty adviser. We could pay for these research assistants out of the funds given to the department in years of overage on M.A.O. admissions. An excess of one M.A.O. student over our target of 13 brings the department $15,000 in unrestricted funds, which would pay for ten research assistants at $1500 each for a one-semester appointment, working 5-6 hours per week. We've enrolled 15 M.A.O.s this fall, and in addition have a few part-time M.A.O.s still finishing up from last year's unexpectedly huge M.A.O. class of 30 (a surprise that caused us to control admissions much more carefully than in the past, which is why we're back about where we should be this year). We have a total of 17.5 f.t.e. M.A.O.s this year, which should yield us a total of $67,500 in unrestricted funds. Even if we have no further overages in the near term, these funds could serve this purpose for the next several years. The number of research assistants would depend on available funds, and would be assigned in priority according to total workload, to faculty who request them and who don't have research assistance already available from other sources. We could begin with eight per year for the next five years, then adjust as necessary thereafter, a number that should provide assistance to the upper range of advisers. It is entirely reasonable that money generated from graduate student tuition should be spent on added support for our students, and we think there is a logical fit between faculty involvement in the graduate program and the allotting of graduate student research assistance. This plan could also be one component in a broader consideration of faculty workloads, and responses could include course reductions and salary increases, though that broader question goes beyond the focus of the present report. back to index 2: The M.A.-only Program To get a full picture of our graduate responsibilities, we need to look closely at our M.A.-only program as well as our sequential cohorts. Until 1990 we had a de facto M.A.O. track, as we regularly denied most of our M.A. students admission into the M.Phil. - we just didn't decide who the M.A.O. students were until year's end. In reformulating our program in 1990, we decided to have an explicit M.A.O. option for people who might want it. The M.A.O. program brings us several benefits. It enables us to mount a larger graduate course offering than we otherwise would, without the problems of upper-year advising and work load that would be entailed with a larger sequential program. M.A.O. students account for roughly a quarter of our graduate enrollments (as the larger number of sequential students take courses for two years and then often some further courses as they prepare for orals). These added enrollments allow interested faculty to offer a graduate course every year, giving a greater range of options for students to choose from. Second, our M.A.O. program attracts an interesting group of students thanks to our metropolitan location. Some are people who had majored in other fields or who have been out of school for a number of years, and who want to do an M.A. in order to prove themselves and get current before applying to Ph.D. programs in English; others are high school teachers or people in publishing who want exposure to the serious scholarly study of literature, and collectively our M.A.O.s bring a range of perspectives to the program. Finally, the M.A.O. program provides critical financial support to the sequential program. Our cohorts of 13 M.A.O. students all pay full tuition, currently $29,640, for a total of nearly $400,000 per year - an important source of funds that helps underwrite our expensive sequential program, whose stipends alone currently total about $2,400,000 per year. We have discussed whether we might offer a separate track or set of courses for M.A.O. students, but our students have no single common base, as some plan to go on to Ph.D. programs and others don't; apart from the fact that some would benefit from more flexibility in course requirements (to be discussed below), they tell us that they want to be full participants in our regular program. The selectivity of our M.A.O. program has improved since its inauguration, from around 50% at first to 30% in recent years. For the current year, we received 96 M.A.O. applications and accepted 28, anticipating about a 45% yield on our offers, to achieve our goal of 13 students. As it turned out, we had better than a 50% rate of acceptances, with 15 enrolling this fall. Almost all of our M.A.O. students these days are doing very well in our classes, but there is certainly a considerable range of preparation in literature among this group, and they have differing levels of engagement with advanced scholarly research. Further, while the CGE accepted only students we affirmatively wanted, we found a rapid drop-off in quality below the final admit list, and we would be very reluctant to go further into the pool. Recommendation 2.1: On reviewing this information, the CGE recommends that the M.A.O. program be maintained at its current size. The applicant pool doesn't permit us to expand without lowering the quality of our program and diluting its focus on preparing future scholars, while the current size is needed to preserve the present scope of our graduate offerings and the full funding of our sequential program. Our M.A. program is overseen by the M.A. Director, who holds regular meetings for the M.A.O.s and for M.A. students generally, and who advises M.A. students throughout the year, in addition to serving on the CGE. Almost always a junior faculty member to date, the M.A. Director has served with no compensation apart from some added research assistance. A modest course relief for this position would be appropriate, and would be consistent with department practice regarding other administrative posts. Recommendation 2.2: The M.A. Director should receive a one-course reduction in teaching for each year of service in this post. This could be set as a course credit for organizing and running the fall series of lectures on theory and method. In earlier years, when the program was less selective and we were often ending up below target size, we admitted students on a rolling basis (now eliminated), and we also allowed for midyear enrollment, with people beginning the M.A. in the spring semester. This option continues to be offered, yet these out-of-cycle admissions have proved to be problematic in various ways. First, they exert an upward pressure on the size of the M.A.O. program, as it's hard to refuse good applicants for spring if we'd already admitted somewhat less accomplished students for the fall. Second, the few midyear-admission applicants aren't being seen together with the full applicant pool, and it's harder to assess them comparatively. Third, students admitted in this way come in out of step with both the preceding and the following M.A. class. This was less of an issue when the department offered M.A. seminars both semesters, but these days the M.A. seminars are only given in the fall, leaving mid-year enrollers with no common M.A. course in their first semester. Recommendation 2.3: Given the disadvantages of off-cycle admissions, we propose eliminating them. Application to the M.A.O. program should be made on a single cycle for the fall of the upcoming year, with no separate provision for beginning the program mid-year. The mid-April application deadline is working well for us, and should become the deadline for all M.A.O. applications. Once they are in our program, our M.A.O. students encounter a recurrent difficulty getting into seminars. This can be a severe problem, as they need seminars in order to graduate. At the same time, it may be that not all our M.A.O. students really need to take the proportion of seminars we expect our sequential students to take (half of total coursework). M.A.O. students who are preparing to go on to Ph.D. programs should be advised to take four seminars out of their eight courses, but others of our M.A.O. students may be served as well by a higher proportion of 4000-level lecture courses. A more flexible course requirement would enable M.A.O.s to tailor the program more to their needs. Recommendation 2.4: We propose affirming that - like M.A. students in general - M.A.O. students should be fully admissible to our courses, and seminar spaces should not be reserved solely for sequential students, or solely for M.Phil. students. At the same time, recognizing that not all of our M.A.O. students have the same degree of need for advanced research work, we propose modifying the basic course requirement for M.A.O.s to specify a minimum of two seminars, instead of the four seminars required of sequential M.A. students. We know that ninety students or more a year ask to join our program as M.A.O.s and that over half of those we admit do come, an unusually high proportion for an unfunded program. What we haven't known in any organized way is whether the program does seem to have been worthwhile to them as they complete it, and whether they are able to proceed on from here in satisfactory ways. Recommendation 2.5: Building on last spring's survey, we should survey our graduating M.A.O. students each spring to assess their satisfaction with the experience and their plans thereafter. This should help us give applicants a better sense as to what they can expect, and can tell us what sorts of students are best served by being here for the M.A. At the same time, we should collect their post-Columbia e-mails, so that we can follow up on their later progress. Having outlined the overall shape of the graduate program, we turn to its specific requirements. back to index 3: The M.A. and the first year of the M.Phil. A: Course requirements and writing requirements in courses: We require a total of 16 courses over these two years, whereas the typical load elsewhere turns out to be 14. Often the 14 consist of three literature courses a term for the four semesters, plus course credit for the M.A. essay and for a teaching practicum. At present, many of our own students in fact do a load of this sort, making up the remaining couple of courses in the third year. Those who do the full 16-course load in the two years usually lag behind in working out their orals fields, while those who are still doing courses in the third year run behind for that reason. Recommendation 3.1: Our MA/M.Phil. courseload should be reduced from 16 to 14 courses. The issue of course work concerns students' research and writing as well as the number of courses, and different schools vary in how much writing they expect from students in