Teacher identities of graduate teaching assistants: how we (De)legitimise GTAs’ role identities

ABSTRACT

Graduate Teaching Assistants (GTAs) are doctoral researchers who are paid to teach. Building on work exploring identity construction in doctoral education and in the GTA context, this study uses role identity theory (McCall and Simmons 1978) as a theoretical framework to explore the factors influencing GTAs’ teacher identity development. Role identity theory emphasises the importance of social interaction and negotiation of identity with those in similar and opposing roles. Semi-structured interviews were conducted with nine GTAs from a variety of STEMM disciplines. From a role identity perspective, we found that the development of GTA teacher identities was strongly influenced by the behaviours and values of others, whilst maintaining the autonomy of the individual. We conclude with a set of recommendations for academic leadership and practice that consider the institutional cultures within which teaching and learning are situated to suggest how GTAs can be supported to develop their identities as teachers.

KEYWORDS:

• Teacher identity
• graduate teaching assistants
• role identity theory
• institutional cultures
• doctoral researchers

Introduction

The roles and remits of graduate teaching assistants (GTAs) have received increased attention over the last three decades. Particularly in the North American context, there have been attempts to professionalise the GTA role since the 1980s and 1990s (e.g. Nyquist, Abbott, and Wulff 1989; Wulff and Nyquist 1996), with GTAs occupying recognised positions simultaneously as postgraduate researchers and as teachers. In the UK, too, employment of GTAs has become more prevalent in recent years. As Park and Ramos (2002) explain, this is largely due to changes in how higher education is funded, with increasing numbers of students but decreasing state funding per capita. This has led many universities to adopt policies of employing greater numbers of part time, often hourly paid teaching staff, many of whom are GTAs (Standen 2018). While we acknowledge that the definition of a GTA varies from one institution to another, in our institutional context – and in this paper – a GTA is defined as a doctoral researcher who is paid to teach and support the learning of undergraduate and/or postgraduate taught students.

The casual nature of GTA work, along with the student-teacher-researcher identities held by GTAs, has led some to characterise GTAs as ‘the donkeys in the department’ (Park and Ramos 2002, 53) and as occupying a role which is ‘neither fish nor fowl’ (Park 2002, 61), never quite identifying as students or as teachers, leaving them in a state of liminality. Winstone and Moore (2017), however, interpret this situation differently, suggesting that GTAs are ‘sometimes fish and sometimes fowl’ (Winstone and Moore 2017, 499, original emphasis), and that they can adjust their identities according to the situation at hand. These two broad perspectives on GTA identity – the former depicting a rather negative situation and the latter focusing on the dynamic role of identity work in a GTA’s academic life – form the starting point for the study presented in this paper, which explores the factors influencing GTAs’ teacher identity development in a research-intensive institution focused on STEMM disciplines (Science, Technology, Engineering, Mathematics, Medicine).

Conceptualisations of identity

In broad terms, identity has been conceptualised as the process of gaining an understanding of who we are and how we relate to others in our environment (Beauchamp and Thomas 2009). Key here is that identity construction is a process, or a series of processes, rather than a static end product. (Gee 2000). From a social psychology perspective, it has long been recognised that identity construction is a bidirectional, negotiated, social process (Goffman 1956). According to this conceptualisation of identity, individuals construct identities on the basis of complex social interactions, and identities can be reinforced or undermined depending on the actions and reactions of others. We will explore briefly two perspectives on identity construction which emphasise socialisation, and which have been applied in doctoral researcher and GTA contexts.

Identity as socialisation

Particularly in the doctoral research context, where GTAs study and work, socialisation models have been used to understand the process of doctoral students becoming part of the academic community. Identity is seen as a complex process in which an individual incorporates the values and norms of the group (Bragg 1976) into their own identity. In the doctoral education context, Bieber and Worley (2006) found that there seems to be a process of socialisation, whereby PhD supervisors tend to prioritise their students’ research skills over teaching skills, but that ‘individuals are not internalising the values and attitudes their graduate school advisors presumably hold regarding the primacy of research’ (Bieber and Worley 2006, 1028). This suggests that identity construction is more complex than a simple observation and subsequent transfer of values from a more experienced academic to a new academic.

