Thinking with certainty or with doubt: a Lacanian theorisation of discursive knowledge in teacher education

ABSTRACT The paper presents a theorisation of pedagogic knowledge formation, as a continuous attempt to understand the positions in discourse we occupy. The paper documents some participatory practitioner research by teacher educators centred on a course development initiative for student teachers of English, at an English university. Students researched their experiences of becoming a teacher within a course that was largely school-based, whilst their tutors researched their own involvement in the process (the main focus of this paper). Drawing on Lacanian theory, tutors are depicted as learning subjects having more or less certainty or doubt about the knowledge they possess. In attempting to understand this interplay of certainty and doubt, tutors arrive at stronger conceptualisations of learning. Through this approach, the paper provides a theoretically informed conception of professional knowledge, as involving a process of renewing ideas about learning, in meeting or resisting external demands.


Introduction
In the psychoanalytical model of Lacan (2008), knowledge can always be renewed. This is not to imply that current knowledge should be viewed with suspicion, as inadequate or inauthentic. Our knowledge is a reflection of the reality we perceive. Rather, in emphasising the need to revise our self-disclosure, Lacan is alluding to the inherited stabilities that guide our lives, sometimes throughout entire lives. Particular forms of discourse, sets of rules, models of thinking, inform us about the worlds we occupy (Badiou, 2009). Indeed much of the regularity we encounter (in the built environment, bodies of professional knowledge, educative discourses, aesthetic styles, modes of school governance), is revelatory of how previous generations perceived and modelled the process of living. Knowledge has accumulated all around us; as inhabitants of social space it is also built into us. Yet we can learn to discern the limitations of such stable arrangements, to question how they define and delimit our thoughts, to reshape the story or image of our own experiences. For such images are always in need of renewal. For Lacan (2008), 'Truth is always new, and for it to be true it has to be new ' (p. 17), because life as lived always exceeds the models that we try to place upon it. And the failures of these models produce desire to get things right. In Lacan's account of subjectivity, desire is the motive force propelling us irresistibly towards new ways of understanding our lives and the people we are. Our desire attracts us to the 'holes in discourse ' (p. 27), in order to re-think, and re-tell the story of our changing relations to the world. A learner would therefore be experiencing the world as coming into being, encountering elements of this world as part of herself, with this encounter yielding fresh motives for personal growth and renewal.
Our depiction of the Lacanian subject, persistently motivated through desire to renew her personal stories, differs from the later work of Foucault addressing the various modes in which the subject positions and defines herself. Here, the subject develops a technology of self by working reflexively across different discursive formations (Foucault, 1997). In contrast, the Lacanian subject always eludes embodiment in discourse, 'the failure of its representation is its positive condition' (Žižek, 1989, p. 175). Thus the Lacanian subject thrives because of this constitutive split between itself and discourse.
We present a theorisation of knowledge formation in reference to a project of practitioner research, in which the authors participated as university tutors on a one-year postgraduate programme of teacher education, in an English university. The research focuses on how the course tutors understand the professional knowledge they deploy with students and each other. According to our theorisation, knowledge is a product of the different positions we occupy in discourse, and our desire to understand these. The main innovation of the paper lies in our use of Lacanian discourse theory to analyse tutor knowledge. The key argument is that having knowledge should not be understood as simply involving the relationship between knowing 'subject' and knowledge-bearing 'objects'. Instead, knowledge is a product of our discursive attitude towards it. Thus, knowledge changes depending on whether we approach it with more or less certainty or doubt.
The authors were frustrated with recent policy driven changes in England, which have contributed to a narrowing of the knowledge base of university-based teacher education. Here university work is often viewed as either 'teaching' or 'research' with insufficient communication between the two (Ellis, Glackin, Heighes, & Norman, 2013). In contrast, we draw inspiration from the relative integration of practitioner-researcher functions in countries like the Netherlands and Finland (Burn & Mutton, 2015;Krokfors et al., 2011), whilst recognising that in the US, Australia and elsewhere, greater tension between these functions is apparent (Gerrard & Farrell, 2014;Paugh & Robinson, 2011;Sleeter, 2014;Tatto, 2015).