courses. Yale and Princeton, with small programs, assume that students will be taking all seminars and will do three seminar papers a semester. Chicago assumes that roughly half of the courses will be seminars, half will be graduate lectures, so that students would take one or two seminars per term. Students aren't required to do research papers for all seminars; they simply designate one seminar per term for which they write a research paper. Harvard doesn't have graduate lecture courses at all, but allows graduate students to take undergraduate lectures for credit, only requiring 2 graduate seminars per term. Writing only one seminar paper per term seems rather little; three is more than most programs require. Recommendation 3.2: Sequential students should be required to take two seminars per term. Students who wish may choose to take more seminars, but they may fill out their program with graduate lectures. So as to allow the students more time to concentrate their research in their prospective field(s) of study, the M.A. seminar as well as the Introduction to the Discipline lecture course should ask for short written assignments rather than extended research papers. Recommendation 3.3: Students who take a third seminar in a single semester may choose to write a research paper for that course as well, but with instructor's permission - and only with instructor's permission - they could also have the option to do presentations and other work but not a final paper, receiving a grade on the model of a 4000-level course. The instructor should inform the office of any such arrangement, and the course would count toward the degree but not to our required number of seminars. Recommendation 3.4: A student who hasn't found two desirable seminars in a semester, or who has applied but not been admitted to two, may write a research paper for a lecture course, with the instructor's agreement (conveyed by email to the office); it would then count toward the required number of seminars. back to index B: Course advising: We have tended to take a laissez-faire approach to course selection, assuming that our students will survey their options and take what they like. Our survey last spring, however, revealed widespread dissatisfaction with the lack of advice on courses. Our advising gets weak reviews at all levels, but by far the lowest marks go to course advice. Fifty students replied to the question "How would you describe the advising you receive in choosing your courses?" Of these 50, exactly zero listed "excellent," only 9 listed "adequate," 14 listed "somewhat adequate," and an outright majority of 27 listed the bottom category, "inadequate." In response to this finding last spring, we have instituted an assignment of incoming students to advisers, and we expect this has helped somewhat, but more can be done. Recommendation 3.5: We should set the adviser for each of our incoming students at the time we first recruit them and receive their acceptance. Faculty in each field should decide who will take the lead on recruiting each student and serving as the incoming adviser. Advisers for incoming M.A.O. students should be assigned by the DGS before the time of orientation. Recommendation 3.6: The CGE should develop a basic set of guidelines to clarify for students and for faculty what should be their mutual expectations as adviser and advisee. Recommendation 3.7: We should expand our student-to-student mentoring system. Currently incoming students are each assigned an upper-year student as a guide for their entry into the program. This mentorship should extend through the second year. back to index C: Registration and admission to courses: Columbia's move to pre-registration in the past few years has enabled faculty to anticipate the size of classes and order books, and helps departments secure rooms of the right size for lecture courses. These benefits, however, largely apply to undergraduate lecture classes; graduate students find disadvantages outweighing the advantages. With pre-registration occurring the semester beforehand, every year incoming M.A. students and advanced admits find that many seminars are full even before they arrive on campus. Students in other departments and in the other schools in the New York City consortium are often caught off-guard by our early pre-registration deadlines, and faculty are selecting their classes without a full knowledge of who would like to take them. Recommendation 3.8: Applications for graduate seminars should be moved to the week before classes begin each semester. Faculty could either set their class list at the start of the first week of classes, or could make decisions at the time of the first meeting. Faculty who are out of town during the week before classes begin could take applications via e-mail, or could select students in the first week of class. Recommendation 3.9: While we have no set minimum or maximum seminar size, instructors should not cap seminars at fewer than 15 students if more than that number of qualified people apply. We believe this is already the general practice in the department, and that it should be spelled out as a matter of policy. back to index D: Course evaluations: We currently evaluate our students in two ways: through grades, and through yellow and blue cards for seminars and lectures, instituted in the days in which grades weren't given in graduate courses - and when entry from the M.A. to the M.Phil. was sharply competitive, the Ph.D. program itself was only partially funded, and the CGE needed to make very fine distinctions each year. We don't find evidence that other schools use the equivalent of our yellow and blue cards, and while they are occasionally used in the office as a record of past work, they are no longer used for their primary purpose, CGE evaluation. Their very modest present uses are outweighed by their generally counterproductive effect: students worry about them if they aren't done on time, and anxiously parse them and over-interpret their phrasing once they are available to read. Much better for us to tell students directly what we think of their work, and to use grades openly as a formal indication of its quality. In the rare instances in which a student is falling seriously short of expectations, this should be communicated directly to the DGS as well as to the student. In sum, the yellow and blue cards are the very type of the archaic requirement surviving as a ritualized practice years after it has come to have a largely archaeological value. Recommendation 3.10: We should abolish the evaluation cards outright. Unused yellow and blue cards may be donated to the Lamont-Doherty fossil collection. Individual faculty, however, are strongly encouraged to keep a record of their evaluations of students' work. Typing up end-of-term comments can give a student a fuller response than a short comment on a paper; and saving a copy of these comments in a file can make life much easier for the faculty member when it comes time to write letters of recommendation later on. back to index E: Grades: Until 1990, we gave no grades in our graduate courses, believing that students should receive detailed feedback in their seminar work, not mere grades; they would also take lecture courses as reading courses with no set requirements that could be graded. We found, however, that grades could have value for giving clearer indication of what narrative comments finally indicated, and we also found that the lack of grades could disadvantage students in fellowship competitions. Accordingly, in 1990 we shifted to requiring grades in all courses, apart from R credit for the preparatory course for Undergraduate Writing. This change has been useful, but the meaningfulness of grades may be somewhat undercut by the fact that most grades are A's, especially in lecture courses. Certainly the high proportion of A's reflects the general quality of our students' work, but it may also be that with the legacy of years of a highly competitive program fresh in people's minds, faculty may fear to give an A- lest they damage a student's career in the program. It appears that sequential M.A. students have a fear that even a couple of A- grades may cause them to be eliminated from the program, though this is not the case. Recommendation 3.11: We should clarify the meaning of our different grades. We propose affirming that both the A and the A- are truly positive grades. Grades of A- are not cause for expulsion from the program or indication of lack of satisfactory progress, but simply register good work that can be taken a step further in future. No doubt some students always produce presentations and papers of straight-A quality, but it does no favor even to very good students to act as though they are always at their very best, even if they sometimes aren't. A one-third proportion of A- grades in a seminar would be a perfectly reasonable result. By contrast, a B+ grade would signal work that raises concerns, and in the case of an M.A. student a pattern of B+ grades would indicate someone who shouldn't go on in the program unless he or she is doing better work in other courses. The rare grade of B or lower would signal an active recommendation that the student not go on, or in the case of upper-year students would indicate some difficulties that should be discussed with the DGS. back to index F: The M.A. seminar and lecture series: Our present pattern of three fall-term, theory/method-oriented M.A. seminars is working well overall, and no general wish has been expressed to return to our older requirement of a master's seminar each semester. The M.A. lecture series has been going very successfully for three years now as a biweekly series in the fall, and students appreciate the chance to see a range of faculty, as well as to bond over pizza. Our survey has shown, however, that many students would like the three seminars to be better coordinated, and to be more specifically geared to the purpose of introducing students to graduate study. Too often, they feel, our M.A. seminar offerings have been hard to distinguish from other seminars the same faculty might offer as 6000-level courses. Our students would also like the lecture series to be genuinely integrated with the seminars; as of now, the lectures often don't relate in any clear way to the seminar's work. Recommendation 3.12: The M.A. seminars should be better coordinated and geared toward their purpose as introductions to graduate study. They should have a broad focus on theory and method (not on a single author or specific strand of theory), and should incorporate practical issues such as advanced research methods