Similarly, criticism has been levelled at communities of practice (Wenger 1998) as a framework for exploring identity (Fuller 2007), where the focus is on emulation. In Lave and Wenger’s (1991) concept of legitimate peripheral participation, newcomers to a community participate in the practices of the group, initially on the periphery, observing existing members and constructing identities on the basis of these observations of practice. Some have cautioned against focusing on emulating the practices of existing group members, particularly in the doctoral research context (e.g. Austin and Wulff 2004), as this reduces the student’s – or the GTA’s – autonomy in the construction of their own identity.

Identity as socialisation: role identity

An alternative socialisation framework for exploring identity is role identity theory, which has been used to underpin research in both doctoral education (Jazvac-Martek 2009) and GTA contexts (Winstone and Moore 2017). Role identity theory was developed by American social psychologists George J. McCall and J. L. Simmons (McCall and Simmons 1978), whose initial work drew loosely on dramaturgy, considering how individuals perform and enact different roles. In line with the bidirectional and negotiated view of identity outlined above, role identity theory places emphasis on interaction with others (McCall 2003). An individual makes sense of their multiple roles either by learning from how others in opposing roles react, or by observing others enacting the same role – or a combination of the two (Burke and Stets 2009). From a role identity perspective, however, there is a focus on the individual’s identity alongside processes of socialisation within the group. As Stets and Burke (2000, 227) put it: ‘[A] role-based identity expresses not the uniformity of perceptions and behaviours that accompanies a group-based identity, but interconnected uniqueness.’ This focus on the development of an individual identity, whilst also maintaining the importance of social interaction and negotiation with those in similar and opposing roles, makes for a more nuanced conceptualisation of identity formation where multiple roles are being enacted. In more recent writings about role identity theory, McCall (2003) has elaborated on this framework by highlighting the potential tensions involved in enacting a new role: ‘Logically speaking, however, identifying with one social object entails disidentifying with other social objects that differ from that one’ (McCall 2003, 12, original emphasis). This is particularly pertinent in the GTA context, where several roles are performed at once. With Park’s (2002, 61) ‘neither fish nor fowl’ (neither academic nor student) characterisation of the situation in which GTAs find themselves, role identity theory may offer a way of understanding the tensions between the different roles enacted when simultaneously studying for a doctorate, conducting research, and teaching other students.

GTA teacher identities

As mentioned above, the last three decades have seen an increase in research on the experiences of and the roles played by GTAs. In the UK, this is set against a political and policy landscape in which teaching in higher education is receiving greater scrutiny than ever before. More specifically, with increased casualisation of teaching-focused academics, which includes a greater reliance on GTAs (Standen 2018), there has also been an increase in research on teacher identity development of GTAs over the last 15–20 years. To our knowledge, the studies that have been carried out on GTA teacher identities have been situated in specific disciplines or with particular groups of GTAs, such as international GTAs. Examples of these discipline-based and group-focused studies will be reviewed briefly in the remainder of this section in order to highlight factors that have been identified as important for GTA teacher identity development in various contexts.

In Physics, for example, Volkmann and Zgagacz (2004) examined their own practice, in which Volkmann was the course leader and Zgagacz was the GTA. The study explored Zgagacz’s construction of her identity as a physicist and as an emerging teacher of Physics. At the beginning of the study, and after observing Volkmann’s teaching, Zgagacz reported that she felt the students were frustrated by Volkmann’s approach, which used inquiry and constructivism to challenge students to think through the evidence for various potential answers to scientific questions, thus constructing their own knowledge. Zgagacz, on the other hand, placed importance on telling students correct answers. When she reflected on her values and beliefs about teaching and learning, it became clear that her conceptualisations of the teacher and student roles were based on how she had learned Physics as a student, which had been a didactic approach, focused on correct answers and achieving high grades. As the study progressed, Zgagacz’s beliefs about her role as a teacher began to change. While she did not agree completely with the rationale for a less didactic approach to Physics teaching, Zgagacz had started to see herself differently as a teacher, and subsequently enrolled on courses in learning and teaching in order to explore the ideas she had been exposed to, and her own conflicting feelings about herself as a teacher, further. According to Volkmann and Zgagacz (2004), the inquiry-based elements in the course design were essential not only to enable students to construct their own learning, but also to provide opportunities for the GTA to consider her role as a teacher. It was also important to create an environment in which the GTA was able to disagree with the course leader and voice the dilemmas she was facing, caused by conflicts between her existing teacher identity and the emerging notions of herself as a teacher.