The empirical research for this paper comprises an ongoing practitioner research project, currently in its fourth year. Data from student teachers are presented elsewhere (Hanley & Brown, 2016). Theoretical accounts of teacher development drawing on a Lacanian perspective have also appeared previously (Brown, 2011;Brown, Atkinson, & England, 2006). In this paper, we provide some data indicative of the university tutors' perspective. In the next section, we develop the theoretical and policy context of our discussion. We outline circumstances specific to England but contemplating issues of wider concern. We then present a theorisation of new knowledge as comprising either continuity (Habermas) or disjuncture (Foucault) relative to previous understanding. We then put our Lacanian tools to work in redefining that discursive space. Two excerpts from tutor data enable us to elaborate the Lacanian influence in our approach. We then offer some concluding remarks, considering some of the wider implications of our theoretical and empirical findings.

Policy and theoretical context
The research took place in England where teacher education has experienced rapid policyled change. Government policy promotes models of teacher learning largely located in schools, with universities performing a supplementary role (Carter, 2015). Our aim is to contribute to a reconceptualisation of the university role in teacher development when time spent with students is particularly compressed. In England, recent iterations of work-based routes into teaching provide for as little as 30 days in university, during a one-year postgraduate course of 'training'. This stands in sharp contrast with teacher education programmes in mainland Europe subject to the Bologna process, where universitybased programmes may extend over four to five years prior to school placement (EC, 2013). Despite the unusual situation in England relative to much of Europe, concerns about reductive conceptions of teacher learning are internationally widespread (e.g. Baltodano, 2012;Evans, 2014;Riveros & Viczko, 2015;Wermke & Höstfält, 2014). A caricature of this learning might depict student teachers as learning 'subjects', learning how to teach through exposure to a particular set of knowledge-bearing 'objects' within a regulative apparatus such as the discourse of 'standards'.
In this paper, we consider what a less reductive approach to teacher learning might become. Taking our cue from Lacan, we suggest the challenge for the learner is to try to decipher for herself, what a learning situation really requires her to do. For example, one data excerpt (below) is from a planning meeting involving both authors, where teaching material for the next sessions with students was being discussed. Two of the tutors offer competing conceptions of what that material ought to accomplish: One tutor aims at fulfilling supposed student expectations of a learning experience, whilst the other wants those expectations to be challenged, even disoriented. Thus the tutors have different kinds of knowledge in mind, reflecting their contrasting desires for the educative encounter (see also Lacan, 2007). In the next section, we offer contrasting accounts of how this uncertainty about the purposes of learning can be mediated discursively. In our first perspective, influenced by Habermas, an ever more nuanced understanding of language closes the gap between learning 'subject' and 'object'. In our second, Foucault's work challenges the existence of any such linguistic understanding. These thinkers' formulations are then contrasted with a Lacanian theory of discourse.

Thinking about knowledge and teacher education: Habermas and Foucault
In the work of Habermas, we learn about our societies and ourselves through a fuller understanding of how we communicate. Habermas' work aims at developing a conceptualisation of an ideal state of communication, free of the distorting influences typical of modern, complex societies (Habermas, 1976(Habermas, , 1981(Habermas, , 1987. In seeking this ideal state, he argues, we should aim to rid our language of subjective bias. As Habermas puts it, ideal communication operates in the 'third-person ' (1987, p. 287). For example, if there is a dispute between two people, there is need of a third, impartial perspective. We can hammer out an impartial perspective by arguing about the validity of rival statements being made, to find out why communication is breaking down and how this should be rectified. Using our shared values and culture as the basis of agreement (p. 229), we can move closer to an ideal point of understanding, free from prejudice and coercion.
In a Foucauldian perspective, no such universal application of language is possible or desirable (1997). Universalist descriptions of the social world, such as Habermas', always exclude other interests, perspectives and ways of knowing (Foucault, 1972). Foucault's work employs an 'archaeological' or 'genealogical' method for recovering these excluded elements. For example, in some of his most notable work, Foucault traces the formation of the 'mad' subject to its origins in the psychiatric and carceral practices of the seventeenth and eighteenth centuries (1989,1997). More generally, our identities are predicated on reflexive involvement in contingent circumstances, not positive identification with a universally postulated referent. As Foucault puts it, we are 'subjects of what we are doing, thinking, saying ' (1997, p. 315). It is this reflexive activity, mediated discursively, that generates a sense of being in the world and not, as with Habermas, a linguistically sensitised relationship with the world itself.