and ways to make the transition from writing seminar papers to writing articles. Several short assignments should be given, rather than one research paper. Each fall's lecture series should be established early in the preceding spring, in discussion the upcoming M.A. seminar instructors, who should meet together to coordinate their courses with each other and with the lecture series. The lecturers should be invited to assign an essay for each talk, and these should be integrated with that week's M.A. seminar readings. The three sections should have some common base of readings and issues, along with whatever specific emphases each instructor wishes to give. We've set up the lectures as a series for the M.A.s, but perhaps incoming advanced students should be invited too. It's true that they are more likely than the M.A.s to come in with a developed knowledge of theory, but many of our M.A.s already do too, and yet they appreciate the chance to hear faculty talking about issues of theory and method in a personal way. Indeed, this series can have special value for new advanced admits, as they won't have two years' worth of course work here and can use the early exposure to the range of faculty. Recommendation 3.13: We should rename the lecture series the Colloquium on Theory and Method and open it to advanced admits as well as to M.A. students. Including the term "method" in the heading allows for talks beyond the field of literary theory per se, such as David Kastan's very successful talk this fall on the history of the book ("the New Boredom") as an emergent field of study. back to index G: The M.A. essay: This seems to be handled just as it should be. No one seems inclined to return to our older pattern of two M.A. essays and a written M.A. exam. The current, article-style length of 25-30 pages gives good practice in writing at a publishable scale, and does regularly produce published articles after revision. Our 1/1-/2+/2/3 grading system, however, is anomalous, instituted decades ago when letter grades weren't given in courses, and it can be confusing outside Columbia. Recommendation 3.14: M.A. essays should be given letter grades in the same way we grade in seminars. The sponsor's grade will be recorded on the student's transcript as the grade for the M.A. Essay Tutorial. back to index H: Introduction to the Discipline: Inaugurated this fall as a first result of our Carnegie discussions, this course surveys intellectual and institutional trends in the development of literary studies from the late nineteenth century to the present. There is general agreement that this course should be continued, but some M.A. students regretted having to take both it and the M.A. seminar at the same time, and this is one area where our M.A.O. students probably don't need to be required to do everything the sequential students do. Advanced admits, meanwhile, say that they are thrown on their own upon arrival; the disciplinary course could make an excellent common course for beginning M.Phil. students. Recommendation 3.15: This course should be renamed "The Discipline of Literary Studies" and should become a requirement for first-semester M.Phil. students. As this year's M.A. students have just taken it, we would skip it next fall, then resume it thereafter. Students who had taken an equivalent course elsewhere could be excused. back to index I: 4000-level courses: Small programs usually have only seminars for graduate students; large programs either have some lecture courses, or allow graduate students to take undergraduate courses for credit (as is the practice at Harvard). Our 4000-level offerings have ballooned or shrunk dramatically over the years, often being used as a stopgap simply in order to mount our full curriculum at both levels in periods and fields of short staffing. These days, with our smaller graduate program, most of our 4000-level courses are largely undergraduate in enrollment. This fall, our W4000 courses (the category freely open to undergraduates) averaged no less than 92% undergraduates. Even the few G4000 courses, which require instructor's permission for undergraduates to enroll, averaged 55% undergraduate enrollment. Of eleven 4000-level courses this fall, only one had a majority graduate enrollment, and even it had 45% undergraduates. Recommendation 3.16: We should offer 4000-level courses when they serve a genuine pedagogical function as an upper-level lecture, but not simply use them to plug holes in the curriculum. Faculty who teach them should keep in mind that the majority of students enrolled will likely be advanced undergraduates and should make sure that the level of discussion and the written assignments will work for them; at the same time, it is appropriate to exclude undergraduates who aren't prepared for a relatively advanced discussion. Graduate students should be encouraged to see these courses as genuine courses for them, and should be able to expect readings and a presentation beyond the level of a typical 3000-level course. back to index J: Course distribution and planning: We currently have fairly typical distribution requirements: one pre-1700 and one post-1700 course during the M.A.; by the end of M.Phil. coursework, at least one course each in four out of five areas (medieval; early modern; British/Comparative 1700-1900; American pre-1900; twentieth century). This seems adequate, but it does somewhat oddly resemble the sorts of distribution requirements used for undergraduate majors, and our four-of-five choice allows students to skip a historical period, such as medieval or early modern. Should our distribution requirement be more coherent for graduate students, and should it serve a distinct purpose at this stage of work? Several students favor distribution requirements as helping prepare them to teach Beowulf-to-Virginia-Woolf-style survey courses. From that point of view, it would be better to require an exposure to each broad period, rather than having the discontinuity implied in our current requirement. Recommendation 3.17: By the end of M.Phil. coursework, students must take at least one course in each of the following four historical periods: medieval; early modern; eighteenth-nineteenth centuries; twentieth century. A further issue is planning: Chicago urges all first-year students to develop a tentative two-year program at the outset. We've typically planned only year-to-year, and sometimes a seminar will be offered two or more years in a row, limiting the range of topics a student can study during the two years of coursework. Recommendation 3.18: As has been done this year for the first time, faculty "druthers sheets" should have a two-year time frame, and as courses are set for the upcoming year, we should also develop a tentative course plan for the ensuing year as well. This will always be subject to change, but it can help ensure that we give students a coherent and varied program across the two years they take courses. With the exception of required courses, graduate seminars should not be offered two years running. Our survey of students found a mixed response to our course offerings. Students are excited to take courses based in the ongoing research interests of the faculty they came here to work with; yet they also register a broad concern that our offerings are weighted too heavily toward our own research preferences, and give them an insufficient training in their field as a whole. Only 13% of our respondents described themselves as "very satisfied" with the course offerings, while 35% were "satisfied," 37% were "somewhat satisfied," and 15% were "dissatisfied." More people, in other words, chose the lower two categories than in the higher two. We've asked our students in recent years to tell us what they'd like to see offered, and have usually added one or two courses a year in response, but this modest level of responsiveness has not seemed overwhelming to them. Asked about graduate students' role in determining course offerings, of those who had an opinion, no respondents at all listed the highest category, "just right"; a mere 12.5% listed the next category, "adequate," 18.5% listed the third choice, "somewhat adequate," while a large majority, 69%, listed the lowest category, "inadequate." At the graduate level, for students who are committing themselves to our profession, we can surely do a better job of consultation as we seek a good meeting of our interests and their needs. Recommendation 3.19: Next fall, we should hold field planning meetings to discuss possibilities for the following year. Sequential M.A./M.Phil. students should be invited to these meetings, which would have the further benefit of helping meet the students' often-reiterated wish for more contacts of an informal nature with faculty. These meetings would help faculty gauge which of their ideas would likely have the best response, and would provide a direct chance for students to talk through ideas, and refine them in discussion with faculty. If these meetings prove to be successful, they should be instituted on a regular basis in subsequent years. back to index K: Language requirements: Our students rate language skills remarkably highly: 38% of our students describe the acquisition of a second foreign language as "essential," 27% as "important," and only a minority as "somewhat important" (29%) or "not important" (6%). Few thought that our language requirement should be reduced (15%); many think it's as it should be (41%), and yet the largest number (44%) would actually like to see it strengthened, reflecting a frustration with acquiring only a minimal competence in two languages. Several express a wish, accordingly, for greater flexibility in fulfilling the requirement. Many schools offer a dual option: students can demonstrate basic reading knowledge in two languages (our sole option), or can acquire advanced knowledge in one language. As one student reported: "The language requirement is both annoying and ineffective. I passed the Spanish exam immediately upon entering graduate school, but when I attempted to continue this training - because I actually wanted to read and speak the language! - I was told that my passing the exam prohibited me from receiving summer funding to continue taking Spanish courses. Why not change the requirement so that speaking and writing in one language becomes the goal?" This seems entirely reasonable, though a basic knowledge of two languages may be best for some students. Recommendation 3.20: We should continue to require that M.A. students demonstrate a basic reading knowledge in one language (the ability to translate accurately with use of a