Fortunately, my experience as a graduate student included an adviser who had genuine concern for my welfare and was willing to consider the opinions of other faculty members regarding major decisions that affected my progress. I have taught only in departments that do not have a doctoral program. Therefore, I feel fairly free to voice my opinions without concern for departmental or career politics. I do, however, have an ongoing interest in the subject because, periodically, the best and the brightest of my students ask for advice concerning a choice of graduate school. My perceptions of the treatment of subordinates lead me to consider two primary causes of the abuse that is certainly present.

First, faculty members are not hired for their advisory abilities. Much of the “abuse” is not mean-spirited or cold-hearted behavior by advisers. It might better be described as neglect or even incompetence. The abuse results from the practice of hiring and promoting physics faculty members on the basis of their promise as researchers (or, for administrators, their promise as grant writers). They are not hired for their promise as instructors or supervisors, or on the basis of their people skills.

A second possible cause of the abuse of doctoral candidates is that advising decisions are made by an individual with no oversight. The culture of many PhD-granting departments is antithetical to external oversight of the treatment and guidance of students, or of acceptable conditions for finishing one’s degree, beyond a vague notion that it should not take “too long.” The first rule of faculty-to-faculty relations is that one does not meddle in the relationship between adviser and student.

Consequently, one finds supervisors who are in a role they have never filled before, and who were hired for abilities and skills unrelated to supervision. They then operate in a culture that precludes almost any form of unsolicited advice or direction concerning that role.

Students who take things into their own hands by trying to transfer to a new research group find a number of major hurdles. First, work done with the previous adviser is effectively erased, and that can cost a year or more of study. Second, the number of advisers willing to pick up transferring students will be limited, because advising a student who has left another group due to a disagreement is viewed by some as meddling.

In essence, the only check on an adviser’s behavior is the long-term effect of chasing away graduate students—namely, the resulting low research and grant productivity. Of course, by the time that has happened, many graduate students will have been served up as cannon fodder, and the professor may have already been granted tenure.

Some form of external oversight is needed, and the adviser’s power must be dispersed. Each department should adopt a formal and openly published policy for the treatment of graduate students. The Statement on Treatment of Subordinates (Physics Today, November 2004, page 43, box 1) is a good starting point, but it needs to be more specific. The published policy should then be a guiding document in the advising that should become the responsibility of the student’s entire research committee.

Currently, most dissertation committees serve mainly as gatekeepers. Instead, a student’s work should be presented twice a year to the committee for a balanced assessment. To mitigate possible bias toward the primary adviser’s opinion, the committee members should put their assessment in writing before they engage in any collective discussion of the student’s progress. As a positive byproduct, the committees would then also have some ethical oversight of the entire department’s research practices. Committee members would be the first “outsiders” to see the results, and they would hear first-hand about the methods and practices. Some common ethics violations cited in the ethics survey in figure 1 of Kirby and Houle’s article could be addressed at this stage, well before publication.

With the involvement of the entire committee, the rules concerning meddling would eventually become irrelevant. The advice and guidance provided would be normalized by including more experience and points of view. New faculty would have some introduction to their advisory duties, starting with their very first student. Research is rarely done in isolation, and advising should not be done in isolation.