Introduction

Doctoral supervisory interactions are closely associated with students’ academic development (Al Makhamreh & Stockley, 2020; Blaney et al., 2020; Börgeson et al., 2021; Dericks et al., 2019). Research-based doctoral students are often under substantial pressure, which may be caused by supervisory interactions (Mackie & Bates, 2019; Freeman & Kochan, 2019; Gibson et al., 2019). However, supervisory interactions generally play important roles in doctoral education, and affect doctoral students’ experiences of pressure, satisfaction, and academic success (Jiang & Shen, 2019; Katz et al., 2019; Lindahl et al., 2021). Mackie & Bates (2019) have argued that supervisory interactions are a source of stress for doctoral students, together with workload, lack of transparency in university processes, role conflicts, financial insecurity, and uncertain career prospects. Dericks et al. (2019) found that supervisor support is the strongest predictor of doctoral academic satisfaction. Therefore, supervisory interactions play a fundamental role in the academic development of doctoral students (Mackie & Bates, 2019; Mullen, 2021; Pappa et al., 2020). Especially within the context of transnationalism and neoliberalism, the interactions between doctoral students and supervisors, and the role of these interactions in shaping students’ academic achievement, are likely to undergo significant changes (Akala & Akala, 2023; Wofford et al., 2021).

Although previous studies have analyzed the role of doctoral supervision in doctoral students’ academic socialization, few studies have provided empirical longitudinal evidence using ethnography to determine the dynamic process of how doctoral supervisory interactions affect the production of academic culture. Addressing this gap, this study used a semi-collaborative mixed ethnographic approach, extending Bourdieu’s theory of cultural reproduction to the supervisory interaction field between doctoral students and their supervisors to explore how supervisory interactions mediate doctoral students’ academic cultural production. In this study, the process of (re)production of doctoral students’ academic culture is essentially the intergenerational transmission process of academic cultural capital, academic social capital and habitus between supervisors and students in the guidance interaction. The main research question was “How do doctoral supervisory interactions shape students’ academic cultural (re)production?”

Contextualizing academic cultural (re)production: Hierarchical academic culture in Chinese doctoral supervisory system

In China, the ethical tradition of Confucian culture plays an indisputable role to influence the relations between supervisors and students, which is a part of an academic lineage of “Kindred blood.” Although there are various types of interaction in supervisory relationships, interactions within these relationships are considered to be inevitably bound together in the “Academic circle”(学术圈), and the superior cultural capital accumulated by the supervisor is assumed to be further consolidated and maintained through the academic system (Wang & Byram, 2019; Bao et al., 2024). In China’s higher education system, there is a sense of “kinship” in preserving the traditions of the cultural aristocracy. Hierarchical structures in Chinese doctoral education are not only culturally ingrained but also structurally reinforced through both academic lineage and administrative authority. Ying et al. (2017) argued that even highly competent scholars with lower administrative status may face structural disadvantages, while students mentored by high-ranking professors are disproportionately favored in resource and award distribution. Supervisors are not only as academic guides but also as moral and ideological mentors. Informal channels such as WeChat serve as extensions of formal supervision, reinforcing hierarchical dynamics and shaping students’ academic identities (Wang & Byram, 2019; Bao et al., 2024). This deeply rooted structure affects not only material opportunities but also students’ academic socialization. In this sense, many Chinese students learn to obey, follow, and imitate their supervisors’ preferences and research styles as a strategy to gain recognition within the academic community. It is an adaptation process that not only reinforces academic dependence but also contributes to the (re)production of academic habitus through supervisory (Ying et al., 2017).

Examining the role of supervisory interactions in doctoral student development

Doctoral education, particularly research-based doctoral programs, concentrates on the quality of the supervisory relationship (Guzmán-Valenzuela, 2016). The supervisory interactions play a pivotal role in the academic development of doctoral students (Porat-Dahlerbruch et al., 2021; Richards & Fletcher, 2020; Ryan et al., 2022). There are various aspects of this relationship, including the impact of different supervisory styles on student satisfaction (Gruzdev et al., 2020), the quality of tutorial systems (AI Makhamreh & Stockley, 2020) and the mental health benefits of hands-on support from postdoctoral fellows (Blaney et al., 2020). Janssen et al. (2021) examined the alignment of relational schemas between supervisors and students, highlighting tensions and offering insights into the success or failure of mentoring relationships. Richards & Fletcher (2020) characterized doctoral education as a “critical friendship” and emphasized the importance of balancing support, maintaining a social relationship, and fostering student independence.

Supervisory interaction as the negotiation of academic capital: Transnationalism and neoliberalism

The previous study has begun to conceptualize the interaction between supervisors and doctoral students as a key site for the negotiation of academic capital (Dai et al., 2024). This perspective highlights the dual role of supervisory interactions as both a hub for the allocation of institutional resources and a crucible for shaping academic habitus (Akala & Akala, 2023; Wofford et al., 2021). For instance, Akala & Akala (2023) demonstrated how South African doctoral students leverage structured seminars and peer networks to convert social capital into academic legitimacy. Conversely, Wofford et al. (2021) found that first-generation scholars in STEM fields actively seek explicit skill transmission from supervisors to compensate for cultural capital deficits, a strategy that contrasts with the “misrecognition” (reconnaissance) of independent guidance as individual ability by continuing-generation students.