dictionary). Thereafter, we should offer a dual option: by the time they take the doctoral orals, students must demonstrate either: (1) advanced reading knowledge of that language (the ability to translate accurately without use of a dictionary); or (2) basic reading knowledge of a second foreign language. Most schools stipulate that these languages must ordinarily come from a short list of classical and Western European languages (Greek, Latin, French, German, Spanish, and Italian); we do have that stipulation on the books, but in practice we simply expect students to be able to say that the languages they present will be useful to them in their later work. Given the comparative dimension of many students' work, it makes sense to continue allowing them to offer whatever languages they believe will be most useful to them as scholars. back to index 4: Doctoral orals The present system works reasonably well, but it has some recurrent problems. First, students take a long time formulating topics and then preparing them. Most other programs expect students to do their orals by the end of the third year at latest, whereas many of our students take them in the fourth year or even later. One reason for delay is that basic uncertainty exists among faculty as well as students as to what the orals should really be. This uncertainty breeds two different problems: unsure what to include, students often overstuff their lists, making them hard to get through in a reasonable amount of time. Conversely, unsure what the orals should do other than lead to the dissertation, students sometimes construct quite narrowly focused lists, for example consisting almost entirely of a single genre (usually the novel), or a single fifty-year period, or a small constellation of issues across all three fields. Indeed, both problems can occur together, yielding lists that are unwieldy in bulk but narrow in conception. A: Definition and structure: Our peer schools display various models for the orals, some looking much like ours, some with two rather than three fields, and in one case (Yale) with no fewer than eight fields. We believe it works well for us to have three fields involving a total of four examiners; no wish has been expressed to reduce the number of fields, or to expand to our former number of four fields with a total of five examiners. Our peer programs, though, are generally clearer than we have been that the orals serve a double purpose: both to help students develop a dissertation topic and also to have them achieve a broad overview of a general field in which jobs get advertised and they'll likely be asked to teach survey courses. Recommendation 28: We should put forward a better definition of the orals. Building on our threefold orals structure, we can say that the orals should do three things: prepare students as prospective teachers to master a field - defined pragmatically as an area in which jobs are commonly advertised - and to talk about it with a group of examiners; serve as a heuristic device to help the student develop and/or refine a dissertation topic; give students a solid grasp of a distinct but related field beyond their primary one (such as the period before or after that of their primary focus, or the literature written across the Atlantic or on the Continent). Recommendation 29: We should give the orals a stronger and clearer structure, so that students and faculty will have a better idea of what they are setting out to do. We propose defining our three fields as follows: an hour-long "general field" surveying a job-related field; a half-hour "related field" in a contiguous period or another clearly distinct but useful area; and a half-hour "thesis field" devoted to works bearing directly on the likely thesis topic. (Note that these terms avoid the invidious yet vague distinction between "major" and "minor" fields.) Since many students don't yet have a thesis topic as they formulate their orals, in place of the "thesis field" they could substitute a second related field or a "single author" field, focused on a full study of someone who might otherwise appear selectively in the general field, or who might come from another period altogether. As Andy Delbanco has argued in print, it is a distinct and valuable enterprise to study an author's full work and the body of criticism around that author. back to index B: Guidance on forming lists: Even when students readily identify what their fields should be, they often spend a lot of time working up their lists from scratch, in effect re-inventing the wheel, though many students before them may have done quite similar topics, and the final list may not even look very different from ones before it. We established a folder of prior orals exams some years back, but students report that they find it hard to wade through them and really decide what's most important. Further, they are wary of imitating a fellow student, and are unsure how well a prior student's list reflects general parameters of the field. If we define the general field as the equivalent of an extended survey course in a job-category field, it may be possible to develop some basic guidelines for the most common general fields, and students could use these lists as starting-points and make choices and additions in light of their own particular interests. Recommendation 4.3: Each of the fields in the department should establish a subcommittee of faculty and students to develop general-field guidelines for the field, either as an overall list or as a set of grouped readings on important subfields. These lists should then be discussed and refined by interested faculty and students, and then set as points of departure for general-field list preparation. These lists are intended to be suggestive rather than prescriptive, highlighting key authors and works most likely to be looked for by hiring committees in the field. Among other purposes, they should alert students to the importance of being familiar with a range of genres, even if their own work focuses largely on one. Students would use these lists selectively, and they would add works of their own choice as well, and they could always propose a field for which no lists had been developed, but we believe that many students would find these lists very useful, guiding them in shaping their orals and giving a direct indication of what such an overview might look like to our current faculty in the field. back to index C: Timing: With this assistance, we should be able to expect orals proposals to be developed in a timely fashion. At present, students often don't submit them until late in the third year, or even later than that, delaying their progress to little or no final benefit. Recommendation 4.4: We should require that students have individual discussions with the DGS and with at least two potential examiners in the spring semester of their second year. The completed orals list should be submitted by October 1 of the third year. Extensions could be given when needed, but this deadline would insure that students have a direct discussion and get advice if they are having difficulties formulating fields or selecting examiners. back to index D: Orals advising and preparation: Once the orals list has been approved, students (and even faculty) are often unsure just what level of engagement is expected and most useful. As one student writes, "It's hard to know what one should expect or demand from advisers, since students' experiences vary so widely." Vagueness can come across as indifference: according to another respondent, "My impression is that no one seems to care all that much about the orals, except the stressed student." Different students of course will want and need differing amounts and kinds of advice, and will prepare in different ways, while different faculty may also have differing ideas on preparation and their own roles as advisers; but it would be a great help if some overall sense could be made clear, at least as a starting-point. Recommendation 4.5: The CGE should develop a set of guidelines on orals preparation, so as to give both students and faculty a better sense of the range of "best practice" employed in the department. These guidelines should include discussion of such matters as frequency of meetings, the writing up of short papers or other modes of focusing analysis, and expectations for the written pre-oral. The CGE should circulate a draft for faculty and student comment before the end of this semester, and a revised draft should be presented to the faculty for discussion. Once a final draft is approved, it should be made into a handout to give to each student and each faculty adviser who signs on for an orals. We know that our students learn at least as much from each other as from their teachers, but the orals phase is typically one of isolated work. Some ongoing form of connection would likely improve the quality of the work and students' confidence as they approach the orals date, and could help reduce the number of nervous students who postpone their orals. Recommendation 4.6: On analogy to our successful dissertation seminars, we should invite the Graduate Student Council to establish a pair of orals discussion groups, to be led by a post-orals student mentor or pair of mentors, who could be paid a small honorarium for hosting a monthly discussion. As we'll only have 18 students preparing for orals at a time, numbers are too small for separate discussion groups in each field. We could establish one pre-eighteenth-century group and one eighteenth-twentieth-century group. back to index 5: The dissertation A: Drafting and advising: While our students clearly produce first-rate dissertations, it often takes them a long time to do so, and the lack of organized faculty guidance surely contributes to this. Most other programs, in fact, have much clearer procedures than we do to ensure that students have meaningful faculty support as they formulate their topics and develop their prospectuses. Our program stipulates that the prospectus is to be completed and approved within six months of the orals, but students often run a year late in reaching this point. Once the dissertation is under way, students take far longer to complete than our program expects. At Columbia as elsewhere, time to completion is very high, and it hasn't dropped with improved funding; our nominal six-year program is in fact an eight-year program for many active students. Our current eighth-year cohort, which began with 26, still has 16 students this year, and while a few of these have effectively finished, many have not. Many factors contribute to this slowness, including the