Professor-in-Training: Status Control of the Teaching Assistant

Abstract

I examine the role of apprenticeship status in controlling the labor of unionized graduate student teaching assistants (TAs). In her book Coerced, Erin Hatton identifies status as a basis of labor coercion—particularly in nontraditional labor regimes—in which managers control workers’ access to status-based rights, rewards, and punishments. I expand Hatton's concept of status coercion to status control and distinguish between two types: despotic, in which status coercion prevails, and hegemonic, in which status consent prevails. I argue that status control of TAs is hegemonic, relying on their investment in a system of apprenticeship in which course instructors are a source of professional advancement, opportunity, and support outside of the TA job. I draw on autoethnographic fieldwork to analyze one expression of TA control, participatory management. In this model, the faculty instructor invites TAs to collaborate on course design and encourages routine discussion of teaching strategies, in which hidden labor is made regulable through “confession”. Identification with the instructor limits TA autonomy by disrupting alliances between TAs, and between TAs and students. I conclude by sketching variations in TA management and by discussing status control as a broader mechanism of extraction in the contemporary university.

One Tuesday morning about a month into the semester, I received an email from the professor asking me to deliver the day's lecture. She was trying to meet a manuscript submission deadline and she believed that lecturing on that day would take too much time away from her writing. I called her and told her I would do it. In reality, I could not have refused. To do so would have been to shortchange the students. And TAs do not say ‘no’ to influential and more powerful faculty members very easily or very often.

(Michael DeCesare, “On Being a Graduate Teaching Assistant,” DeCesare, 2003 , p. 152)

A picture of the teaching assistant's (TA’s) labor process would be obviously incomplete without a discussion of the faculty who manage them. Yet in the autoethnography seminar of Berkeley TAs that eventually developed this special issue, we TAs initially did not think to “study up” and examine our relationship with the faculty who supervise us. The needs of students were palpable in our fieldnotes, they directly shaped and constrained how we went about teaching. Instructors, on the other hand, were in the background, never present in our classrooms or with us while we graded. Indeed, most teaching assistants in Berkeley sociology will say that the characteristic style of instructor management is no management at all.

In retrospect the instructors were always present in our work, just indirectly. In my field notes, I often described the strain of reconciling the expectations of “the course”—a heavy reading load or the style of an assignment, for example—with what I felt I could reasonably expect from my students or could execute as a teacher. At the start of one ethnography seminar meeting, while doing fieldwork check-ins with the other TAs and our advisor, Michael Burawoy, I described how the group of TAs I was teaching with had attempted to diplomatically negotiate some demands with our instructor, with limited results. Michael responded by asking, if they’re so burdensome and no one is watching, why do you follow the instructions at all?

This question baffled me. The answer seemed obvious yet inarticulable. As for DeCesare reflecting on being a TA 20 years ago, it's just what you do. Such uninterrogated “common sense” provided an early clue to the argument I ultimately make in this paper. As I will show, the power of faculty instructors over TAs relies heavily on TAs’ consent to an ideological system of apprenticeship, one that coordinates the interests of graduate students with those of faculty. By being a collegial or invested TA—or by doing what an instructor asks of them—a graduate student may be rewarded with a letter of recommendation, a personal referral for a research position, or a mentoring relationship with an overextended professor. Certainly these rewards are not always, or even usually, top of mind, but PhD students work in an economy of status, in which they rely upon higher status professionals for their chance to attain a secure and privileged academic position. And as the growth of undergraduate enrollments are met not with more tenure-track positions but with the growth of part-time and contingent sources of labor (Colby, 2020), graduate students rely ever more heavily on faculty for any opportunities to advance professionally.

As Hatton (2020) shows, for graduate research assistants (RAs) in STEM—“apprentices” who are not legally recognized as workers—the relationship between graduate students and faculty supervisors is characterized by status coercion, “the power to discharge [workers] from a particular status…and thereby deprive them of the rights, privileges, and future opportunities that such status confers” (p. 13). Far more than economic coercion, the power of these faculty supervisors comes from their ability to unilaterally bestow and revoke status rewards: RAs’ access to publications, credentials, and future academic positions. Under threat of status-based punishments, and with no legal recourse for abuse of power, research assistants can be made to work long hours in the lab and endure abuse, harassment, and dehumanization from their advisors, while being exploited for faculty and university gains.

In my five semesters as a TA in sociology, I have seen little that resembles these coercive expressions of control. TA work is largely self-directed and unsupervised: TAs have significant discretion in how they teach course content (Germain, this issue) and how they allocate their time between TA work and the other demands of graduate school (Gepts, this issue). Consider the concerns one professor shared with me about managing TAs before the start of a semester: How would she be able to standardize our teaching and grading? How could she ensure that students were not advantaged—or worse, disadvantaged—by being assigned to a given TA? Without direct control over TAs, she seemed to suggest she had no power at all. To capture this divergence from Hatton, I broaden her concept of status coercion to status control and distinguish between two types: despotic, in which status coercion prevails, and hegemonic, in which status consent prevails. I argue that TA status control is hegemonic, relying on the TA's investment in their own apprenticeship status, and on the coordination of graduate student interests with those of faculty.