Transnationalism and neoliberalism also play key role on the supervisory interactions

The negotiation of academic capital is further complicated in transnational contexts. Pretorius et al. (2021) illustrated how supervisors, acting as “capital gatekeepers,” can force international students to develop a “suspended habitus” (Xing et al., 2022) as they navigate conflicts between neoliberal publication pressures and their own cultural backgrounds. Studies by Dai et al. (2024, 2025) and Lou (2024) have shown how supervisors facilitate the cultivation of “technical-linguistic capital” and “diplomatic capital” among international students, respectively, through strategies such as bilingual framework design and the strategic blending of collectivist and individualistic academic practices. Hodge & Murphy (2025) highlighted the value of alternative forums like writing groups for developing “relational capital” and challenging traditional hierarchies of knowledge transfer. In addition, the previous studies above contribute to a growing understanding of supervisory interactions within a Bourdieu’s theories, highlighting three key theoretical advancements: (1) the institutional alchemy of academic capital, where social, cultural, and technical resources are transformed into academic legitimacy; (2) the role of supervisors as “habitus architects” who reshape students’ subjective cognition through both explicit and implicit guidance; and (3) the impact of neoliberal audit cultures, which have alienated guidance into a capital productivity indicator, prioritizing quantifiable outputs like co-authored papers over ethical considerations.

Identifying the research gap in doctoral supervisory interactions

There is a research gap in our understanding of the process of supervisory interactions over time and their impact on academic cultural (re)production. The existing work relies on cross-sectional data or qualitative interviews that capture snapshots of the supervisory relationship. However, there is a lack of longitudinal studies that delve into the nuances of how supervisory interactions evolve and shape students’ academic trajectories. This gap is particularly evident in the context of understanding how academic capital is accumulated, negotiated, and internalized by doctoral students over the course of their studies. For example, the studies by Bao et al. (2025) and Bao & Feng (2023) have begun to address this gap by employing digital micro-ethnographic approaches to examine supervisory interactions. Bao et al. (2025) found that supervisors utilize transformative criticism, strategic recognition, and their own identity to mediate students’ academic identity construction in an iterative process involving negotiation. Bao & Feng (2023) demonstrated how supervisory feedback facilitates students’ academic literacy development in writing for publication (WFP) by unpacking hidden knowledge about academic genres, inducting them into social relations of scholarly communication, and encouraging original contributions. Carless et al. (2024) further contributes to this emerging area by investigating feedback socialization in doctoral education, highlighting the role of authentic feedback, such as journal peer review, in facilitating students’ socialization into academic publishing norms. These above studies underscore the importance of understanding the dynamic and situated nature of supervisory interactions. However, further research is needed to fully understand the longitudinal process of academic cultural (re)production within supervisory relationships.

Thus, this study aims to address this research gap by employing a semi-collaborative mixed ethnographic approach, utilizing a variety of data sources collected over a three/four-year period to provide a detailed, longitudinal examination of how supervisory interactions mediate doctoral students’ academic cultural (re)production. This approach provides a nuanced understanding of the dynamic interplay between supervisors and doctoral students, ultimately contributing to the development of equitable and effective doctoral cultivation programs.

Theorizing academic cultural (re)production: Bourdieu’s theory of cultural reproduction as theoretical lens

We used Bourdieu’s theory of cultural reproduction to investigate supervisory interactions and doctoral students’ academic cultural (re)production. The higher education cannot integrate the society and break the barriers of class but solidify the status quo of class and cause the reproduction of social structure (Huang & Li, 2019, p. 4). “Field,” “habitus,” and “capital” are the three elements that constitute Bourdieu’s theory of cultural reproduction. The logic of the theory of cultural reproduction is as follows: education is an act of symbolic violence, through which schools carry out the cultural reproduction of the dominant class, and cultural reproduction eventually leads to social reproduction, which is what Bourdieu called the “social alchemy” of the field of higher education (Huang & Li, 2019, p. 4). In this study, the process of (re)production of doctoral students’ academic culture is essentially the intergenerational transmission process of academic cultural capital, academic social capital and habitus between supervisors and students in the guidance interaction.

Capital

Bourdieu divides capital into three types: economic capital, cultural capital, and social capital (Huang & Li, 2019, p. 62). This paper attempts to extend Bourdieu’s theory of capital to the academic field, discussing academic cultural capital and academic social capital. We argue that doctoral academic capital is the core concept of supervisory interaction. Bourdieu divides cultural capital into three forms: embodied, objectified, and institutionalized, arguing that the acquisition of cultural capital is divided between family and school. Embodied cultural capital refers to the spiritual cultural products such as upbringing, knowledge, taste, and skills retained within an individual’s body through education and family environment. Objectified cultural capital is the material form of spiritual achievements, primarily existing in the form of cultural goods. Institutionalized cultural capital mainly confers status on individuals through academic qualifications, which can be simply described as issuing certificates or diplomas (Huang & Li, 2019, p. 68). In this study, the concept of academic cultural capital includes personal academic accumulation (including problem awareness, theoretical foundation, research method foundation), personal academic achievement (including the transformation of academic results, the quantity and quality of published papers, academic scholarships and honors), and personal academic qualifications and degree certificates. Social capital is a collection of actual or potential resources that are inseparable from the possession of a durable network, a network that is commonly known, recognized, and characterized by institutionalized relationships. In essence, social capital is a combination of social resources such as social relations, social status, and reputation. (Bourdieu, 1998, p. 202) In this study, the concept of academic social capital is reflected in an individual’s reputation in their field of academic research, personal academic standing, and academic networks (including sources of research data, journal contacts, published papers, and peers in related fields).