inadequacy of our "full" funding and the uncertainties of the job market, yet as advisers we surely bear part of the responsibility. Our students report severe difficulties in receiving timely feedback on their work in progress. Perhaps no single slippage is as common in our program as a lag of two, three, or even more months between the receipt and the return of draft chapters. During such months, students often put off serious work on the next chapter, or else they go ahead without the benefit of guidance on the work already done, and they may go ever further down unproductive paths that could have been avoided by timely advice. This is particularly perverse, as the delay gains the faculty nothing; it takes the same number of hours to read a draft two weeks or two months after receipt, yet the difference is enormous to the student. As one student put it, "The GSAS guidelines of two weeks [for return of drafts] are a joke. I've waited months, often over a semester, to get comments on a chapter." As another student says, students as well as faculty need better guidelines, to clarify the responsibilities on both sides: "I believe that the lack of structure and support at Columbia is the greatest obstacle towards progressing. At Columbia, one can often feel that if one disappeared from the program no one would care, or even notice." There are a variety of things we can do to improve our own effectiveness as mentors throughout the dissertation process. We recommend adopting the following methods, most of which are already being used elsewhere: Recommendation 5.1: At the end of every successful orals exam, following a brief break during which the examiners confer, the student and examiners should have a twenty-minute conversation about current dissertation ideas and plans. (A student who wishes could be allowed to defer this conversation for up to two weeks; if so, its date should be set at the conclusion of the orals, while everyone is still in the room to arrange it.) Recommendation 5.2: To embody our collective experience on dissertation development and help students feel less in the dark as they start the process, the CGE should create a short pamphlet on prospectus and dissertation writing, drawing on published articles on the subject and on local practice. This document would give students guidance on various ways that dissertations can best be structured, researched, and written, including discussion of the appropriate range of length, possible chapter organizations, balance of "primary" and "secondary" discussion, anticipated audiences, and relation to field and job market concerns. This pamphlet should be given to each student at the time they begin preparing for orals. Recommendation 5.3: Not later than October of the fourth year, the student should have a half-hour conference with at least two of the dissertation committee members to discuss a draft prospectus (or progress toward a draft). In cases where these people aren't yet identified, the conversation should be with the DGS and one other faculty member. Recommendation 5.4: If the prospectus isn't in fairly final form at the time of the October meeting, a follow-up meeting would take place to review the finished prospectus, in January or early February of the fourth year. Recommendation 5.5: By June 1 of the fourth year, the student should have a one-hour conversation with the dissertation committee, discussing a first chapter or the progress toward its completion. Anyone who is out of town should if possible join the discussion by phone. The office should keep track and make sure that the meeting is being held, and the DGS should receive a brief email report from the committee's chair following the meeting. Recommendation 5.6: Students and their readers should agree to certain basic parameters of their work together. In order to make satisfactory progress on the dissertation, the student must visit each of the three readers at least once per semester to discuss progress and plans; they should be in contact via email if a direct meeting isn't possible. And in order to satisfactorily meet the responsibilities of service as readers, faculty will make every effort to return draft chapters, with substantive comments, within the two weeks mandated by GSAS if feasible, and in any event within a month. Recommendation 5.7: When possible, students should drop off drafts in person with each reader and discuss any lingering questions or problems with the draft. They should put a full date on each chapter draft, and should include their e-mail address at the top of the first page. This will help the reader keep in mind how long the draft has been waiting to be read, and will make it easier to respond. Recommendation 5.8: To build in timely feedback on chapters as a normal and expected part of our work with students, we should have students notify the office (the person in charge of orals scheduling) when a chapter is being handed in. The staff person should then email the three committee members and set up a time for them to meet with the student either three (or if necessary four) weeks later. We propose reserving the 1:00-2:00 time slot on Thursdays (a time normally free of classes) as the standard time for these meetings. The meeting would include discussion of the submitted chapter and of the next stage of work to be done. As faculty already meet individually with students to discuss completed drafts, these meetings would add little or no time for the faculty, but they would yield much more coherent and useful advice for the student - which would in the end produce better results and even save time for students and advisers alike. Chapters often go through multiple drafts, but only one formal meeting need be scheduled on each chapter, with no more than one per student per semester. Recommendation 5.9: As these and other recommendations involve setting specific dates, we propose that the DGS, M.A. Director, and Departmental Administrator work together each spring to create a graduate calendar for the upcoming year, setting out all relevant deadlines for all cohorts in the program. Students and faculty alike will benefit by having a ready reference guide to what is to be done when. back to index B: Sponsorship: In the German system from which American Ph.D. programs derive, the single "Doktorvater" was expected to give birth, patriarchally and parthenogenetically, to the newborn Ph.D. Less unitary than that model, our structure provides for three advisers, supplemented at the defense by two outside examiners representing a wider readership. The language of "sponsor," "second reader" and "third reader" suggests a primary working relationship between student and sponsor, lesser involvement by the second reader, and less again by the third. We give a certain parity to sponsor and second reader, both of whom have to approve the prospectus and are listed by the department as involved from the start; they can also serve as co-sponsors. The third reader is marginal in formal terms. Two questions are raised by the current system: first, does this pattern reflect our actual practices, or does it assume a clear division of labor more pervasive in the past than now? Students today often have interests that don't mirror those of one or even two sponsors but connect in different ways to all three readers, and third readers are often fully as involved as the others. This raises the second question: is it good to have the third reader spared the level of involvement of the other two, or does the present system fail to credit the work that many "third" readers now do? A number of junior faculty have expressed the feeling that they're often third readers, as engaged in advising as the sponsor and second reader, yet unacknowledged: the department hasn't even kept track of third-reader assignments at all. The trade-off that comes with having three ongoing readers is that the student gets more advice but has to respond to comments from three different and perhaps contradictory perspectives. Primary reliance on only one or two readers simplifies the issue of feedback. On the other hand, the dissertation will eventually be the student's prime showpiece on the job market, where it will have to appeal to a much wider circle of readers. We believe that, in balance, a dissertation gains by having active involvement from its three readers; the university already provides separately for an after-the-fact reading by having two people from outside the department on each defense. Further, the task of reconciling different readers' advice is an important skill for students to develop, one that will constantly be useful in responding to readers' reports and to colleagues' advice in pre-tenure reviews. Recommendation 5.10: So as to acknowledge - and encourage - active involvement by all three readers, we propose that each dissertation be directed by a three-person Dissertation Committee, one of whom the student will designate as Chair of the committee but all of whom will be considered as common participants in the shaping of the dissertation. All three would sign the cards to approve a prospectus, and all three would be credited as members of the student's committee in department lists and in the department's accounting to Arts & Sciences of each faculty member's service at salary time. back to index C: Dissertation seminars: For several years, we've had dissertation seminars in which students read chapters of their work in progress, and when these have worked well, they've been very successful. Some dissertation groups have been run on a student-only basis, but after several years' experience it appears that this option is generally not ideal. Either discussion drifts in an unfocused way, or else students feel it's a marginal enterprise, attendance drops, and the seminar fades away. If we'll now be having about 15 students in each cohort beginning dissertations, with three years a common period of active work, there is a potential membership of 45 students in dissertation seminars. Some fields may have enough students to have a seminar based in a single field, but in most cases it may work best to have a cross-field group. An advantage of such cross-field groupings will be to help students avoid the problem of tunnel vision, of assuming that their topic is unique to their particular field. Overall, we should have three or four dissertation seminars, organized in such a way as to offer in principle a seminar of use to each of our students. Recommendation 5.11: Each field (or group of contiguous fields) should establish a monthly dissertation seminar if it doesn't now exist, and the faculty should