That my professional interests rely on my relationships with faculty poses an obvious problem for a graduate student writing about the workings of this very power relation. I conducted an ethnography of the same field of power in which I already do, and will continue to, live and work. This peculiarity has meant that the very power dynamics I study in this paper are those I have had to navigate as I wrote and considered publishing this research. And not only the status rewards and punishments that may have affected me—the loss of amiable relationships with faculty, the gains of my first publication—but those that could be inflicted on the instructors with whom I have taught, and who are embedded in the same field of power, though at different and variably vulnerable locations within it.

These concerns are compounded by a more typical concern of ethnographic research: the difficulty of reconciling participant observation with a medical model of informed consent (Thorne, 1980). Initially focused only on the goings on of my discussion sections, my research turned to faculty only after a period of time in the field, after I’d begun to wonder about the instructors who seemed largely absent from the labor process and yet still exerted constraints on it. As with much participant observation, the parameters of such an analysis could not be entirely known or negotiated in advance. Following the guidance of ethnographers before me, I have approached consent in this project as an ongoing series of negotiations with my subjects and field site, during both data collection and drafting (Fine, 1993; Duneier, 1999; Grindstaff, 2002). Such negotiations take on an added gravity in autoethnography, where one's ability to anonymize is limited and where accountability to and relationships with one's subjects do not end when one moves on from the project. The result is that I have limited my analysis to just one case of instructor management: that of a tenured faculty person with whom I shared a working draft after our teaching relationship ended, and who shared an interest in—and belief in the importance of—an analysis of TA labor control.

In the article that follows, I first engage directly with Hatton's construction of status coercion to formulate a broader concept of status control that encompasses both hegemony and despotism. Next, I examine one concrete expression of status control of TAs: Professor Adams's ¹ participatory model of management. This model has proven a useful case because its expression of status hegemony is an extreme one. I show that even when instructors attempt to manage TAs through collaboration and consent, the basis of the status control of professor over graduate student remains. In Professor Adams's participatory model, TAs were invited to collaborate on tasks typically reserved for the professor: designing the assignments, rubrics, and course policies that would direct their teaching labor. TAs also routinely engaged in discussions of the ups and downs of their teaching practice, discussions in which hidden labor was made regulable through “confession” (Foucault, 1990[1976]). These hegemonic mechanisms forged a fixed alliance between the TAs and the instructor, thereby limiting potential strategic alliances between TAs and between TAs and their students. I conclude by sketching two additional types of hegemonic status control and by arguing that this form of control is a broader mechanism of domination and exploitation in the contemporary US university.

To be clear, I do not argue that the labor control I describe here is the only force extracting effort from TAs, nor do I believe that faculty are maliciously exploiting their TAs. As this special issue illustrates, faculty management is but one of many constraints and liberties that shape the TA labor process and which collectively produce an autonomous and committed labor pool from which the university extracts pedagogical, administrative, and care work.

From Despotic to Hegemonic Status Control

The question of labor control has been at the heart of studies of the capitalist labor process: How do employers extract surplus value beyond the wages workers are paid? Much of this literature has focused on economic coercion as the primary basis of labor control. Yet, as Hatton (2020) shows, this focus excludes the primary basis of employer power in many nontraditional labor relations and elides the noneconomic sanctions that discipline all workers. In a surprising comparison of prisoners, workfare workers, college athletes, and graduate student research assistants in STEM, Hatton (2020) shows that all of these workers experience the same form of coercion, status coercion––“the power to discharge [workers] from a particular status … and thereby deprive them of the rights, privileges, and future opportunities that such status confers” (p. 13). While any worker can experience status coercion, it is intensified in these labor regimes by their respective constructions as something other than work. To disobey a supervisor is to risk severe status-based punishments, such as losing eligibility for parole for prisoners, access to healthcare and food assistance for workfare workers, playing time for college athletes, or publications for RAs.

Hatton's RAs in STEM offer a useful comparison to the TAs in this study. Legally and institutionally constructed as student apprentices, not workers, RAs purportedly are paid in education and training and require considerable guidance from faculty. Faculty supervisors wield expansive punitive power with little oversight, controlling RAs’ “education, academic credentials, and future careers” by way of their control over RAs’ status as “graduate students in good standing” (p. 71). This power is concentrated in the hands of the faculty primary investigator (PI) who is both the RA's boss and academic advisor. The RA's own research is inseparable from that of their PI, and the PI in turn depends on the labor of their RAs for the execution of research and the grant funding it produces. Hatton emphasizes the coercive strategies of subjugation in a labor relation one RA describes as “indentured servitude”: bullying and harassment, surveillance to ensure RAs are working weekdays, weekends, and holidays, and even delaying the graduation of RAs who have become particularly efficient workers (p. 198).