Another reflective account that focused on one GTA’s experiences was contributed by Fairbrother (2012), this time in Public Health. In contrast to Zgagacz (Volkmann and Zgagacz 2004), Fairbrother began her GTA work with a student-centred view of teaching. As was the case with Zgagacz, Fairbrother attributes her beliefs about learning and teaching to her own previous experiences as a student as well as the fact that she had recently completed a Postgraduate Certificate in Education (PGCE). Fairbrother (2012) notes the well-documented pitfalls of the GTA role as a halfway identity between student and teacher (Muzaka 2009; Park and Ramos 2002), but she is also clear that her researcher and teacher identities are intertwined. A particularly important aspect in Fairbrother’s teacher identity development was the autonomy she was afforded in designing learning and teaching activities, rather than always working purely in an assistant role: For Fairbrother (2012), the unique position of GTAs, as researchers and emerging teachers, brings potential benefits for all involved, especially if the GTA role is conceptualised as a traineeship rather than a marginalised assistant role. This point was made by Park (2004) several years earlier in his review of the GTA role in North America. However, it seems this notion of GTAs as academic apprentices (Park 2004; Nyquist, Abbott, and Wulff 1989) is yet to be fully established in UK universities (Jordan and Howe 2018), perhaps reminding us of the importance of institutional culture (Garsten and Nyqvist 2013) in influencing and nurturing academic identity formation, especially in its formative stages.

A particularly important piece of research in relation to the present study was conducted by Jazvac-Martek (2009), who explored the experiences and identities of doctoral students using role identity theory (McCall and Simmons 1978) as a framework. As explained above, role identity theory postulates that people engage in a variety of roles and construct a variety of identities. This is particularly relevant in the doctoral research context, where an individual is likely to be in the role of student, researcher and teacher. Jazvac-Martek’s (2009) work was carried out with nine Education doctoral researchers at a university in Canada, focusing on their identities as doctoral candidates in general, rather than their identities as teachers specifically. Jazvac-Martek (2009) found instances where doctoral researchers’ role identities were not confirmed by others, causing a conflict between self-image and the image portrayed to others.

Winstone and Moore (2017) built on Jazvac-Martek’s (2009) work by exploring role identity theory in the context of GTAs’ teacher identities, this time with nine GTAs in Psychology. The researchers conducted two focus groups in order to explore the GTAs’ experiences as teachers. They found that teacher identities were unclear at the beginning of GTAs’ teaching experiences, but that teacher identities developed over time and with increased opportunities to ‘experiment with provisional selves’ (Winstone and Moore 2017, 497). Winstone and Moore’s (2017) work also echoes Fairbrother’s (2012) reflection on her GTA experience, which highlighted the importance of autonomy and freedom in the teacher identity construction of GTAs. In contrast to some earlier studies (e.g. Park 2002), Winstone and Moore’s (2017, 497) participants generally found the ‘identity malleability’ – the fluidity of the student-teacher roles – to be positive. However, the participants conceptualised their researcher and teacher identities as separate entities, and GTAs’ teacher identities can potentially be threatened if others, such as PhD supervisors, do not value or support the (teacher) role being enacted. The researchers identify key factors influencing GTA teacher identity development, including how others treat the GTA, and the emphasis placed on research versus teaching. Winstone and Moore (2017) conclude, in line with Jazvac-Martek (2009), that GTAs oscillate between different role identities, and that there is no discernible point in time when the student identity is discarded and the teacher identity is adopted.

As mentioned towards the beginning of this section, some studies have also focused on particular groups of GTAs rather than specific disciplines. Williams (2007) found that with international GTAs (ITAs), there is often a high degree of awareness of cultural differences between education systems, but that some ITAs are cautious about drawing comparisons between their home education system and the system in which they teach, especially if they did not attend higher education in their home country. Williams also found that ITAs often spoke more positively about the education system in which they were teaching, in this case in the USA, as compared with their home education system. Furthermore, she found that ITAs’ teacher identities became stronger with experience in the teaching role. Williams (2007) concludes that discussions about teacher identity ought to be included in training programmes in order to prevent GTAs from simply emulating the practices of their own previous teachers. However, in contrast, several other studies (e.g. Volkmann and Zgagacz 2004; Sandi-Urena and Gatlin 2013) found that previous educational experiences of GTAs were an important influence on the development of teacher identity.