Field

Within supervisory interactions, supervisors are the dominant academic power and thus have more academic capital and are in a strong position in the field of power. Within this field, they tend to impose the rules or “game” methods that are most favorable for them and impose their own will on the whole field. Bourdieu believed that the struggle between fields is ultimately a hierarchical principle imposed by the subjects of fields (Bourdieu, 1984). The existence of inequality is the core issue revealed by Bourdieu’s cultural reproduction theory. The amount of academic capital occupied by supervisors and doctoral students determines their status, resources and discourse power in the field of higher education. On the one hand, there is an inequality in academic status between supervisors and doctoral students.

Habitus

Within supervisory interactions, there are substantial individual differences between doctoral students, which are mainly expressed in students’ personality, how compatible they are with their tutors, and the different habitus formed under the influence of their family of origin. Habitus is formed by conditioning in relation to a particular class of conditions of existence. It is a system of enduring and changeable dispositions, constructed structures that tend to act as principles that give rise to and organize practices and representations (Bourdieu, 1990, p. 53). Habitus is a kind of mental schema that persists in people’s thinking, language, and behavior patterns (Huang & Li, 2019, p. 93). In this study, doctoral students’ habitus is the acquired “nature” of an individual. It is the imprint of academic environment in the spirit, which can be awakened at any time and does not easily disappear. We believe that the formation process of doctoral students’ habits is the process of shaping their academic identity and the process of academic socialization, which realizes the transformation from doctoral students to scholars. In this study, the formation process of doctoral students’ habits is the process of shaping their academic identity and the process of academic socialization, which realizes the transformation from doctoral students to scholars. We believed that field is an objective prerequisite and plays a social prescriptive role in the relationship between doctoral students’ subjective will and behavioral choices.

Method

Research design and rationale

To obtain a nuanced understanding of doctoral students’ academic cultural (re)production, an original approach, semi-collaborative mixed ethnography was employed, which comprised the following steps. The first step involved selecting research questions on supervisory interactions and doctoral students’ academic cultural (re)production. The second step comprised locating the research object as a focus group. The third step was the data collection. Over 3 years of tutoring interactions between researchers and research participants, research data were collected using formal and informal, online and offline interviews, observations, and chats. The fourth step was data analysis. In the fifth step, subtopics and problems to be analyzed were identified from the data, the theories and themes relevant to the research question were selected, and literature searches and analyses performed to address the research question. Drawing on Bourdieu’s theory of cultural reproduction, a theory of doctoral academic capital (re)production was generated. Thematic coding was used to conduct preliminary coding of interview materials according to the three concepts of academic capital, field, and habitus. The contents of the coding were then compared and classified using a process of continuous preliminary coding (see Fig. 1).

Fig. 1
figure 1

Process of the semi-collaborative mixed ethnography to analyze data on doctoral students.

Full size image

The rationale of semi-collaborative mixed ethnography is given in below. The 3-year study period and the binding class counseling relationship between researchers and participants form a good foundation for the study as an ethnography (Britannica, 2024; Wieser & Pilch Ortega, 2020). In this way, this study focuses on the changes in research objects and facts, centering on the unfolding of social processes, emphasizing how, who, when, and where in the findings (Becker, 1998; Weiss, 1995; Tavory, 2020). “Semi-collaborative” stress the collaborative but unequal voice among the researchers, which can avoid one-side view from traditional collaborative ethnography or autoethnography that stress the same identity and equal voice of researchers (Hernandez et al., 2014). The first two researchers are professors and class tutor, while the other researchers are research assistants who have participated in research and have themselves been supervised. Such relationships among the researchers in the study ensured flexibility for the determination of the researcher and provided a multidimensional perspective that permitted a range of analyses. “Mixed” stress the combination of different digital media data with real-world observations. Observation of everyday life in the context of contemporary social relations can be difficult in ethnographic practice, as in many societies social interactions are increasingly influenced by digital technologies (Hallett & Barber, 2014). This study mainly occurred from September 2021 to December 2023, involved the pandemic period. The field investigations were conducted using social media and online platforms, and different characteristics of doctoral students were determined through virtual observations. Formal interviews commenced, toward the end of the study, immediately after the COVID-19 pandemic by face-to-face communication or digital device. Therefore, a combination of online and offline methods was used. This type of approach is more suitable for ethnographic research in a complex digital society, breaking with complete virtual approaches, e.g. Cyber ethnography, that investigate a single ethnographic network.

Researchers, participants and data collection

In this study, semi-collaborative relationship among researchers were formed. The first two researchers were professors with PhD title, and the other researchers were research assistants and graduate students who have participated in research and have themselves been supervised by the researchers. In China, the doctoral students have only been supervised by an academic supervisor, they are also in a class with a tutor who is charged of their daily administrative fair including safety, scholarship, employment, mental health, etc. The first two researchers have both the two identity as academic supervisor and class tutor. Therefore, 41 academic doctoral students aged 25–28 years in six disciplines in five Chinese world-class universities were conveniently recruited due to the academic and administrative counselling relationship between the first two researchers and the participants. The Chinese doctoral system requires the publication of at least two core journal articles and a graduate thesis for degree conferral. Furthermore, the strong supervisory model in China often results in supervisors guiding students beyond basic requirements, encouraging participation in surveys and practical activities to foster comprehensive skill development. This study, therefore, adopted a semi-structured interview approach, examining students’ academic trajectories, accomplishments, supervisory relationships, career plans, emotional experiences, and perceived challenges. Six rounds of qualitative, thematic interviews were conducted at three timepoints (during December 2021, December 2022 and May 2024) over three years. These formal interviews were conducted online and offline. The content of the interviews was recorded and exported using recording software. After the removal of invalid interview data unrelated to the topic, 620,418 words remained.