Teaching assistantships have traditionally also been structured on an apprentice model, in which graduate students “sell” their labor to a supervising professor in exchange for a modest stipend and the training required to become a professional themselves. But as tenure-track prospects have declined and undergraduate enrollments have soared, TAs have mobilized through unions for recognition as wage workers, demanding an altered exchange with better pay and protections in the present. Unionized for over 20 years, UC Berkeley TA positions are governed by a union contract, setting pay, duties, and working hours. But one's entry into a TA position primarily reflects not economic necessity—though these positions are the basis of material survival—but the peculiar agreement of their apprenticeship: one must pay tuition and fees to obtain a PhD, but these can be waived by working as a teaching assistant. TAs are thus bound to university positions—barring other sources of tuition remission—for the duration of their degree.

Sociology PhD students undergo a long period of training and face uncertain professional prospects on the other side. As the “master status” becomes increasingly unattainable, graduate students’ need for status distinction only increases, intensifying anxiety and reliance on faculty for status rewards. PhD students in sociology will typically TA for a variety of professors, and though they rarely work for their academic advisor as RAs do, they often rely on many different professors for letters of recommendation, guidance (and sometimes funding) for research, co-authored publications, and program requirements like dissertation committees. Even a faculty instructor of no relevance to the TA's research might share a positive or negative assessment of the TA with their colleagues. To show oneself to be a capable and collegial TA, then, is to potentially advance one's own future interests. Take, for instance, a fellow TA in my department who was offered a short-term administrative position with a faculty member after being recommended by a professor she had TA’d for. She would be nearly 8 months pregnant at the start of the project, and despite having planned to sharply decrease her work commitments as she approached her due date, she accepted the position. I later asked her why, when no one would have thought less of her for turning it down. She reasoned that she didn’t want to say no to such a prominent professor: What if I build a relationship with this professor? What if she offers me another position in the future, based on this work? What if the instructor I TA’d for doesn’t recommend me to other things because I turned this down? The TA's compliance was secured not through threat of punishment, but through the coordination of her future interests with the interests of faculty.

So coordinated are these interests, that I rarely thought about the specific power faculty held over my career when interacting with instructors as a TA. There are times, however, when one enters a TA position very consciously aware of the impact of the present relationship on one's future. For example, graduate students sometimes request to TA for a professor with whom they hope to take a qualifying exam, or most starkly, graduate students occasionally TA for their own advisors. Remaining a “graduate student in good standing” is thus key to the TA's professional advancement, yet our labor process does not resemble the regimes of status coercion described by Hatton, in which “the possibility of punishment is omnipresent” (p. 27). Whereas Hatton focuses only on coercion, I broaden her theory to distinguish between despotic status control, in which the application of coercion prevails with little concern for obtaining consent, and hegemonic status control, in which “coercion and consent balance each other reciprocally without force predominating excessively over consent. Indeed, the attempt is always made to ensure that force will appear to be based on the consent of the majority” (Gramsci, 1971, p. 80). I argue that TA control is hegemonic, relying on the TA's own investment in the apprenticeship structure and its potential rewards.

Expressions of Hegemonic Status Control

How does apprenticeship status shape the labor process of TAs? Apprenticeship ideology frames workers as dependent, requiring more guidance and direction than other kinds of workers. In academia in particular, apprenticeship status foregrounds the “collegial relationship” of faculty and graduate student, in which the graduate student is a pupil of the professor and requires their guidance and regulation. As Dirnbach and Chimonas (2003) argue, this ideology masks a “rigid academic hierarchy, in which administrators and professors dictate terms, and graduate students take orders” (p. 141). Apprenticeship also constructs the TA's labor as compensation in and of itself. The apprentice labors but is rewarded—in addition to any monetary compensation—in the training and experience received, training that will allow the apprentice to fully enter the profession. This construction serves both the faculty and universities for whom TAs work because teaching labor is constructed not as an extraction, but as an “opportunity” or the “gift” of education, rendering workers “unexploitable.” Given these features of apprenticeship status, we might then ask more specifically, how is the TA's labor process organized to secure status consent?