Another important aspect of ITAs’ teacher identity development relates to role identity theory. As noted in Jazvac-Martek’s (2009) study, a particular role identity might be legitimised or delegitimised by the reactions of others. There is some evidence that ITAs may be even more susceptible to the reactions of others than home GTAs, mainly due to issues around teaching in a second language and working in an unfamiliar education system. Wang and Mantero (2018), for example, explored three Mathematicss ITAs’ experiences in a US university, where students’ feedback, often critical, about ITAs’ English proficiency, introduced an unhelpful power dynamic and complicated the process of teacher identity development. The proposed solution is often to create specialised training programmes for ITAs (e.g. Meadows et al. 2015; Dimitrov et al. 2014; Dawson et al. 2013), which can be useful, but it is important to avoid seeing ITAs in a deficit light, and to resist engaging in native speakerism (Holliday 2005), which perpetuates the notion that ‘native speakers’ are uniquely and best placed to use the English language in educational contexts.

Across the studies reviewed in this section, a variety of factors emerged as important for teacher identity development, including previous teaching and educational experiences, pedagogic training, the learning and teaching environment, interactions with peers (other GTAs) and academics, and the GTAs’ beliefs about learning and teaching, which may also be influenced by cultural backgrounds of individuals as well as the broader cultures within the institution. The context of each study was quite focused, exploring a particular discipline or, in the case of Williams (2007) and Wang and Mantero (2018), focusing on the experiences of international GTAs. The present study broadens the scope of previous research by exploring the teacher identities of Home and International GTAs in a range of STEMM disciplines.

Methodology

The study took place at a research-intensive STEMM-focused institution in the UK, where a high number of GTAs are engaged in a variety of teaching and learning activities, from laboratory demonstrating, facilitating tutorials, supervising student projects, to marking and providing formative feedback. From informal conversations with GTAs and from discussions in educational development sessions for postgraduate researchers, there was anecdotal evidence that GTAs had diverse experiences in relation to their teaching roles and their multiple identities as academics, researchers, teachers, and students. The study presented in this paper was an attempt to explore these experiences more systematically, focusing on the following research questions:

1. How do GTAs describe who they are as teachers?

2. What factors influence the development of GTAs’ teacher identities?

3. How can GTAs be supported to develop their identities as teachers?

Data were collected using semi-structured interviews with nine GTAs who had recently attended workshops as part of our institution’s GTA training programme. Participants were selected using purposeful sampling as the aim was to glean rich information rather than produce generalisable results (Patton 2002). More specifically, maximum variation sampling (Savin-Baden and Howell Major 2013) was used in an attempt to ensure that a range of experiences could be captured in the data. The notion of saturation was excluded in this study, as it was not considered possible to reach true saturation, as individual experiences and perceptions of a phenomenon, in this case identity development in GTAs, are potentially unlimited. Sampling was therefore conducted by ensuring the following inclusion criteria were represented across the participants:

• Different amounts of GTA experience, from those in their first year of doctoral study and GTA work to those in their final year of doctoral study and GTA work

• Different fee statuses – Home, EU and International, in order to explore potential differences in experiences related to cultural and linguistic backgrounds

• Different departments and academic disciplines

• Different genders

These inclusion criteria, and the maximum variation sampling applied, are markedly different to other studies reviewed earlier (e.g. Volkmann and Zgagacz 2004; Sandi-Urena and Gatlin 2013), where participants were drawn from a single discipline. It was made clear in the recruitment email that participants who met these criteria would be selected on a first come first served basis. Each interview explored the following broad themes:

• The participants’ GTA experience

• The participants’ values and beliefs about teaching and learning

• The participants’ identity development as teachers

• The participants’ future plans

• Other themes raised by the participants

Each interview lasted around one hour and was audio recorded for subsequent transcription. At the end of each interview, the participant was invited to select a pseudonym from a list of names provided. The use of pseudonyms conserves the anonymity of participants whilst also maintaining the personal essence of the narratives that emerge (Saunders, Kitzinger, and Kitzinger 2015). The list of pseudonyms consisted of names from a variety of cultural and linguistic backgrounds, and participants were at liberty to choose any name on the list. As Clark (2006) points out, this strategy could risk obscuring important information about the ethnic and cultural identities of the interviewees; however, we considered it to be important to allow any potential links between teacher identity and participants’ cultural, linguistic or ethnic background to emerge organically, if at all, rather than attempting to attach cultural information to each participant through the pseudonyms. Table 1 provides a brief profile of each participant.

Table 1. Participant Profiles.