support it by their presence. One or two students would present a chapter at each session. For each year, one faculty member from the field should serve as the seminar's director, and should receive a course credit for this work. This assignment should rotate from year to year among the interested faculty in the field. Recommendation 5.12: To give students a general forum for sharing of their work in progress as they go on the job market, we should make permanent the advanced, open-field dissertation seminar inaugurated by Jonathan Arac and currently conducted by Bruce Robbins in his capacity as Placement Director. This has met twice a month throughout the year, carrying a single course credit for the faculty adviser who runs it. At each meeting one member presents a 30-minute talk from work in progress on the dissertation, which is then discussed by the group for the rest of the time (75-90 minutes). The colloquium gives practice for job-talk experience (speaking to a department, not just fellow inquirers in the same subfield, which is the case in the period-based colloquia), as well as drawing out common methods or issues that join otherwise divergent areas of our discipline. back to index 6: Fellowships and funding We have one of the best-funded programs in the humanities at Columbia, and better than at all but a few better-endowed schools. All of our sequential students are funded for six years (five if they come in with an M.A.), with tuition plus an annual stipend of just over $17,000 (rising to $18,000 for 2004-05). Our annual budget for graduate student support is now just over $5,000,000, roughly half for tuition, half for outright stipends. Three years of funding are tied to teaching; the teaching load was set in 1990 at 7 courses over six semesters, but has since been reduced to 6 sections of Undergraduate Writing in three years, taught to small classes of 12 students. The M.A. year is pure fellowship, TAships are done in the second year, and there is a dissertation fellowship year, originally set as the sixth year but now able to be taken in the fifth year by students who have a completed chapter - a much better arrangement, giving students a concentrated year to complete a good deal of the dissertation before going on the job market in the sixth year. We hope that with speedier progress through orals and beyond, most students will be able to take the dissertation fellowship in the fifth year. We do still lag behind some of our competitors on funding. As one student says, "If NYU is not our competitor but Harvard is, why is Columbia not paying us a base rate of $21,000?" The competition for some students is heightened because some schools offer more money to their top candidates than to others. Since 1990, we have provided equal funding for everyone we fund, and we continue to feel strongly that differential funding breeds anxiety and bad feeling far in excess of the benefits of capturing the occasional extra rising star. Indeed, students often turn down higher stipends elsewhere in part because they prefer to be in a program with equal funding across the board. A: The cost of living: Even when our fellowships match those elsewhere, the cost of living is higher in New York than almost anywhere else, notably for housing. Columbia's subsidized graduate student apartments can run to $850 per month for a room in a shared apartment, so a year's rent alone consumes over $10,000 of the $17,000 stipend. One consequence is that most of our students work ten or more hours per week beyond the duties that come with their fellowships, and almost all have to work full-time during the summer. This is clearly a major drag on time to degree, and one over which we have no real control. Foreign students, on the other hand, can't legally work off campus, can't secure government loans, and are often under severe difficulties in making ends meet. Recommendation 6.1: We should share with the Graduate School our recently developed information about students' off-campus workloads, and continue to press for improved funding. To begin with, an expansion in the Graduate School's supply of $3000 stipends for summer research would very materially aid students at the time of orals and early dissertation work. Award of these funds should be tied to timely completion of the orals proposal and dissertation prospectus. back to index B: Conference travel: Another modest but very valuable kind of funding is for travel to present papers at conferences. These experiences enhance students' sense of themselves as members of a national field, give them contacts elsewhere, and improve their profile when they go on the job market. Both the department and the Graduate School offer limited support for graduate student conference travel, though the grants typically cover only a quarter of the expense. Given most students' marginal financial situation, the cost can be a heavy burden. They further report that faculty seem to assume that they are aware of what professional organizations they should join and which meetings are best to attend, yet often this is not the case. Recommendation 6.2: The department should lobby the Graduate School for better conference-travel funding, and also explore any internal sources of increased funding. At the time that students set their orals, if not before, their advisers should discuss with them the major organizations and conferences in their primary field(s), and talk through which would be best for the student to join or attend. In principle, each student at the orals stage and beyond should be participating in a field-based organization apart from the MLA. back to index C: Outside fellowships: One way our students can improve their funding is by winning an outside fellowship; outside funds can be used to raise their stipend to $22,000 before the balance is offset by reduction in the Columbia stipend. The financing of our full funding also depends on a modest but steady inflow of external funds (Javits, Fulbright, Newcome, etc.). For this reason we require all our students to apply for at least one outside fellowship as they begin dissertation work, and our students have had fair success in winning these; several are currently being held, and our students also do well in the intra-university Whiting competitions. Even so, we regularly hear from Whiting committees that our students' proposals are jargon-heavy and don't really make clear the significance of the project. Edward Mendelson, Jenny Davidson and others have generously served in recent years as fellowship advisers, but it appears that a fuller program would be worthwhile, and skills honed in grant-writing will also serve students as they go on the job market and in their careers thereafter. Recommendation 6.3: The department's Fellowship Committee should be expanded to include two advanced students who themselves have won Whitings or other fellowships. We believe these students would enjoy sharing their skills with younger peers, and would greatly supplement the strategic advice and editing work of the faculty members of the committee. 7: Graduate Student Teaching We begin from the principle that gaining meaningful teaching experience should be a central part of every graduate student's professional training. We know that teaching experience is very important on the job market, not only as a qualification but also because experienced teachers interview well, more readily appearing as colleagues to hiring committees. The newly organized University Writing program run by Joe Bizup gives our students excellent training in the teaching of writing, along with the autonomy and small classes that have always characterized the Columbia writing program. Yet the present system has some serious limitations. The TAships held in the second year do very little to begin preparing students as teachers; instead, in most instances the TAs are passive auditors in a lecture class, swinging into action only to grade a share of papers and exams written by students with whom they have had little or no real contact. The TAs then go into University Writing, which is a far more positive experience, but after two years of teaching writing, they're eager to move into the teaching of literature and related texts. Lit Hum and CC provide excellent opportunities, but typically only about 30-40% of our students enjoy that option. Under the old Logic & Rhetoric writing program, a number of students could go on to teach introductory literature surveys in General Studies, but those courses have been abolished in recent years, so that opportunities to teach literature at Columbia have significantly decreased. As a result, students often seek out adjunct courses elsewhere, more for the teaching experience than for the minimal pay. As Henry Pinkham recently remarked ruefully, "Our students do a little teaching for us for a lot of money, then go out and do a lot of teaching at other places for a little money." The cost of living may require students to supplement their fellowship income, yet that can be done far more efficiently with almost any other kind of tutoring, editing, or writing than by going out to a distant community college campus to teach fourteen weeks for $1800. Such poorly-paid work is a major factor in delayed time to degree. After extended discussion of these problems, the Carnegie committee and the CGE would like to propose a three-part model for graduate student teaching, one that begins with an enhanced TAship - a controlled first exposure to running a class before going on to full-scale teaching in University Writing and then in literature classes in the fifth year (or the sixth year for students who opt for the fifth-year dissertation fellowship). A: Second-year TAships: We believe that the best way to make TAships a meaningful part of students' training is to expand them to include the leading of sections. In order for this to be workable, the sections should be small, no more than the 15 mandated by Arts & Sciences for recitation sections. TAs should have real guidance by the faculty, and the overall workload should be kept to an amount that won't fill the whole year and extend time to degree. Further, as not all faculty want to use sections, we need to allow for students to be available to serve simply as graders in a good proportion of courses. Recommendation 7.1: In order to meet these requirements, we propose the following plan: 1. Each graduate student would teach in a sectioned lecture course one semester of the second year, and would serve as a simple grader the other semester. Sections would take the form of an added hour, rather than reducing lecture time. Following the model used in Art History, TAs would run two one-hour sections of 15 students each, thereby addressing the problem of numbers, as they would be responsible for the 30 students which Arts & Sciences currently provides as the number per TA. While this solution adds an hour of class time, there is only one preparation to do, and a much more manageable number for discussion, with the same amount of grading in the end. 2. In order to make the assigning of sections a simple and transparent process, we propose that sections be instituted in all classes that are expected to have 60 or more undergraduates and whose instructor requests sectioning. This cut-off puts the sections where they are of the greatest benefit, in classes whose size inhibits discussion during the lectures themselves, and it works well in terms of numbers, if second-year students are to run sections one semester and serve as simple readers the other; see below for a detailed model on this basis. 