Tenure-track faculty who teach at R1 universities face constraints similar to those of TAs described in this issue. Wright et al. (2004) describe the conditions of work in US universities, which have become “greedy institutions,” making multiple, conflicting, and increasing demands of faculty. Tenure-track faculty are hired and evaluated almost exclusively on the basis of their research (Mahaffy & Caffrey, 2003), producing a time bind between professionally rewarding research and mandated teaching loads (Wright et al., 2004). Nonetheless, faculty in my department have shared that the importance of teaching (specifically teaching evaluations) for tenure is unpredictable—it all depends on the tenured faculty in the department, the Dean, and a higher committee, known at Berkeley as the “Budget Committee,” made up of faculty from across the campus. Gill (2016) illustrates the intense anxiety among academics, particularly early-career faculty who must excel in teaching, research, mentorship, and department service in the hope that it will lead to job security. New faculty report their departments as under-supportive and lacking in teaching advice and, much like the silence among TAs regarding teaching emotions and practices (Eby, this issue), professors keep teaching accomplishments private (Wright et al., 2004). Indeed, my observations suggest that faculty rarely discuss teaching with each other and are unaware of the common difficulties others also face in their classrooms.

Instructors also face the peculiar challenge of managing a new group of workers each semester whom they cannot select or train. And, as Germain (this issue) shows, TAs have significant autonomy in their teaching practice, as their work is unsupervised and resistant to routinization. Nor do faculty have the tools of economic coercion; it would be against their interest to fire a TA when there would be no adequate replacement partway through the semester. ² Indeed, unlike at many other public universities where TAships are scarce and thus coveted for their tuition waivers and pay, TA positions in my department have, in recent years, outnumbered sociology graduate students who will take them. All in all, faculty have a limited set of practical tools: they may create expectations or “rules” for TAs through the syllabus, assignments, course policies, and interpersonally in weekly TA meetings. But how to enforce them?

Scholars have shown how workplaces leverage cultivated or pre-established worker statuses to control work that is autonomous or not easily surveilled. Labor processes may be structured around rewarding such ethics as the high achieving (Sharone, 2004), diligent (Vallas et al., 2022), entrepreneurial (Purcell & Brook, 2022), or flexible worker (Boltanski & Chiapello, 2005) in order to align worker interests with those of the company. In Sharone's (2004) study of engineers at a tech firm, status rewards and punishments were key to self-management. In examining why engineers self-impose such long work hours, he finds a managerial system that fosters intense anxiety among workers about their professional status relative to each other while requiring workers to set their own performance goals and working hours. The case I present below also relies on extensive worker participation in management, but rather than drawing on the worker's anxiety over low professional status, it draws actively on worker identification with a manager of higher professional status. Nonetheless, I see apprenticeship status as a disciplinary mechanism, instilling in the TA a sense of professionalism, from which the university extracts pedagogical labor in the present, and which trains the TA to one day become the self-managing, responsibilized academic, whether tenure-track or contingent (Gill, 2016; Weeks, 2011). Further, unlike most scholarship on hegemonic forms of worker management, TA management relies on a status hierarchy that is external to the labor process, between graduate students and faculty rather than between TAs and instructors.

To illustrate management of TA labor, I examine an extreme expression of hegemonic control: an empirical account of a semester in which the faculty instructor appeared to neither dictate rules to the TAs, nor to enforce them. Professor Adams invited me and my fellow TAs to participate in the work of the professor: to collaboratively discuss and design the course policies, assignments, and grading rubrics that would direct our teaching labor. In weekly meetings, TAs were encouraged to discuss their teaching strategies and struggles, thereby revealing hidden labor to the instructor such that it could be regulated. I analyze this process in terms of Foucault's confession, “a ritual that unfolds within a power relationship, for one does not confess without the presence [of]…the authority who requires the confession, prescribes and appreciates it, and intervenes in order to judge, punish, forgive, console, and reconcile” (Foucault, 1990[1976], pp. 61–62). Indeed, through ritualized sharing, TAs not only voluntarily confessed when they had knowingly deviated from rules, but provided knowledge about their labor such that new rules and regulations could be created.

I then show how the effects of this participatory management stretch beyond the weekly TA meeting. TA work can be understood as a form of service work located in a “service triangle” (Leidner, 1993), a site of shifting, situational alliances here composed of instructor, TA, and students. When faced with demands from both above and below, TAs may strategically draw on their ambiguous status as both student and instructor to manage these relationships and their workload. Further, TAs may ally with each other, whether to collectively deviate from rules in secret or to oppose the instructor directly. As I will show, the expression of instructor control shapes the possibilities of both horizontal and vertical alliances. When working for Professor Adams, the trust and transparency between worker and manager rendered alliances between TAs risky, as TAs routinely brought up their own, and others’ teaching practices in discussions with the instructor. Further, TAs’ public alignment with the instructor posed difficulties for strategic deflection of student demands and complaints.

One Model of Status Hegemony: Participatory Management

In the sections that follow, I draw on autoethnographic fieldnotes to examine the dimensions of hegemonic status control in one expression of TA management, a participatory model. I focus in the first two sections on the weekly TA meeting— the site of collaboration between TAs and instructor—where hegemonic mechanisms are most clearly displayed. In the third section, I examine how these mechanisms forge a fixed alliance between the TAs and the instructor, thereby limiting potential strategic alliances between TAs and between TAs and their students.