The positions adopted by Habermas and Foucault centre on the question of how we locate and confer significance on what we do. With Foucault (1972) the researcher's task is to mark the separateness and singularity of discursive formations as they occur, in the absence of totality or synthesis; 'We must not imagine some unsaid thing, or an unthought, floating about the world, interlacing with all its forms and events' (p. 229). Habermas criticises Foucault for over-stating the importance of contingency. 'No place is left for any overarching meaning', he argues, 'Instead the space of history is seamlessly filled by the disordered flaring up and passing away of new formations of discourse ' (1987, p. 253). With Habermas the significance of our actions is secured by mutual participation in the lifeworld of others.
A recent related research study (Hanley, 2017;Hanley & Brown, 2016) involved groups of English literature and language graduates making the transition from graduate to teacher on a one-year school-based course of teacher education. In relocating discipline-specific knowledge from university to school, students were tasked with recalibrating earlier experiences as students and lovers of English in terms of classroom-based demands. Making the unfamiliar switch from university to school contexts, student teachers face a quandary. Should they hold on to the familiar 'academic' identity or redefine themselves in terms of the new professional environment, where the status of their graduate knowledge is less secure? In the Habermasian perspective outlined above, the emergent identity would be understood in terms of an overall continuity, with new understanding assimilating earlier knowledge. In the Foucauldian perspective, the emergent identity would be viewed as effecting a disjunction with earlier knowledge, the new understanding requiring new kinds of explanations and conceptualisations.
This quandary could also be referenced to Piaget's concepts, 'assimilation' and 'accommodation'. Lacan, to whom we now turn, did not subscribe to Piaget's conception of a learner moving through successive stages of ego-centred development. Lacan understood the actions of children as motivated by a socially expressed demand, so that both the child's and later the adult's identities are conditioned by how they perceive and relate to that demand (Lacan, 1977).
In the next section of the paper, we utilise Lacanian tools in scrutinising how we have been framing our practices as tutors and how these might be thought differently. We are interested in the Lacanian notion that a relationship with a given situation is cemented by a perceived demand, which the subject persistently tries, and fails to satisfy (Žižek, 1989). Thus Lacan's work stands in contrast with both Habermas' theory of communication and Foucault's tracing of discourses through their contingent historical circumstances and effects. As Verhaeghe (1995) puts it: The Lacanian theory has nothing to do with either of those two. His theory is even in radical opposition to communication theory as such. Indeed, he starts from the assumption that communication is always a failure: moreover, that it has to be a failure, and that's the reason why we keep on talking. If we understood each other, we would all remain silent. Luckily enough, we don't understand each other, so we have to speak to one another. The discourses stretch a number of lines along which this impossibility of communication can take place. This brings us to the difference from Foucault's theory. In his discourse theory, Michel Foucault works with the concrete material of the signifier, which puts the accent on the content of a discourse. Lacan, on the contrary, works beyond the content and places the accent on the formal relationships that each discourse draws through the act of speaking. This implies that the Lacanian discourse theory has to be understood primarily as a formal system, i.e. independent of any spoken word as such. (p. 81, emphasis in original) Next we examine the formal relationships made available through Lacan's conception of the four discourses, comprising the four positions of master, university, hysteric and analyst. We suggest that learners occupy these positions at different points in a process of learning, reflecting changing motivations in an evolving narrative of 'becoming a teacher'. First we address the more general idea of conceptualising learning as conditioned by formal, or structural (as well as symbolic) relationships.