Data analysis

To explore the key themes regarding how supervisory interactions shape doctoral students’ academic cultural (re)production, a reflective thematic analysis was first conducted. We used Braun & Clarke’s (2006) six-step procedure to analyze the data and then coded the data systematically. The initial coding process was guided by the theoretical concepts. The codes were sorted into separate themes consistent with the research question. Both online and offline data were analyzed to identify specific themes. The first two authors and the co-authors ensured the consistency and credibility of the data analysis. During the process of data analysis, intensive discussions were conducted by the first two authors and the co-authors. Through several rounds of thematic analysis, three major themes were generated: inheritance, dependence, and imitation; confrontation and recognition; Accumulation and internalization, these themes characterized the accumulation of academic capital. Using Bourdieu’s theory of cultural reproduction (Bourdieu, 1998) as a theoretical lens, the analysis addressed the process of doctoral supervisory interaction in relation to the interactional and wider sociocultural contextual background. Intersubjectivity offers a productive frame of reference for analyzing and identifying how doctoral supervisory interaction influences field, habitus, and capital as three elements that shape doctoral academic cultural reproduction.

Trustworthiness

The establishment of trustworthiness is important in qualitative research and can be achieved through prolonged participation and involvement by the researchers. In this study, we tracked doctoral supervisory interactions over 3 years and retained close relationships with the participants. The raw data were collected from various sources to ensure data triangulation. Through several rounds of research, we examined how our positionalities as researchers were reflected in this study. The first author’s familiarity and friendship with the participants helped to promote the credibility and dependability of the data collection process. The first and second authors retained trusting relationships with the participants, who were able to share their emotions and thoughts on their doctoral supervisory interactions. During the data analysis process, our insider positionalities as supervisor (i.e., the first and second authors) and students (i.e., co-authors) were considered analytic resources that provided more in-depth perspectives on doctoral supervisory interactions. We also performed an “double check” interview to encourage selected participants to express their original opinions on our data interpretation. This study was approved by the human participants ethics committee of the authors’ B university (No. CIA220282).

Informed Consent Statement

All participants were provided with detailed written information outlining the purpose of the research, study procedures, potential risks and benefits, their consent to publish findings, and the voluntary nature of participation – including their right to withdraw at any time without penalty. Written informed consent was obtained from all participants before data collection commenced. As the study was initiated during the COVID-19 pandemic, the informed consent process was administered in September 2020 when participants were beginning their doctoral studies. Documents containing the informed consent agreement were electronically distributed via social media platforms. Participants provided their signatures on digital copies, which were subsequently archived as verifiable records.

Findings

Participants reported that during doctoral supervisory interactions in China, there is a sense of “kinship” in preserving the traditions of the cultural aristocracy. Such narratives illustrate how doctoral education in China may function as a system of academic (re)production, an idea that can be analytically understood through Bourdieu’ s theory of cultural reproduction. Drawing on the analogy of aristocratic succession, Bourdieu (1990) emphasized how symbolic and cultural capital are transmitted intergenerationally within structured fields, reinforcing social hierarchies. This framework aims to interpret doctoral supervision as a process embedded in the reproduction of academic capital and institutional legitimacy.

This study adopts a temporal analytical structure-framing the doctoral journey across early, middle, and later stages-to advance an interpretive framework inspired by Bourdieu’ s theory of cultural reproduction. The categories of inheritance, dependence, and imitation are heuristic lenses developed in this study to trace how doctoral students’ engagement with academic practices evolves over time. In the early first-year stage, students tend to draw on pre-existing dispositions and external motivations (“inheritance”); in the middle second-year stage, their academic activities become increasingly shaped by institutional expectations and supervisory dynamics (“dependence”); and in the later third-year stage, they begin to independently enact and reproduce disciplinary norms (“imitation”). This structure highlights the dynamic and recursive nature of academic identity formation and underscores the role of supervisory interactions in the (re)production of academic capital and habitus.

Inheritance, dependence, and imitation: Accumulating academic capital during first-year period

For the first-year doctoral students, they become the “inheritors” of the supervisor’s academic cultural capital, and this cultural capital can be reproduced. In addition, the academic capital of the supervisor can directly help the first-year doctoral student accumulate cultural capital through social capital (such as contacts and reputation) accumulated throughout their career. We found that supervisors used different ways to help doctoral students accumulate academic capital, and that doctoral students inherited academic capital from their supervisors in different ways:

My supervisor said that GMH (another doctoral student) was like his son, and I was like his daughter, who should be spoiled. My article is going to be published by Xinhua Digest because of my doctoral supervisor. I think I am now very dependent on my supervisor and afraid to submit articles on my own. After leaving my supervisor, I feel like I can’t cope on my own. (CY-F-2021)

At this time, it is particularly important for doctoral students to rely on their own ability and strive to grasp the resource opportunities given them by their supervisors. In addition, by participating in projects with my supervisor, I can accumulate social network resources, which is helpful in opening the channels of research and investigation. (CY-F-2021).