Collaborative Construction of Rules

In my first few weeks teaching with Professor Adams, I was surprised by, and perhaps skeptical of, her emphasis on collaboration. On the first day of lecture, she told students that she would be co-teaching this course with the TAs, referring to us as a teaching team. In our first TA meeting, she said she hoped to learn from the TAs this semester, especially those of us who had taught this course before. But this was not just talk. Professor Adams's approach to management would sincerely foreground the participation of TAs through conversation, collaboration, and collective decision making.

In our weekly meetings, TAs were invited to participate in the work of the instructor. We typically spent at least half the meeting discussing upcoming assignments. Sometimes, we collectively wrote prompts from scratch in these meetings, but more often Adams presented several preliminary options for us to work from and comment on. This led to long discussions about how students should be evaluated and how the assignments should be phrased. A few days after such a discussion, Professor Adams would email the TAs a final draft of the assignment to read over one more time for any final comments. Once the assignment was designed, consecutive meetings were spent collectively writing detailed grading rubrics.

This collaboration in some ways obscured the hierarchical relation between TA and instructor. As a participant in our conversations, Adams could make a suggestion or praise an idea, effectively encouraging certain behaviors over others. Of course, she also controlled the pace and agenda of the TA meeting: she determined what would be the result of collaboration and what she would decide on her own, and when we would shift discussion from one topic to another. Even when we were given many opportunities for input, I sometimes didn't say anything when I disagreed with the content of an assignment or rubric. Giving feedback on a professor's work proved to be a stressful task, as I struggled to balance sharing what I thought was pedagogically best with my desire to show respect and avoid offending her. For example, when Professor Adams sent the TAs ideas for the first paper prompt asking for our thoughts, I read through them and recorded the following in my field notes:

A lot of the ideas are for autobiographical assignments; how will I grade those? Some don’t explicitly ask students to draw on concepts or readings from the class; how would I go about teaching knowing that this is how they’ll be assessed? How should I phrase my comments to Adams? Can I propose an entirely different idea or should I just comment about bringing more course material in?

I decided to make just one comment, suggesting that we ask students to explicitly bring in concepts from the texts. Another TA agreed and my feedback was incorporated into the assignment.

While collaboration may have ceded only limited control to the TAs, it was nonetheless a powerful tool for securing TA investment in assignments and policies, and for limiting the TAs’ use of discretion outside of the designated spaces of collaboration. Grading rubrics offer an illuminating example. Much like with assignments, Professor Adams and the TAs extensively discussed and then determined the criteria and content of the rubrics. These were not general guidelines for grading, but detailed and technical systems to routinize it. Three to six pages in length, the rubrics provided a description for every half point that could be earned in a series of categories, including the student's use of evidence, specific types of analysis, and writing style. Once the rubrics were finalized, they effectively constituted the “rules” of grading. Early in the semester, some TAs deviated slightly from one of the grading rubrics, assigning points in smaller increments than was allowed.

Professor Adams asked how grading was going. Nick said he’d had a tough time with the 5% increments on the rubric, “Like some students weren’t 100s but they weren’t 95s and others were not quite 95s, they were 92s or 93s.” Professor Adams said she would not be able to write descriptions for every single point between 0 and 100 on the rubric, so we could not assign them. Nick would need to go back and assign grades in the 5% increments from the rubric.

As a solution to Nick's problem, Professor Adams suggested that we think about how to define the rubric categories more optimally for the next assignment. In this way, collaboration had a contradictory effect on TA autonomy: on the one hand, TAs had more control over their work, as they had direct input on the assignments and how they would be expected to grade them. On the other hand, the fruits of collaboration solidified into “rules” that TAs were expected to follow, thereby limiting the TA's use of individual discretion.

Confession as Rule Enforcement

Much of TA labor is hidden: TAs teach, grade, and communicate with students alone and without direct supervision. Given the autonomous nature of this labor, we might ask how rules for TAs are enforced. Professor Adams's participatory approach encouraged frequent discussion of hidden labor in the form of problem solving, commiserating, and even celebrating. While these practices of sharing felt largely nonhierarchical and emotionally satisfying, I examine them as an incitement to discourse (Foucault, 1990[1976]), in which TAs are encouraged to describe in detail their teaching actions, motivations, and concerns, producing both a sense of satisfaction for the TA and rendering their teaching labor more regulable. Through routine sharing, TAs “confessed,” both revealing when they knowingly deviated from established rules, and generally providing knowledge of their labor such that new rules could be created.