Participant	Discipline	GTA Experience	Fee Status	Gender
Ava	Physics	3 years	Home	F
Camilla	Earth Science and Engineering	3 years	Home	F
Dan	Electrical and Electronic Engineering	3 years	International	M
Elijah	Computing	2 years	International	M
Felix	Medicine	1 year	Home	M
Kari	Computing	1 year	International	F
Phoebe	Materials	2 years	EU	F
Sean	Medicine	2 years	International	M
Vera	Life Sciences	3 years	International	F

The interviews were transcribed and then analysed using reflexive thematic analysis in an experiential orientation (Braun and Clarke 2006; Braun and Clarke 2022), focusing on the meanings, experiences, feelings and perceived realities of the participants in their roles as GTAs. The transcripts were coded, initially using open coding line by line. The codes were then reviewed to find similarities between codes across transcripts, and to identify overarching themes with shared units of meaning in relation to the research questions. As Braun and Clarke (2006, 82) point out, ‘the “keyness” of a theme is not necessarily dependent on quantifiable measures – but rather on whether it captures something important in relation to the overall research question.’ This means some themes are exemplified by more data than others. The initial themes were reviewed to check that ‘the themes “work” in relation to the data set’ (Braun and Clarke 2006, 91). Next, the underlying meaning of each theme was identified in order to apply names that accurately captured the essence of the data in each theme. The final themes were:

• Conceptualisations of the GTA role

• GTAs’ teacher and researcher identities

• Length of experience in the GTA role

• The impact of others’ behaviours on GTAs’ teacher identities

The study was reviewed and gained ethical approval via institutional education ethics review processes.

Findings and discussion

Examples from the data were selected to illustrate the themes and to ensure that each theme represented a significant part of the overall story. We begin with conceptualisations of the GTA role in teaching and supporting learning.

Conceptualisations of the GTA role

It emerged that GTAs’ conceptualisations of their work as teachers varies widely depending on the context and the type of class, such as lecture, tutorial or lab session. Vera captures this variation using a metaphor of a more service-oriented role of a flight attendant in some of her teaching encounters.

Vera: I've noticed that some classes I am kind of in the role of in a way being a flight attendant […] So I'm mostly walking up and down the aisles. People are asking me to fetch them things and I'm reminding everybody of the safety protocols. […] Now some classes I've taught […] I felt like I had many more discussions with the students.

The variation in Vera’s role, with a more passive ‘flight attendant’ and a more engaged facilitator of learning in small-group discussions, in some ways echoes earlier research which characterises the GTA role as ambiguous (Muzaka 2009; Lueddeke 1997; Linehan 1996), which, it might be argued, could hinder the development of teacher identities. All nine interviewees emphasised the supporting, facilitative and mediating nature of their role. Though this was done in slightly different ways by each individual, it is evident that the participants felt very strongly that their role is clearly situated as an intermediary between students and lecturers. This conceptualisation is well aligned with role identity theory, which emphasises the multiple roles enacted by an individual and the variety of ways in which roles are enacted, depending on different (teaching) contexts and others in opposing roles (Stets and Burke 2000). Camilla captures this dual role of supporting learners to learn, and supporting teachers to teach, as follows:

Camilla: I see my role is there to support both the lecturer and the students […] I think without any GTAs present in the session, like, it would just be really hard for the students to actually take in the material […] I guess my role is supporting the students in their learning and the teacher and the lecturer in their teaching.

The interviewees also made reference to the small age gap between students and GTAs as a unique feature of the GTA role. More important than age, however, is the similarity between GTAs’ recent – and current – experience as learners and the current experiences of their own learners:

Felix: So, now I can see myself as maybe someone a bit closer to the students in terms of where they've come from. I am a learner, you know, […] I'm an achievable step.

Felix’s description of his role is evocative of Vygotsky’s (1978) zone of proximal development. As the GTAs highlighted in this study, they are of a similar age to their students and have also had similar recent student experiences. As such, the GTAs as more capable peers, in Vygotsky’s terms, are near peers (Ryan 2014), who are slightly further along in their learning. Felix’s use of the phrase ‘an achievable step’ portrays the GTA role as an embodied, personified form of scaffolding (Bruner 1978), whereby the GTA is a visual, tangible reassurance that the learning task at hand is achievable.

A final point to make about the role relates to GTAs’ perceptions of how their role is viewed by others. As well as engaging in a variety of teaching contexts and tasks, most of the GTAs also commented on the varying expectations of teaching staff.