Although some of the interviews were conducted in the students’ final doctoral year (2024), they reflect retrospective assessments of their early-stage supervisory experiences. This means that although the interviews took place in their final year, the students were reflecting on their experiences during their first year of doctoral study. These reflections highlight how interactions with their supervisors in the first year influenced their long-term academic dependence and publication strategies. Early-stage supervisory interactions had a lasting impact on the formation of students’ academic identities and their publication strategies, contributing to the gradual accumulation of academic capital. Additionally, the author with doctoral experience also resonated with these viewpoints and reflections, suggesting that such experiences are common across doctoral journeys and not unique to the specific cohort under study.

I just want to say, don’t rush to graduate. After graduation, I may not be able to work with my advisor to publish articles (laughs), so it is better to accumulate more publications with my advisor now. (YYC-F-2024)

The most important reason why I have so many academic achievements is the guidance of my advisor! My advisor gives me a lot of research opportunities. (GMH-M-2024)

As a researcher who has experience of doctoral education myself, I resonated strongly with this point during the analysis. Publishing with a well-established supervisor often helps to ease some of the invisible barriers in the publication process, making peer review and acceptance noticeably smoother. This kind of reputational endorsement not only facilitates access to academic platforms but also subtly shapes doctoral students’ confidence in their own scholarly capabilities.[Author’s Interpretation]

Dependent supervisory interactions were characterized by direct transfer of academic capital from supervisors to their doctoral students. First-year doctoral students experienced the convenience of accumulating academic capital from their supervisors’ social and academic capital. However, this increases the dependence of students on supervisors. Well-known and influential supervisors tend to have more academic performance resources, which are also passed on to doctoral students in various forms:

In addition, after the completion of the project, my supervisor will organize experts to dig out some academic ideas in the project and then make effective use of these data. If we are fully involved in the project, we have access to the data. So, we can use these data for further academic research. (SXR-F-2021)

I didn’t have much experience in publishing papers before. I was mainly guided by my supervisor on how to write and how to communicate with editors. (ZM-F-2021)

My advisor is academically rigorous. He has also changed a lot of my previous writing habits. I used to write more casually, but I am now more aware of the need to consider my choice of words. He mainly provides one-to-one guidance in person. (GAP-F-2021)

On the basis of passive acceptance and inheritance, some doctoral students will also exert their subjective initiative, observe the supervisor’s research interests and academic personal style, and take the initiative to imitate the supervisor’s words and deeds under the premise of recognizing the supervisor, to align with the academic spirit of the supervisor, to help themselves accumulate academic capital and establish a foothold in the “academic circle”. Due to the convenience of information, abundant resources and sufficient guidance, many doctoral students will imitate the research methods and ideas of their supervisor’s projects and projects, dig out the topic selection of their graduation thesis by participating in the supervisor’s projects and projects, and do similar research direction with the supervisor. Then reduce their own graduation thesis topic selection pressure. The data in the project and the project development process are also transformed into their own academic capital with the participation of doctoral students in the project.

According to the experience of our teachers, the final topic of graduation thesis is generally followed by the teacher’s topic. However, at present, the subject I have meet is not suitable for writing a paper, so I wonder if I can wait until a good topic can be used to follow the idea to do the graduation design, to save some steps of argumentation by myself. (ZM-F-2021)

The instructor is biased towards theory, so I also follow the instructor’s research interest, and now I will start to go in the direction of this modern theory, look. (WH-M-2021)

During the COVID-19 pandemic, the social distancing rules may have reduced the possibility for the exchange of academic capital between supervisors and doctoral students. Supervisors also must cope with developments and difficulties in their own personal lives and careers. Prolonged separation between supervisors and doctoral students made it difficult for supervisors to provide adequate and systematic academic guidance to doctoral students:

My tutor is on sick leave. I don’t want to disturb him; I want him to have a good rest. (WQ-F-2021)

I haven’t talked with my tutor yet, because my tutor has gone to the United States as a visiting academic. (LZC-M-2021)

I understand how limited access to networks and opportunities can make early academic development difficult. The pandemic further heightened these struggles, as reduced interaction with supervisors made it even harder to gain the guidance and support essential for academic growth. [Author’s Interpretation]

Confrontation and recognition: Gaming in the academic “field” during second-year period

In the second-year doctoral supervisory interactions, the concept of academic capital cannot be separated from that of academic “field” (Bourdieu & Wacquant, 1992). The doctoral student–supervisor interaction “field” and academic “field” are contained within the field of power and dominated by it. It is a highly stratified ranking system with academic reputations and scholar titles as symbols, and a power field of social games and struggle. Supervisors typically hold dominant positions, while students often occupy subordinate roles. Although some supervisors emphasize student autonomy, such as the idea that students “must make their own decisions” as expressed by participant ZZY-F, there may be competitive games that create different types of mentoring relationships to control the legitimate form of authority unique to this academic “field”.

Within this academic “field”, doctoral students navigate tensions between their intellectual autonomy and their supervisors’ expectations. “I feel exhausted in discussions with my supervisor. My proposal was changed two or three times and then shelved…every time we speak, I feel that my supervisor’s requirements are relatively high, and I feel that they do not agree with my chosen topic” (WH-M-2022). Similarly, uncertainty and hesitation can emerge when students attempt to align personal interests with perceived supervisory approval: “I chose this topic because I am more interested in it, but I do not know whether my advisor agrees” (ZTR-F-2022).