At the start of each TA meeting, Professor Adams opened the floor to any TAs who wished to share “highs” or “lows” of teaching from the past week. A “high” might include a story of a particularly successful discussion section, a kind comment from a student, or a sign of student progress. “Lows” typically revolved around a recurring set of problems: low engagement, absenteeism, poor performance, or interpersonal issues with students. Discussion of “highs” was typically very brief, while “lows” took more time, as they invited commiseration, support, and suggestions from other TAs and the instructor. In comparison to semesters with other instructors, this repeated practice generated a remarkable amount of transparency about what was actually happening in discussion sections. The TAs generally enjoyed this practice and actively participated: they often expressed relief that they were not alone in their teaching struggles. But such detailed sharing sometimes brought to light practices that Professor Adams, ultimately responsible for the course, objected to. In this way, transparency also allowed intervention in work routines that would have otherwise remained secret.

In one meeting, a TA shared that he suspected very few of his students were doing the readings, and that this was making discussion sections difficult to get through. A few TAs shared suggestions and one TA, Alexa, mentioned that she had been occasionally giving pop quizzes at the start of class so that students felt they needed to come prepared to every discussion section. Our conversation moved on, but a few minutes later Professor Adams stopped the discussion to follow up with Alexa: the pop quizzes are fine as long as they don’t factor into student grades. Alexa admitted that they did factor in but insisted that they were “not a big deal,” just a few sentences on a question about the reading, plus everyone had been getting full points so far.

Adams responded, there's also the issue of equity across sections. What if they’re getting more work just because they’ve randomly been selected into this discussion section? At this, another TA, Megan, spoke up and shared that other sections are having students write short memos or a few questions before class…so it seems like there are different definitions of “more work.” I was one of the TAs requiring memos before class and interjected that they are graded entirely on completion. Alexa went on to say that she was not grading based on correct answers, but that honestly, she needed students to think she was so that they would read for class. Adams said that Alexa could not suggest that they were being graded.

This example shows the ways in which solicitation of TA feedback and open dialogue can lead to greater regulation of TA labor. Alexa was using pop quizzes to improve her experience teaching discussion sections: students came prepared for class knowing that they may have an assessment. While forbidding pop quizzes may seem like a benign change, incentivizing students to read the assigned texts before attending discussion sections is a central struggle of TA work. I required short writing assignments for the same reason, and while students knew that these assignments were graded on completion, completion required a “quality” answer, in other words, that they had read the text. My own strategy was threatened when, in the above example, Megan revealed that it was also a common practice among the TAs.

Our collective transparency also allowed Professor Adams a view into our administration of course policy. These policies might include how we were meant to evaluate participation and attendance, or to handle cases of plagiarism or assignment extensions. Open-ended conversations sometimes led TAs to voluntarily confess when they had broken explicit policy. Take, for example, one conversation about late penalties and assignment extensions. At the time of my fieldwork, the conditions of COVID-19 had led to a proliferation of extension requests, which often could not be documented in the way traditionally dictated in course syllabi. Students requested extensions because they were ill, did not have regular access to the internet or a computer, were burdened with family care duties, had taken on additional waged work that interfered with their courses, or had new or deepening mental health issues from the pandemic, among other reasons. TAs are often the first point of contact for struggling students (Torres Carpio, this issue), and many of us became accustomed to granting extensions and accepting late assignments more freely, without documentation, and typically without consulting with our instructors.

In a meeting following a grading deadline for the TAs, Professor Adams said she’d noticed that many of the students’ papers had been turned in late. She asked what the TAs had done in terms of extensions and late penalties. She phrased this as an open question, despite having a clear no-extension policy and a late penalty policy listed on the syllabus. As we had learned to do in these meetings, one TA, Sophie, started to talk through her thinking on extensions.

Sophie said she’d felt very conflicted about extensions. She’d had a conversation with a different faculty member and some other TAs about extensions where they determined that it is unequitable to deny extension requests. So she granted all the requests she’d gotten. John spoke next and said that he had been surprised by the number of extension requests he’d received and had also granted them all. Another TA said she had not assigned late penalties to the papers that had been turned in late.

Rather than pointing out that these TAs had clearly broken with course policy, Professor Adams continued the conversation:

She said she wasn’t sure she agreed with that professor's approach. “I would love to hear more about it, but I think the most egalitarian thing to do is to have no exceptions, no performance, nothing.” Professor Adams then asked what each of the rest of us had done. One by one the TAs revealed how they had dealt with late assignments, some following policy, some not.

The professor listened, and when we finished, she reiterated her concerns regarding equity given our different approaches. Without a clear resolution, we moved on to another task. At the end of the meeting, Adams revisited the topic, saying that we would stick to the “original plan,” meaning the policy she had included in the syllabus. She acknowledged that we TAs would each have to figure out how to handle extensions for our own courses in the future, but for now we needed a consistent standard. The TAs who had not given late penalties agreed to correct the grades before they were released. In this case, the TAs had not actually come to an agreement with the instructor about how to handle these, but she had “heard us out” as future colleagues before doubling down on her policy. It is, of course, not remarkable that the instructor would make the final decision on matters of course policy. But the participatory process of this decision-making was a powerful hegemonic mechanism: TAs consented to the outcome of collective rulemaking and engaged in “confession,” bringing to light their deviations (and those of others) and making visible their labor such that it could be regulated.