Dan: Some staff would probably see us as guys who do the marking […] There are some lecturers who are really interested in the teaching and how you explain concepts. […] so each course director has very different views of what the GTA should do.

From a role identity perspective, identities are constructed through negotiation and interaction with others (McCall 2003), and this was evidenced in our data by the different expectations placed on GTAs by course leads. The GTAs in this study typically work with several course leads on a variety of modules, which means the GTAs have to navigate different, potentially conflicting, expectations, which might create the ambiguous and problematic conditions set out by Muzaka (2009). However, Eick and Reed (2002) suggest that role conflict can be mitigated by providing space for reflection, so that the individual can develop a strong role identity, which is robust and resilient when faced with differing expectations from others. Furthermore, the role conflicts which arose in Volkmann and Zgagacz’s (2004) study were reframed as positive encounters between the course lead (Volkmann) and the GTA (Zgagacz); as useful starting points for conversations about learning and teaching. This seems to resonate with our data and leads us to question the extent to which GTAs are provided with opportunities to reflect on their identities and sense of selves as teachers, but also the extent to which teaching staff are aware of how their interactions with GTAs might facilitate or hinder the development of GTAs’ teacher identities.

GTAs’ teacher and researcher identities

Some interviewees conceptualised teaching as an adjunct part of their identity. Camilla, for example, referred to teaching as an additional role, alongside the main role of researcher.

Camilla: If I'm ever explaining to someone what my role is at university, I would always say, oh, yeah, I'm a researcher […] And then the teaching thing is, yeah, it tends to be, like, I almost see it as, like, a side job.

The literature provides some explanation for why some GTAs may view their teacher identity in this way: As Henkel (2005) explains, the discipline identity often takes precedence over other identities in the wider higher education institution, which might include identities related to teaching. Therefore, while links between the two roles may be recognised, there is a tendency to reflect on how the main (discipline) identity is enhanced by other role identities.

Several interviewees, however, saw deeper links between their researcher and teacher selves. Elijah, for example, reflects on how his involvement in research and teaching contributes to the overall mission, or ecosystem, of the university:

Elijah: I'm doing the research part and in my teaching, I’m also contributing to creating the next generation. So in that sense, I guess you could just contribute to one aspect of it, but I feel like the university has to have both these things, and then each feeds into the other. And so yeah, I think that's a contribution to the ecosystem in that sense.

Elijah’s concept of how he fits within the researcher and teacher roles aligns with Fairbrother (2012), who reflected on her intertwined identities as a researcher and a GTA in Public Health. Malcolm and Zukas (2009) characterise academic identities in similar ways to Elijah and Fairbrother (2012), with researcher and teacher identities inextricably linked. They suggest that the bounded nature of disciplines is restrictive and that we should focus instead on how individual academics enact their various roles and construct their identities (Malcolm and Zukas 2009). As mentioned earlier, the focus on the individual is important from a role identity perspective. As Stets and Burke (2000, 226) put it: ‘the basis of social identity is in the uniformity of perception and action among group members, while the basis of role identity resides in the differences in perceptions and actions that accompany a role as it relates to counterroles.’ This suggests that socialisation within the bounds of a discipline, alone, may impede the development of multiple role identities, such as teacher identities. On a practical level, GTAs do work within disciplines, and their primary reason for enrolling at the university is usually to carry out research for a doctoral degree. However, as pointed out by Williams (2007) above, perhaps pedagogic training should provide space for GTAs to reflect on their different identities more frequently. This might involve highlighting that GTAs are (becoming) part of a secondary discipline – the discipline of learning and teaching.

Length of experience in the GTA role

One of the most common responses, mentioned by all nine GTAs, was the impact of teaching experience on their sense of self as a teacher.

Felix: When I’m doing the GTA role, I feel, you know, I feel alive. I feel like I'm doing what I like doing. […] And I think there was a part of me that thought, I don't think I want to do University teaching in the future as a career. But I had such a positive experience the last term that I’m now really thinking about it.

Ava: Because now, like when you're a PhD in the first year, like, you're more, you're kind of above the undergrads, kind of just barely, but now, I'm also senior PhD, and I help lower PhDs […] like I feel qualified enough to teach another PhD now, whereas at the start of my PhD, I had no authority to do those things at all.