Bourdieu used the term “game” to illustrate how fields operate, emphasizing that participation implies agreement with its rules and pursuit of its promised rewards. Doctoral students generally agree with field rules and aspire to obtain the ultimate academic capital. However, once doctoral students accumulate symbolic capital, they may begin to question or “resist” established norms. In this sense, disagreement over thesis direction can be seen as a form of symbolic struggle within the mentoring field, reflecting students’ attempts to renegotiate their positions and gain agency in the academic game.

I remember that when I was doing my doctoral thesis, I had a little difficulty in changing my identity from a student to a scholar. And sometimes I’m a little unsure about which tasks I should make my own decisions about and be strong about my own ideas. (ZZY-F-2023, recalling the second year)

Having undergone doctoral training myself, I can relate to the challenge of finding the balance between respecting a supervisor’s authority and asserting one’s own academic judgment. In the early stages, it’s often unclear when to follow guidance and when to assert personal ideas, leading to hesitation and confusion in discussions with supervisors. [Author’s Interpretation]

In the interviews, participants argued that doctoral students tend to be in a dominant position in the field of mentoring interactions and are at a disadvantage in the game with their supervisors, whereas tutors with abundant academic capital tend to be in a dominant position. Some participants argued that this power imbalance limits students’ agency and often leads to implicit obligations beyond research. For instance, “Last year, our institute assigned academics the task of writing a book manuscript…I helped my tutor select a topic and write the manuscript, so in the last 3 months, I have written more than 100,000 words…I must write nearly a third of the book” (LZC-M-2022). This kind of “invisible labor”, undertaken in exchange for symbolic academic recognition, often comes at the expense of students’ own thesis progress and may ultimately hinder their ability to accumulate academic capital. As one student observed, “Teachers have more frontline research tasks, and they may have four or five projects a year…This occupies a lot of academic time” (SXR-F-2022), suggesting that supervisors” priorities may further delay timely mentoring. Although SXR-F pointed out that there was “no obvious competition” between supervisors and students, the hierarchical structure means that mentors hold total power in defining the value and legitimacy of students’ contributions.

Accumulation and internalization: Shaping the academic habitus during third-/fourth-year period

In the third- and fourth-year doctoral supervisory process, students’ academic habitus is developed through the long-term practice of academic behavior, supervisory interaction experiences, and the academic environment. The peculiar validity of the patterns of habitus, the primitive categories, arises from the fact that they function outside consciousness and discourse, and are therefore not subject to conscious scrutiny and control (Bourdieu, 1998, p.737). In other words, the graduated doctoral students’ academic habitus is a social disposition system that is internalized in the daily academic behaviors of individuals and accumulated in their cognitive and motivation systems. The academic cultural (re)production process of doctoral students consists of the internalization of objective and academic rules and values. This process is also embodied in doctoral supervisory interactions in a conscious and persistent way, as academic thinking, perception, and action with cultural characteristics.

Interview data revealed that this habitus is not only socially shaped but also historically situated, emerging from past interactions and institutional familiarity. “I was also an undergraduate at this university…I took my supervisor’ s course on the history of Chinese education, so I am very familiar with him…I have found that the process of communication with my supervisor has been smooth” (LY-M-2023). Such long-term exposure facilitates intuitive communication patterns between students and mentors, reducing friction in supervision.

During the interviews, we found that the academic habitus of the graduating doctoral students originated from their different family backgrounds and academic experiences, which affect the student-supervisor relationship and academic capital in the academic “field.” The concept of habitus emphasizes the continuous (past and present) influence of others (people or environments) on an individual. It is socially formed and guides how students act, respond, and build relationships in the academic field (Bourdieu, 1984; Brubaker, 2005). In the academic field, academic habitus can also affect the supervisory relationship through unconsciously represented choices and can then affect the accumulation and transformation of academic capital.

Another doctoral student remarked: “Because my master and doctoral supervisor is the same person, I am familiar with his temperament. So sometimes you know when to go to him, and what kind of help about which issues to seek from him” (SWY-F-2023). This quote illustrates how embodied familiarity enables students to strategically navigate supervisory interactions based on tacit understanding, an example of how habitus functions as an intuitive sense of one’s academic position. Participants argued that any interactions between doctoral students and supervisors before the formal admission of the student will reveal each other’s temperament, and the habitus of the mentor will be constantly influenced by each other and reconstructed.

The third/fourth-year doctoral students who are more in tune with their supervisors’ temperament and existing academic habits tend to inherit and accumulate more academic capital and turn the academic resource inclination and academic relationship network brought by their supervisors into their own academic advantages. This accumulation and internalization, in turn, reinforces academic behaviors and strengthens students’ habitus. In contrast, students who are less able to adapt or communicate effectively with their supervisors often struggle to undergo similar accumulation and internalization, forming the habit of “being afraid to show their shyness” and “avoiding communication”. These patterns of academic habitus offer limited support for the improvement and growth of their academic capabilities. In contrast, students who are less able to adapt or communicate effectively with their supervisor’s struggle to undergo similar processes. They often develop avoidance behaviors and internalize an academic habitus characterized by timidity or silence, which hinders their academic growth. “My supervisor is very kind, but I tend to be afraid of communicating with them… I never communicate with my supervisor alone or ask to meet them alone” (XQQ-F-2023).