Managing Horizontal and Vertical Alliances

The TAs enjoyed working with Professor Adams, and often expressed this sentiment to each other after our weekly meetings. We received support and help in these meetings and Professor Adams's insistence on collaboration made clear that she saw our opinions as valuable and essential to the construction of an effective course. This level of trust also had significant practical benefits: we had remarkable influence over the construction of the course we were teaching and could directly bring up issues we were encountering without fear of reprimand. However, such strong identification with the instructor posed difficulties for private alliances among the TAs. In the open discussions in the TA meetings, a TA might bring up another person's teaching practices, like when Megan brought up the additional memos some of us had been assigning during our conversation about pop quizzes. These practices were never brought up as accusations, rather they were often invoked as a compliment or suggestion, such as when Alexa had suggested her own practice of pop quizzes when troubleshooting a fellow TA's engagement problem. Yet this shaped the ways in which I interacted with my colleagues in private: I could not count on them to take my side or keep quiet if I deviated from decisions made collectively or by Professor Adams.

In a few instances, a TA would privately share with the other TAs that they had defied a course policy. For example, after a major assignment, one TA sent a message to the others that she was thinking of canceling her next discussion section since there wasn’t much to cover and her students were exhausted. A few other TAs encouraged her and said that they would follow suit. I texted that I would plan on canceling too. The next morning, John texted, incredulous, are we really allowed to do that? Another TA, Ava, reiterated the sentiment. I was surprised to hear this from Ava who had been a TA several times. Managing the TA workload often involves making calls that one would rather keep from the instructor. In my experience, TAs would occasionally cancel their sections without checking with the instructor when there was no new content to cover or when they had an emergency. While Jess and John's responses did not suggest they would bring it up with Adams, I held my breath during the TA meeting that day, certain that I would be found out. When they didn’t bring it up, I vowed that I wouldn’t discuss bending the rules with the group of TAs anymore. Later in the semester, a different TA asked the other TAs if we were really following the professor's policy of no extensions except in the case of emergencies. A few responded saying that yes, they were. I said that I too had been following the policy. This was true, but I hadn’t yet had a student request a nonemergency extension. I wrote in my field notes that I probably would grant one if I was asked, as I had in previous semesters, but I didn’t share this with the others.

Professor Adams's public emphasis on collaboration also had consequences for the strategic alliances we could make with students. TAs often make use of their ambiguous status between students and the instructor to manage competing demands from above and below. In the presence of students, a TA might express sympathy for the grueling demands of the course, building goodwill with students while suggesting the problem is out of their hands. In the presence of the instructor, the TA might commiserate over the difficulty of getting students to show up to discussion sections, prompting the instructor to intervene by sending a stern course-wide announcement about attendance. These shifting alliances allow TAs to strike a delicate balance of legitimacy and powerlessness in the eyes of students while maintaining a collegial and sympathetic relationship with the instructor.

When the first assignment of the semester was released, Professor Adams shared the essay prompt during her lecture. She asked one TA to read the first paragraph, then asked another to read the second. She told the students she was having TAs read it “as a reflection of the collaboration between TAs and instructor that had gone into designing this assignment.” While this was intended to enhance the TA's legitimacy, announcements like these created a lasting public alliance between TAs, the instructor, and the assignments. I faced two difficulties because of this alignment. First, I could not easily build solidarity with my students over challenging course demands. Rather, I was seen as a co-creator of those demands. As a secondary effect, students more often held me accountable for course policies, assignments, and grades that I did not fully control. Students perceived my influence over the course to be quite high and were more likely to direct complaints at a TA than at the professor. In one instance, when I prompted my students to ask questions about an upcoming paper, one student raised her hand and asked if she could go over the page limit.

I said no. A few lines is okay but beyond that isn’t. She said, well, I don’t see how you can expect us to answer all the questions in just 3 to 4 pages. Her tone seemed accusatory.

I had faced frustrated students before, in class and in office hours, and would typically manage these interactions by aligning myself with the students: expressing sympathy for their situation and suggesting I would be reasonable in my grading. In this instance, while I agreed that there might be too much to cover in three to four pages, I did not feel I could bring it up with the instructor nor did I feel I could publicly agree with the student and deviate from the rubric I’d helped design. I’d had an opportunity to give my feedback when we designed the assignment, so I felt I had to defend the expectations it contained:

It took me a second to think of something with everyone looking at me. I told her that everyone has to work within the same limitations, so if that ends up being an issue it will be an issue for everyone.

Thus, while the TA labor process is significantly self-directed, instructor management constrains TA autonomy by shaping the possibilities and limits of the alliances TAs can form, through which they may manage, contest, and deviate from instructor directives.