Occupying the symbolised world
We run a teacher research session where students analyse the symbolic spaces which beginner teachers occupy. Inevitably, students focus on how these spaces are shaped in their own expectations, reflecting their histories as daughters and sons of parents with certain expectations, as graduates of a western system of education operating within certain priorities, as aspiring teachers wanting to achieve 'Qualified Teacher Status' and so on. One of our purposes with these discussions is to highlight to students, the functioning of structure. That is, to help students become sensitised to the impact on their thinking of positions they occupy. Another purpose is to help students to notice the interplay between structural awareness and perhaps more familiar ways of seeing the material, where what matters are specific phenomena and how they respond to them (e.g. what they think and how they feel about what they are doing). The Lacanian conception of the 'Other' provides a useful analytical tool for thinking about this process of self-definition in relation to a pre-existing symbolic framework. In Lacan's (2000) words, the 'Other' -… is not a world system, but a system of reference for our own experiencethis is how it is structured, and we can situate within it the various phenomenal manifestations with which we have to deal. We shall not understand a thing unless we take this structure seriously. (p. 74) Human interaction continuously reveals the functioning of this 'Other'. For example, I can say something with a specific meaning ('I'd like a glass of water') whilst also asserting a particular state of affairs as being true (I deserve a glass of water, you ought to give it to me, water is wonderful), thereby revealing the structural motivation of my statement. Such underlying affirmations (or denials) may reveal more about us than we intend.
For example, I might argue I am a free and independent being, whilst acting like a compliant conformist. And the situation could be reversed, when I flout all prohibitions whilst imagining myself as constricted by circumstances (Lacan, 2000). It might be, as Bailly (2009) suggests, that in the course of a single conversation, my imaginings are diverted in multiple different directions as I attempt to explain, define, assert and excuse myself. With each change, I am hoping my desire for the conversation will be realised in my occupancy of the new position.
In Lacan's (2007) seminar XVII, he outlines the notion of the 'Four Discourses' (master, university, hysteric, analyst), as a structure for understanding such shifting discursive alignments. Each discourse presents a way of understanding the structure of different social phenomena (governing, educating, protesting, revolutionising (Bracher, 1994, p. 107)). Lacan (2007) maps the schema onto concrete historical events, but we are primarily concerned with its general applicability to educative experience and the social and psychological motivations of individual agents in particular. Each discourse models a different way of understanding 'interrelationships between knowledge, truth, subjectivity and otherness, and how particular configurations among these elements are produced' (Clarke, 2012, p. 51). The invariant structural relationship in all four discourses means that conscious agency is driven by an underlying truth that the agent has suppressed but nonetheless reveals. (The discourse of the master is somewhat different, see Verhaeghe, 1995, p. 97). First, the discourses of the master and the universityare what we might describe as authoritarian discourses of mastery, insofar as both are dominated by master signifierswhether in the place of agency, as in the discourse of the master, or in the place of truth, as in the discourse of the university. (Clarke, 2012, p. 55) In this quote, 'place' relates to Lacan's (2007) visual representation of the four discourses in diagrams or 'mathemes', not presented in this paper. The 'agency' of the discourse of the master relates to the agent's power to dominate the field of discourse, a power which in truth is tautological, resting on the mere assertion of that discursive prerogative. Thus, the master suppresses self-knowledge of the essential vacuity of their identity as master. In the discourse of the university, the exercising of 'rational', 'neutral', 'scientific' knowledge is supposed to replace relations of mastery and dominance, but the 'place of truth' relates to science's suppressed political/hegemonic dimension. As an illustration of how an agent might operate within these discourses in an educative context, suppose I am debating an issue in public. I may decide the best way to win support for my position is being seen to assert myself with rival speakers, rather than having an informed view of the issue itself (master). Soon, the truth emerges that I do not have the necessary knowledge and cannot justify the authoritative stance I have adopted. Alternatively, I may adopt a neutral stance to the debate on the assumption that I am simply being reasonable and rational about the issues (university), whilst betraying my desire to dominate through the forms of knowledge (logic, arguments, reasons) deployed to support my position.
Conversely, in the discourses of the 'hysteric' and 'analyst', the relation between the individual agent and master signifiers or dominant ideas, is more ambivalent or open to question. The agent in the hysteric's discourse protests against the dominant discourses of the master and university, but in truth 'the subject is still underpinned by an unacknowledged and repressed other' (Clarke, 2012, p. 56). The hysteric agent merely substitutes one master (or stable form of knowledge) for another in searching for a secure and stable identification. Only in the analyst's discourse does the agent resist new forms of certaintyinstead speaking from a position of acknowledged dividedness leading to a more exploratory and perceptive discursive attitude. For example, at the public debate I may be aware of being undermined or flattered by another speaker so that my own contribution appears insignificant. I may opt to combat the speaker by exposing the hidden politics of their strategy with an alternative politics (hysteric). Or, I might speak to what is being suppressed, that which makes some things and not others in the debate sayable (analyst) (see Bracher, 1994).