The comments from Felix and Ava capture the sentiment of the other GTAs’ reflections on their teaching experience. For Felix, the positive experience had such an impact that he has started to identify more with the role of a university teacher, which he had originally discounted as a career option. Ava’s self-efficacy (Bandura 1994) seems to have increased, where she speaks about feeling qualified to teach. In addition to gaining an increased affiliation with the role of teacher, Sean spoke about a shift in his identity from learner to teacher.

Sean: I'm always a majority learner, but now majority a teacher, actually.

It is perhaps unsurprising that gaining teaching experience should strengthen an individual’s teacher identity. In the context of doctoral researchers becoming lecturers, Van Lankveld et al. (2017, 329) use the phrases ‘expert becomes novice’ and ‘loss of expertise’ to explain the feeling that researchers experience when beginning as a teacher, where they have been accustomed to being an expert in their discipline and then suddenly become novices in the field of education. Particularly in the GTA context, the importance of gaining experience and having opportunities to teach was also noted by Fairbrother (2012). In Sean’s case, he has started to see himself more as a teacher and less as a learner, though he retains facets of himself as a learner. This aligns with McCall’s (2003) notion in role identity theory of identifying with a new role while identifying less, or ‘disidentifying’ (McCall 2003, 12, original emphasis), with another role. Winstone and Moore (2017, 499) found that ‘GTAs engaged in a process of enactment to “try on” the professional identity and experience what it may be like.’ It seems that Sean liked the new (teacher) identity he tried on.

The impact of others’ behaviours on GTAs’ teacher identities

Of the nine GTAs in this study, six grew up outside the UK. Some of the GTAs reflected on cultural differences they have observed, and on how their cultural background might affect their identity development. Kari, for example, reflected on her general demeanour, which she says was a marker of difference when she came to the UK.

Kari: I get told I'm enthusiastic and bubbly all the time here. I studied abroad here. […] And I just remember having teachers from the UK, from Eastern Europe, from all over the place and I could never tell what they were thinking.

Until coming to the UK, Kari had not thought of her ‘enthusiasm’ and ‘bubbliness’ as a particularly special trait, but comments from students and staff highlight, for her, that her personality might have a cultural dimension. In this case, there was no negative impact on Kari, as the comments were received in a positive light. However, this highlights the potential for aspects of a GTA’s cultural background, and any resulting comments about this from others, to affect how GTAs perceive themselves. As Jazvac-Martek (2009, 255) puts it:

Role identities form as a person categorises, classifies or associates oneself in relation to a social grouping, with knowledge of role identities derived from social interactions with others in combination with the internalised version of a particular identity.

The comments about culture in Kari’s example have the potential to cause identity conflict (Leong and Ward 2000) if the internalised identity, in Jazvac-Martek’s (2009) terms, cannot be reconciled with the identity attributed by others in the social group. It is clearly not possible to prevent all potentially damaging comments from being made to GTAs. Even a seemingly positive comment, such as the one in Kari’s example above, might be received negatively by a different person or in a different situation. However, as discussed above, the solution is not necessarily to design training programmes exclusively for international GTAs, but rather to embed conversations about intercultural competence (Deardorff 2006) in programmes of study and educational development provision.

Another key factor mentioned by all nine GTAs was the reactions of students, other GTAs, and course leads. Ava sums up the impact student reactions have had on her developing teacher identity.

Ava: I think I've grown into it a bit more. I think maybe I do have it in my identity to be a teacher now, rather than just a learner, I'm not sure, but I think the response from the students has changed as I've gotten through the PhD.

Ava’s comment is similar to Sean’s reflections on shifting from a majority learner to a majority teacher identity, but she focuses on how her students’ reactions to her have changed since the beginning of her PhD. The interviewees also spoke about how the reactions of course leads, and the fair payment for GTA work, can have an empowering effect on GTAs.

Camilla: It's good to know, like, you know, that the role is valued enough that the teachers actually have expectations of you because you are also being paid to do this.

From a role identity perspective, the reactions of others have a vital function in legitimising or delegitimising the identity enacted by the individual. In the GTA context, Winstone and Moore (2017, 495) explain:

if a GTA is enacting the ‘teacher’ role identity, behaviour from students that respects this status has the potential to legitimise this role. However, if others’ behaviour leads the GTA to feel that they are not a ‘real’ teacher, the construction of this role identity will be impeded.