3. What would happen with a course whose enrollment comes to 70 or 75 students, well above the 60 for two TAs but below the 90 for which a third reader would now automatically be provided? We propose hiring additional TAs from upper-year cohorts for this purpose. Arts & Sciences already does provide such supplemental funds, providing from 6 to 12 added TAs per semester depending on need, and we expect we can get more such funds if we can show the need for them. 4. A more meaningful TA experience of this sort will naturally take more time than our present system, and the question was posed at our fall faculty meeting of how best to compensate students for this added time. The student and faculty members of the CGE and Carnegie committees agree that an appropriate method is to count it directly into the program's requirements. We would do this by giving the TA a course credit for the sectioned course, on analogy to the "R" course credit now given for the Undergraduate Writing teaching preparation course. Students are currently expected to spend up to 10 hours per week on their duties as TAs or RAs; running a section should really not take much if any more time than this, but the assigning of the course credit acknowledges that this is a course they're doing on an active basis, and the net result would be no increase in time over our present system. We say this in the context of an intense concern for time to degree, which is addressed at several other places in this report, in connection with the known major bottlenecks in the program. 5. To be workable, lecture-class sectioning does require direct engagement by the faculty member who is teaching the course. Those who use sections will be agreeing to give their TAs initial guidance on running a class, then to meet weekly or biweekly to assess how the sections are going, and to attend one or two section meetings. 6. To prepare the TAs, we would institute a half-day TA orientation at the start of the year, to be run chiefly by upper-year students who have served as section leaders in the past. We would pay these upper-year people an honorarium for running this session, at which some faculty members would appear at appropriate points to discuss their practices and take questions. This session should be coordinated with the general GSAS teaching-preparation conference. 7. Student TAs will gain important experience from teaching sections. Several members of our committees have said that leading sections worked well for them as graduate students and helped form their professional identities. We have also heard from several current students who have run sections for Ann Douglas and Andy Delbanco, and all have been enthusiastic about the experience. Rather than considering that the enhanced TAship would take unfair advantage of graduate student labor, we see the current use of TAs as intellectually and ethically problematic, an unsubstantive experience that takes advantage of the students and serves only the faculty's narrowest interest in avoiding unwelcome work. Instead we are proposing an educationally much more meaningful TAship of real use to the student, compensated for in practical terms by the course credit. As for pay, our second-year students would still be receiving a $47,000 package (plus medical benefits) with only modest service required in return. 8. Teaching a section under faculty guidance gives a controlled first experience in running a class, without the heavy preparation time required to do one's own free-standing class. Bringing in a poem to analyze, taking up issues touched on but not developed in the week's lectures, responding to student questions and confusions, all form a manageable first teaching experience, which will give valuable grounding for the more demanding teaching the next two years in Undergraduate Writing. Our undergraduates are not served well when they are required to take Undergraduate Writing from a third-year graduate student who has never faced a class or run a discussion before. A substantive TAship in the second year will provide a welcome preparation for UWP teaching the following year, to the benefit of our graduate students and undergraduates alike. 9. Our undergraduates will also benefit by the offering of sections in our larger lectures. Reports from History, from Art History, and from our own colleagues who have used sections all indicate that our extremely talkative College students welcome the opportunity to get together and discuss the issues raised in the lectures. Sections will be particularly valuable in larger courses of 60 or more students, too large for an effective lecture-discussion of the sort that can be carried on with smaller lectures, and we know that several of the people who teach large lectures are actively interested in using sections. As the large courses use more TAs than smaller lectures, we are confident that we can readily place half our TAs in sections each semester; no faculty member who doesn't want to use sections will be under any necessity to do so. No doubt there are some undergraduates who don't enjoy the experience of sections, but the majority of our lectures would remain unsectioned, so that undergraduates could avoid most such courses if they wish. The addition of sections to some courses and not others also reduces the strain on undergraduates' time and on room assignments. 10. The faculty who choose to use sections will enjoy extending their mentoring to the realm of teaching, and will be in a much better position to write effective letters of recommendation later on. An often-reiterated complaint from faculty at non-research-intensive schools is that letters of recommendation often talk only about the students' writing and show no knowledge of their teaching, which is a critical qualification for jobs - including at research universities. 11. Faculty who use sections would be taking on extra responsibilities in overseeing them, but would most likely do less grading. Those on our committees who have served as section leaders report that they work best when the section leader is the primary grader for the students in that section. Under our present system, undergraduates in large lecture courses are having two thirds or three quarters of their work graded by graduate students whose faces they may not even know; with sections, there would be a direct connection and an enhanced chance for real feedback. Under such a model, the faculty member would give guidance on grading, work through samples with the TA, and read borderline papers and deal with any appeals. Alternatively, faculty might choose to grade the term papers, but have the TAs do all the exam grading. 12. We have discussed the question of fields and of pressures on the graduate program's size. As our proposal uses the same number of second-year TAs as the present system, and in the same proportions, it will exert no upward pressure on the graduate program. It will continue to be the case, as it is now, that students are sometimes assigned to TA outside their immediate field of interest, but most are placed in a contiguous field, and in any event the real focus of the TAship is on developing pedagogical skills, learning how to lead a discussion and control a class. Several of our current faculty say that they had very positive early experiences as TAs outside their direct field of interest. back to index To give an idea of how sectioning would work in practice, the following table outlines how these proposals would have translated into practice in the current academic year. Table 4: Section assignments as they would have looked for 2003-04 Course Size TAs Course Size TAs 3335: Shakespeare 60 2 3336: Shakespeare 62 2 4822: 19c Novel 101 4 3267: Foundations Am Lit 95 3 3269: Mod Brit Lit I 92 3 3270: Mod Brit Lit II 79 3 3208: 20c Comparative 119 4 3230: Joyce 68 3 4604: Am modernism 79 3 3219: 20c Poetry 101 4 4930: Pol's in Am Film 86 3 3390: Myth & Lit 73 3       4560: Contemp Theory 63 2 Total TAs needed   19     20 On this model, a total of 39 TAs would have been used this year to provide sections for 13 courses, at least if all those instructors wished to use TAs. Under our present system, these 13 courses are using a total of 31 TAs, so the new plan would require hiring an additional 8 TAs (either 8 individuals, or 4 doing a section each semester). As our current system has a total of 17 sections run by upper-year TAs, the increase of 8 doesn't represent an enormous increase to Arts & Sciences, and we believe that our proposed system would give us a strong claim on those added funds. (Note: This year was somewhat anomalous in that only two pre-nineteenth-century courses, both on Shakespeare, would have been sectioned. We've often had one or two other 60+ early-period lectures, particularly on Chaucer and Milton. We expect that as our newly-arrived medievalists become known to the undergraduates, we'll see more medium-large medieval lectures. With our current cohorts of 18 sequential students, we would typically have 2 medievalists and 3 early-modernists in each second-year class. Under this year's situation, most of them would be assigned to TA in Shakespeare. In a year with more large early-period offerings, the medievalists would likely be TAs in Chaucer.) This year, we have a total of 35 lecture classes with undergraduate enrollments of 30 or more and consequently with TAs assigned as graders. Under the proposed plan, 22 of these 35 courses would have remained unsectioned, with regular graders. So students disinclined to take sectioned lectures would still have about 2/3 of our lectures available to them. Instructors not wishing to use sections could simply