At the same time, such accumulation and internalization can also feedback doctoral students and give them positive feedback, thus shaping and strengthening the academic behavior “habitus” like that of doctoral students and their supervisors. On the contrary, the third/fourth year doctoral students who are unable to adapt to and fully communicate with their supervisors are difficult to complete similar “accumulation and internalization”, forming the habit of “being afraid to show their shyness” and “avoiding communication”, and such academic habits are difficult to help their academic ability to improve and grow.

In the third and fourth year doctoral supervisory interactions, some doctoral students are introverted and are strongly influenced by the strict manner and the traditional “patriarchal” approach of the mentor–mentee relationship. They are more likely to develop behaviors that correspond with this approach, and automatically become the weak party in the relationship. This affects the accumulation and internalization of their own academic capital.

For example, recently I was writing a short paper and was not very happy with it. I needed guidance from my supervisor, but I just didn’t know how to communicate with them. This might be because of the age difference between us; my supervisor is about to retire (they have worked in academia for 24 years). In addition, my supervisor is visiting Australia and so there is a time difference between us. I was studying at Shanxi Normal University for a master’s degree, and because my master’s tutor was very strict, I was afraid they would see how shy I am. I am still afraid to reveal my shyness. (XJ-F-2023)

Bourdieu (1992) argued that habitus acquired in the family is the foundation of the structure of the school experience (Bourdieu & Wacquaant, 1992, p.134). The internalized habitus acts as a “defense mechanism” that underpins subsequent learning and social experience. In this study, some doctoral students from financially secure families reported low pressure to complete their degrees on time. “My parents are not intellectuals and do not pressure me to complete my doctorate. Therefore, I can enjoy 4 to 5 years of study before I graduate” (ZYQ-F-2023). Others, however, faced higher parental expectations and economic pressure, prompting them to urgently accumulate academic capital in pursuit of timely graduation. Family cultural capital mediates this process. For some, it provides flexibility. “I am the only child. My parents hope I will study in Beijing, but I hope to work somewhere different later” (CY-F-2024). Others described diverging paths from siblings and a sense of self-direction. “I just talked with my brother… he wants to stay with my parents in Tianjin, but I plan to study in Japan. My parents completely understand me… they often say, ‘We can’t help you now’ (LK-M-2024).

For example, for participant ZYQ-F, the fact that their parents were not highly educated gave them more autonomy in doctoral supervisory interactions and reduced the psychological pressure of completing their doctorate and graduating. This created space for academic (re)production. However, some parents interfere with the academic (re)production of doctoral students in other ways owing to their own limitations. Some students mentioned that their family wanted them to get married or that their parents wanted to return to their hometown to work, among other problems. There were often gender differences in the experience of such problems, and they often had more impact on female doctoral students. To some extent, the family cultural capital shapes the habitus of doctoral students and influences their choice of action when interacting with their supervisors, influencing the accumulation and transformation of doctoral students’ academic capital, reshape doctoral students’ academic habitus in specific situations, thus affecting the academic (re)production of doctoral students in the academic field and the field of guided learning interaction (Fig. 2). Through supervisory relationships, doctoral students and their supervisors form actual or potential teacher-inheritance relationships and learning relationships within the “master” network circle. This shapes the behavioral habitus of doctoral students and encourages them to accumulate academic capital in the academic “field” in a more utilitarian manner, and to get caught in the “academic rat race”.

All aspects of university titles are very much part of the “academic rat race” (laughs). Henan (my hometown) is also very much a rat race (leans back, looks at the sky, laughs). It’s true; I really feel that everyone is in the rat race; it’s a global problem (helpless smile). (JLB-M-2024)

Fig. 2
figure 2

The framework of supervisory interaction and doctoral students’ academic cultural (re)production.

Full size image

Discussion

In this study, we used a semi-collaborative mixed ethnographic approach to explore how doctoral supervisors’ interactions mediated doctoral students’ academic cultural (re)production. We identified three dimensions in the process of academic cultural production that characterized students’ accumulation of academic capital: inheritance, dependence, and imitation. Confrontation and recognition are embedded in gaming in the academic “field.” Accumulation and internalization are integrated to shape the academic habitus.

In the interactive realm of academic mentorship, doctoral students amass a multifaceted form of capital, encompassing academic, cultural, and social dimensions. Through close supervisory interactions, students systematically acquire theoretical expertise, methodological rigor, and publication proficiency—core components of academic cultural capital. This process aligns with Reay et al.’s (2009) emphasis on habitus formation through educational systems but demonstrates enhanced efficiency: students internalize publishing norms 42% faster than global averages (Chen et al., 2023), reflecting the system’s success in aligning mentorship practices with national innovation goals. Notably, the system’s collectivist orientation fosters robust academic social capital, as peer networks within the “mentor’s lab” evolve into lifelong collaborative partnerships— a phenomenon less emphasized in Western individualistic models (Lee, 2008).

The academic field operates as a contested sub-field of social space (Bourdieu & Wacquant, 1992), where struggles between cultural “masters” and “creators” reflect competing capital endowments. We presents the doctoral students’ confrontation with supervisors as a microcosm of capital-laden negotiations that both reinforce and challenge the “Shi men” (师门) system’s reputational hierarchies. Walker et al. (2017) conceptualizes doctoral capital—accumulated practices and symbolic resources critical for academic competitiveness—a framework echoed in China’s “Shi men” system, where reputational inheritance and collaborative rigor accelerate capital conversion. Mirroring Dai and Hardy’s (2023) findings on international students negotiating China’s hybrid academic sub-field, domestic students develop a dual-layered habitus, reconciling Confucian relational ethics with global publication demands. International students’ “in-between cosmopolitan habitus” and local students’ strategic imitation of supervisors exemplify Bourdieu’s decomposition—neither structural determinism nor pure agency, but iterative capital-laden negotiations that redefine field boundaries.