Conclusion: Status Control in the Extractive University

In this paper I have broadened Hatton's concept of status coercion to theorize status control as a form of domination that includes two types: despotic status control in which coercion prevails over consent and hegemonic status control where consent outweighs coercion. In studying TAs, I have argued that control relies on TAs’ investment in their status as future professors. My empirical example illustrates one expression of this hegemony, in which the instructor secures consent through participation: organized sharing of teaching practices and challenges, and (limited) TA involvement in course design. This study links expression of control in the labor process to possibilities for worker identification and alliances, a key concern for organizing worker populations obscured by other statuses. Duty and loyalty to students and faculty, as well as personal identification as a student and future professional, have been key roadblocks over decades of organizing academic labor (Dirnbach & Chimonas, 2003). Such a finding may be instructive for understanding organizing challenges among workers who have nonworker statuses, or those whose work is understood to “transcend” the imperatives of wage work, such as healthcare workers.

Of course, not all instructors manage TAs as Professor Adams did. In the experience of the authors of this special issue, the participatory model was unusual, though instructors often incorporated TA participation in select elements of their management. While the basis of power between faculty and graduate students remains the same, the dimensions I examined in the participatory model—the collective construction of rules, the enforcement of rules through confession, and the resulting alliance with the instructor—may vary in different expressions of status hegemony. Based on the experience of the authors of this special issue, I sketch two additional models of TA management.

In a second expression, which we may call “bureaucratic,” many instructors create a unilateral system of rules and procedures for TAs to follow, typically distributed in a manual of sorts. Rather than resolving issues of equity through extensive discussion as the participatory model does, the bureaucratic model attempts to resolve the issue of TA and student variation by pre-imposing standardized, if arbitrary, processes and outputs. Such an approach is welcomed by many TAs, as it relieves the burden of individual or collective decision making, though it can limit autonomy. As rulemaking doesn’t rely on discussion and consensus, enforcement may rely more on punishment. The bureaucratic model features prominently in examples of TA management that are circulated widely in my department each semester, offered as guides for managing TAs without infringing on their workload protections. In one such example, a professor outlines expectations, best practices, and some firm rules for teaching a social theory course in a long document distributed to his TAs. One rule established in this packet is that TAs must grade on a curve by assigning their students a predetermined distribution of A, B, and C grades, a system which attempts to impose a degree of equity in student outcomes. If a TA's grades deviate notably from the distribution, the professor will read through the papers himself. The manual makes clear that TAs may always make a case for deviating from the curve if they believe the quality of their assignments warrants it, so I asked this professor how often TAs take him up on this. It is apparently a rare occurrence. The possibility of a professor directly evaluating a TA's grading appears to be deterrent enough to ensure compliance with the policy.

In a third expression, some instructors develop rules for the course but make it known, implicitly or explicitly, that they do not expect TAs to follow them. Extension policies are an illustrative site for this “mock” ³ model of management. Some instructors will have an official “no extensions” policy in their syllabus, but make it known to their TAs that this policy is meant to reduce student demands on TAs and the instructor, and that they may deviate from the policy so long as they handle it themselves. As rules are not intended to be followed or enforced, this model gives TAs significant autonomy, with its accompanying isolation and potentially burdensome workload. In one such case of management, the instructor noted explicitly that the TAs would be the “shield” between the instructor and the students.

The instructor–TA relationship is not the only site of status control; indeed, work framed as “training” proliferates in the extractive university. ⁴ Some of this work can more convincingly be considered training, such as university workshops for the development of teaching statements, which will serve graduate students when they enter the job market. Others are heavily in service of the university, cloaked in the language of apprenticeship and opportunity. As I was writing this paper, TAs at my university were invited to join a “working group” to address inequalities in educational access brought to light by the COVID-19 pandemic. The group requires a 5-month commitment and active contributions to bi-weekly meetings, individual research of best practices, and meeting facilitation. Participants will all the while be creating resources and best practices that can be used by faculty and staff campus-wide. They will also have the “opportunity” to serve as ambassadors for these resources within their departments. Further framing the work as a professional opportunity, TAs must apply to join, and—given the benefits to their present and future teaching careers—they will not be paid. The invitation notes that it is TAs who are best positioned to address educational inequalities as they interact with students in the smallest settings. The “heroism” of excellent teachers is often invoked alongside apprenticeship status to secure consent from TAs to do this additional work, work they hope will confer distinction in a shrinking job market. TAs are training not only to be teachers, but excellent, heroic teachers, who will, through individual effort, address the inequities of the student body. As universities rely increasingly on graduate students and contingent faculty for teaching labor (Colby, 2020), the work of TAs can less convincingly be constructed as apprenticeship. Rather, apprenticeship becomes a way of securing and obscuring extraction, the growing contributions of TA labor to the university.