use graders in the traditional way. A further question concerns the availability of rooms for holding these sections. We've spoken with the Registrar about this; her advice is that we should make our sections a required part of the course if we wish to guarantee rooms for the sections, as courses with optional sections get last priority. We believe that, in fact, the overall faculty sentiment is that our sections should be required. In that event, the Registrar doesn't anticipate that we'll have difficulty securing the rooms needed for the number of sections shown above. She says that sections can be scheduled on any day of the week between 9-11 a.m. and between 6-7 (and, to a more limited extent, between 4-6), as well as all day on Friday. A number of our sections could also be held in the two seminar rooms we control in Philosophy Hall, 612 and 408A, which often have a free hour at the middle of the day. In order to have mandatory sections, they need to be set in place before pre-registration begins, so that students can plan their schedules in full knowledge of when the class will meet. Most of the courses listed above are known quantities that regularly run above 60 students when offered. We propose that the Curriculum Committee would estimate the anticipated class size as part of the curricular planning process and arrange for sections accordingly. Any class that had had enrollments of 60 or more when last offered would have a presumption of sections, unless the instructor preferred not to use them. In borderline cases, the benefit of the doubt would be given in favor of sectioning; one or more sections could always be cancelled once pre-registration figures were in, if the numbers came in substantially below what was anticipated. We believe that sections will provide a genuine enhancement of our graduate students' training and of our undergraduates' experience in large lectures. Sectioning is a concept well known in the discipline, and one that has been tried successfully by several people in our own department. At the same time, we recognize that a broad program of sections is something new for the department, and so we propose that it should be re-evaluated once it has been tried out, so that we can see how it works for us in practice. Recommendation 7.2: The CGE should evaluate the use of sections after they have been in place for three years, and should report to the department, at which point the faculty will determine whether to continue sectioning on this scale, to expand it to include more courses, to scale it back to fewer, or to eliminate it outright. back to index B: University Writing: We know that experience in teaching writing is highly valued on the job market and that teaching in the UWP benefits our students in many ways. We believe it should continue as a basic component of at least most of our students' professional training. It remains the case that not all of our graduate students take to the course or really do it well, and we have a longstanding problem that we have few if any options for funding middle-year students if they aren't teaching writing. Further, the expectation that all students will teach the course constrains us in our admissions: we have to turn away some exceptionally qualified foreign students merely because their English, while very good, isn't idiomatic enough for us to expect them to teach composition by their third year. Our students' staffing of the old "Logic & Rhetoric" writing program underwrote our six-year funding package at a time when most departments didn't even have four years of funding. Other departments have been catching up in recent years, however, and our students are now expected to do more teaching than in History and various other departments. If we could offer a funding option comparable to History's, it would involve five years of funding, but with no UWP teaching requirement at all. This would be a year less than our UWP-centered package, instead entailing two years as TA and a year of literature teaching (or one as TA and two of literature teaching). Many students will of course actively desire the UWP experience, as well as the added year's funding, but others might prefer to take the five-year option, and this would also give a safety net for the occasional student who just doesn't prove to be a good writing instructor. Recommendation 7.3: We should explore opening up more varied funding options for our students, looking close at patterns currently used in other Columbia humanities and social science departments. The UWP will remain the right option for most of our students, but it would be good all round if an alternative could be made available. back to index C: Upper-year teaching: We have begun discussions with the College and the Graduate School on opening out the third year of our students' teaching to include writing-intensive introductory literature courses. The administration has responded favorably in principle to this idea, and we will continue to work on this over the next year in conjunction with college-wide efforts to create discipline-based courses that could follow University Writing. Recommendation 7.4: We should work with the College and the graduate school to build in a year's literature teaching for each of our students, with the goal of having a program in place for fall 2006. Recommendation 7.5: We should support proposals now under development by the Graduate School and the College to provide post-docs for new Columbia Ph.D.s to teach in the Core and in other capacities. Together these courses would effectively round out our graduate students' teaching experience at Columbia, contributing very substantially to their professional development, their sense of themselves as academics, and their attractiveness on the job market. back to index 8: Going on the Market All good things must end, however, and we need to do our best to help our graduate students make the crucial transition to the job market. As shown in Table 1 on page 6, our placements in recent years have consistently been strong, with an average of 15 placed each year in tenure-track jobs and post-docs. In addition, some students in each cohort will have found their way into another career during the course of their studies. Some leave before finishing, while others complete the Ph.D. without intending to go into college and university teaching. We can certainly count as success stories those of our graduates who go on to careers as reporters, editors, and film makers, and they continue to use the skills in research, analysis, and writing that they honed while in our program. It remains the case that each year there are students who want academic jobs and who deserve them but who fail to prosper on the market - even with our successful placements, after all, not all of those students are going to jobs they're completely happy to have. Can we do a better job of helping our students get the jobs they should have? Several years ago we instituted mock interviews, and these have certainly been useful. Our smaller cohorts these days should also give us time to offer better advice and direction in the job search. Yet the 13 upper-year students who responded to our survey don't rate our efforts very highly. Asked "how useful was the advising you received around your job search," only 2 said "very useful," 4 said "useful," while 4 put the third category, "somewhat useful," and 3 put the lowest category, "not useful." Of course, these are the responses from students who went on the job market but didn't get jobs; probably the successful candidates would - rightly or wrongly - rate our assistance more highly. Yet there are no magic solutions in a perennially tough job market, and by the time students are on the market, they're largely formed, so advising can only do so much. Our advising has been active and ongoing, both by individual sponsors and by our very engaged placement directors over the past decade and more. Yet like many aspects of our department and its programs, this collective experience exists largely in anecdotal fashion. These oral traditions may work adequately for current students who are fully connected to the department, but many students are disengaged from day-to-day contact with faculty by the time they apply, and may even be far from campus; purely oral traditions are of limited use to them. Recommendation 8.1: A committee composed of current and past placement advisers should draft a pamphlet on job search preparation and strategies. This draft can be circulated for comment by the full faculty and by experienced students, including those who have had good success on the market in recent years. The final version can then be given to every student who signs on as a candidate. It would be appropriate to give this pamphlet to students already at the time they pass their orals; this would be a logical point for them to begin thinking seriously about such job-related matters as conference participation and the sending out of articles. back to index In Conclusion While the focus of a report such as this is on problems and improvements, it's important to keep in mind the fundamental health and vitality of our program. More and more students are applying to join us, and we have no difficulty attracting extremely talented students to come. As one respondent succinctly summarized the benefits of being here: "Excellent faculty, collegiality among graduate students, location in a cultural mecca, interdisciplinary ethos and possibilities." For our part, we as faculty benefit greatly by teaching and advising these students. Given these strengths, the number of recommendations made in the preceding pages is much higher than our committees anticipated when we began this process. It's a little shocking to realize just how many aspects of our program have gone along without direct discussion for many years, often despite real changes in the wider environment and lurking problems with present practices. We have tried to assess every aspect of our program carefully, building on everything that's working well, and proposing changes that can work in practice and be sustained over time. As we put these changes in place, there will be a period of adjustment to them, but ultimately a coherent and well-planned program is easier to run than an amalgam of half-forgotten ideas and compromises. The proposals in this document are intended to help our students get the best training, retain momentum, and do their best work as they prepare for a life of teaching and scholarship. We hope that the result will be an improved experience for students and faculty alike. back to index