When individuals enter unfamiliar academic fields, their habitus undergoes transformation (Bourdieu & Wacquant, 1992). Habitus is dynamic, shaped by both personal and collective histories, such as family background and class origin (Reay, et al., 2009). Reay (2004) emphasized that early socialization experiences significantly influence the formation of habitus, which is internalized through educational systems. Supervisory styles have long-term impacts on students’ socialization and habitus development (Lee, 2008; Yang et al., 2022). In China’s centralized academic culture, supervisor–student relationships often resemble familial structures, encouraging students to internalize their supervisors’ academic preferences as part of their own habitus (Weng, 2020). This quasi-kinship model reinforces hierarchical roles and influences how students interact, communicate, and adapt. Those with limited interpersonal experience may develop an “avoidant habitus,” marked by reluctance to seek help, which hinders the accumulation of cultural capital. Such patterns, though less common in Western systems, are also observed in other East Asian settings. This study thus enriches culturally sensitive interpretations of habitus and offers new insights into academic culture (re)production in doctoral education.

This study offers new insights into academic culture (re)production in doctoral education. To further situate these findings globally, it is critical to recognize that supervisory practices differ across countries. For example, the UK and Australia typically focus on individual independence and maintaining strict project schedules (Taylor & Wisker, 2023; Helfer & Drew, 2019). The doctoral education in the United States often centers around individualized academic socialization, and preparation for diverse careers through flexible program structures and strong integration into federally funded research projects (Cui & Harshman, 2023). By contrast, Chinese doctoral supervision emphasizes a hierarchical, collectivist model that prioritizes loyalty, long-term relationship-building, and integration into the mentor’s academic lineage. Compared with Japan and South Korea, where similar Confucian values shape doctoral supervision, the Chinese “Shi men” model further institutionalizes reputational inheritance (Kim et al., 2009; Dai & Elliot, 2023). These comparisons across different countries emphasize how local academic cultures influence the development of hybrid supervisory environments spaces shaped by the ongoing interaction between national traditions and international standards. This hybridization process not only affects students’ academic habitus but also reshapes institutional logics of doctoral cultivation worldwide. In addition, China is a country that pays special attention to social capital. China’s thousands of years of history are not only a history of rule by man, but also a history and evolution of the inheritance of relationship network (Huang & Li, 2019, p. 72). We conducted within the unique context of the “Shi men” (teacher-student community) in China’s higher education system, presenting a distinct national context compared to existing studies, thereby enriching current scholarship on doctoral advisor-student interaction.

Conclusions, implications, and limitations

The present findings increase the limited understanding of how supervisory interactions shape doctoral students’ academic cultural production. The findings have both methodological and theoretical implications regarding investigations of the association between supervisory interactions and doctoral students’ academic cultural production, locally and contextually. A semi-collaborative mixed ethnographic approach was used to examine how supervisors mediate doctoral students’ academic culture in their supervisory interactions. Various sourced data were collected over three years, including the focal participants’ interactions through instant messaging, interviews, and artefacts. Bourdieu’s theory of cultural reproduction was used as a theoretical lens to characterize three core academic cultural production dimensions (inheritance, dependence, and imitation) that characterized the accumulation of academic capital through confrontation and recognition in gaming in the academic “field” and accumulation and internalization in shaping the academic habitus.

Along with the findings above, a key policy recommendation is to formalize and structure supervisory relationships in doctoral education. This could involve implementing formal mentoring programs that pair experienced faculty with junior supervisors to enhance their mentoring skills and share best practices. Furthermore, developing clear guidelines and expectations for both supervisors and students regarding communication, feedback, resource allocation, and progress evaluation is crucial. Regular meetings and progress reviews should be mandated to ensure consistent support and guidance. It is essential to involve all stakeholders, such as supervisors, students, faculty, and administrators, in the development and implementation of these policies to ensure buy-in and effectiveness. Finally, a system for ongoing evaluation should be established to regularly assess the impact of these policies and adjust as needed based on feedback and evidence of their effectiveness. By implementing these measures, institutions and policymakers can create a more equitable, supportive, and productive environment for doctoral students, ultimately contributing to the advancement of knowledge and the development of a diverse and thriving academic community.

This study has several limitations. Primarily, the qualitative, ethnography approach, while insightful, does not establish causal relationships between supervisory interactions and doctoral students’ academic cultural production. Future research employing quantitative methods is needed to explore these causal links. Furthermore, this study’s context and cultural setting may limit the generalizability of its findings. Additional grounded theory or ethnography studies are required to investigate how diverse contextual and cultural factors, such as disciplinary norms, institutional values, and funding models, influence this dynamic. Expanding the research to different countries, institutions, and disciplines will provide a more comprehensive understanding of the complex interplay between supervisory relationships, student agency, and the broader academic environment in shaping doctoral students’ academic trajectories and contributions. These limitations highlight the need for further research to build a more robust theoretical framework for understanding the doctoral experience.