The same can be said for the reactions of other ‘actors’, such as course leads and professors. This aligns with the long-held view of identity explored above, where Goffman (1956) characterised identity construction as a bidirectional process in negotiation with others. As Camilla illustrates, the positive interactions with and high expectations of course leads were experienced as empowering. This is, fortunately, in stark contrast to the well-documented notion of the GTA role as undervalued and ambiguous (Muzaka 2009; Park 2002; Park and Ramos 2002).

A related but less positive factor refers to the status of teaching in higher education, and how others’ perceptions of teaching impact GTAs’ teacher identities. This was raised by four of the GTAs in this study. For example:

Ava: I think I'm still trying to get out of the conditioning that lots of people in academia have, where like research should be the most important thing and teaching is the secondary thing. […] I’ve actually been told by one of my supervisors to cut some of the teaching from my CV.

Phoebe: My supervisor knows that I do it [teaching]. She's okay with it. She doesn't want it to take too much time off my project.

Several studies have shown similar findings regarding the status of teaching alongside research. Dugas et al. (2020) found that academics were more likely to express a preferred identity as researcher, as opposed to teacher. Similarly, Henkel (2005) seems to foreground the researcher identity in her conceptualisation of academic identity. This is shown in the following exchange with Vera, highlighting the perception of teaching as a lower status endeavour compared with research. Vera began by asking the interviewer to confirm that her comments would not be shared with her doctoral supervisor, and then she made the following comment:

Vera: I was gonna say that I'm really interested in developing this identity more because I'm finding that I'm much better at the teaching side than the researching side and it gives me more personal satisfaction. […] I don't want my supervisor to lose faith in me. I don't want him to start to see me as somebody who's not worth investing his time and effort in for the rest of the PhD. […] So for my supervisor’s ears, I'm still interested in research, but this is also to help him stay interested in me, too.

It is particularly troubling that Vera was concerned about her supervisor discovering the extent of her interest in teaching. Whilst this was an isolated comment in the context of the present study, other research has produced similar findings. For example, Winstone and Moore (2017) found that GTAs often experienced conflict between their researcher and teacher identities because of less favourable views on teaching from their doctoral supervisors. Supervisors are in positions of authority and are potentially figures of inspiration for PhD students and GTAs. Therefore, GTAs exposed to such a negative view of teaching may feel impeded in the development of their teacher identities.

Concluding remarks and recommendations

As this was a small-scale study conducted in a specific institutional context, it is not possible to make generalisations about the relevance of the findings for other contexts. Nonetheless, we propose the following recommendations to help stimulate discussions around institutional cultures in relation to teaching and learning, and to support those working with GTAs.

First, in order to strengthen GTAs’ ongoing processes of negotiating teacher identities, designers of academic development courses for GTAs might consider including multiple and ongoing opportunities for GTAs to reflect on how they are enacting the role of ‘teacher’ and how their beliefs and values about teaching might be changing.

Second, to explore the different ways that staff work with GTAs in specific institutional contexts, course leads and lecturers who work with GTAs should be involved in the training conversations about GTAs. On a practical level, it may be problematic to require staff to attend a workshop on this topic, but a set of guidance should be created outlining aspects of good practice when working with GTAs. Similarly, the GTA role needs to be clarified to students, so that expectations are clear about how the role of GTA is enacted, both at institutional level and within specific departments and disciplines. Where possible, GTAs should be listed as visible, formal members of the teaching team, which includes being listed in module handbooks, for example. This may not always be possible, particularly where GTAs are appointed after module handbooks have been created, but teaching teams should be encouraged to plan ahead so that GTAs are included as ‘junior colleagues’ from the outset, helping to legitimise the teacher role identity.

Third, to address perceived imbalances in the status of teaching alongside research, and to nurture the academic identities of GTAs, training for doctoral supervisors should incorporate a holistic exploration of what it means to be a PhD student, acknowledging the unique multidimensionality and value of a role that incorporates GTA, educator and academic, and not only trainee researcher. This would open a narrative around doctoral supervision as a more rounded role, which is about mentoring the whole person. It would also communicate to new supervisors that teaching is an important and highly valued part of the academic role, including in the context of doctoral research and the work of GTAs, some of whom are aspiring academics and higher education teachers.

These recommendations acknowledge the value of exploring the nuances of the institutional and disciplinary cultural contexts within which the work of GTAs and teaching more broadly is situated, in order to examine some of the general and localised values, expectations, tensions and relationships that influence the construction of teacher identities.

Disclosure statement

No potential conflict of interest was reported by the author(s).