Our aim is to critically examine the knowledge university tutors employed with students and each other, but to quote Verhaeghe (1995) again, 'Lacanian discourse theory has to be understood primarily as a formal system, i.e. independent of any spoken word as such' (p. 81). In the following discussion, we acknowledge this gap between Lacan's view of the conditions of discourse and our application of the Lacanian schema. Data groupings reflect student and tutor desires for educative encounters, but we recognise that 'The discourses, existing as a formal structure even before one speaks, are continually interchanging through the interrelationships between their disjunctions' (Verhaeghe, 1995, p. 98). For example, in asserting their non-identity with master signifiers, the hysterical subject restores and assumes the master's function (p. 98). Whilst our aim is to explore the educative possibilities of the analyst's discourse, we also note that something unique in the subject always exceeds the structure (Alcorn, 1994). We try to do justice to this ambiguity in what follows.
Tutors kept reflexive diaries throughout the research period, whilst other data included transcriptions of planning meetings, analysis of taught sessions, reflections on seminars, reviews of teaching materials, responses to interactions with students, etc. Two excerpts are included as illustrative of systematic empirical inquiry into the tutor function.
Working with an image of certainty: the discourses of the master and the university In psychoanalytic discourse an object of perception does not sit quietly in its perceptual frame. Rather it pushes at the limitations of the frame, creating a 'remainder' (Lacan, 2007) that challenges the social and psychological order. But we might resist our perceptions of a remainder, in favour of a conception of the world that shores up what we already believe, further endorsing our certainties. Here, 'certainty' relates to an identification with the idealised object of powerful discourses, where we repress our troubling knowledge of a gap (Cho, 2007) between the representation and the truth. And, knowledge thus forms a completely coherent but static, tautological (i.e. self-referential, selfenclosed) system, and it is precisely such a knowledge/system that, rejecting truth as dynamic, produces the a. (Bracher, 1994, p. 125) The 'a' here refers to the 'reject', something misrecognised in the discourse but essential to it (Lacan, 2007, p. 43). For example, the discourse of 'human capital' both draws attention to and covers over relations of dominance, in its depiction of learners as engines of economic efficiency (Davies & Hughes, 2009). In Lacan's schema the 'university' also fulfils a function of the master, by restricting a conception of learning to the activity of a learning subject produced within a closed system of knowledge (Boucher, 2006;Brown, Rowley, & Smith, 2014). Bureaucracy is an example of how 'Individuals are to act, think, and desire only in ways that function to enact, reproduce, or extend The System' (Bracher, 1994, p. 115).
As suggested above, in teaching sessions tutors assume an ambivalent relation with the discourse of mastery. Often, our purpose is to surrender authority to students with the aim of challenging received wisdom. Yet these challenges might merely be repositioning the students' expectations of the master. In the following excerpt from a tutor-planning meeting, there is uncertainty about how to proceed. At the beginning of the academic year, tutors re-appraised the forthcoming sessions by questioning their fundamental assumptions about the purposes of learning. The discussion centres on a course of initial teacher training where students were graduates in English language and literature aiming to become secondary teachers of English. Its apparent purpose is to assess current conceptions of English pedagogy.
Tutor 1: What affects how English plays out in the classroom? … Is it more about the people who do it or is it in the internal workings of the subject itself? What do we want them to understand about what they're doing … Does English stand alone or do students generate it differently every time they teach?
Tutor 2: I can sort of see what you're saying. They (students) need to see a connect, they have to see that it starts with them, they build it up from their own reading, thinking, whatever else … I want to pin this down to particular conceptual areas … authorial intention, reader response theory, grammar, literacy.
These reflect contrary impulses we often have in a process of planning. The tutors want to shift student thinking towards messier, more critical appraisals of how teacher identities are formed. What is at stake is not acquired knowledge but cultivating a tolerance for productive uncertainty (analyst). Tutors, however, also wish to be recognised as 'knowing subjects' (Fink, 1995, p. 132). That is to say, as experienced tutors we are aware that student learning inheres in an image the learner has of herself as being able to learn, supported by an image she has of particular tasks as appropriate to her inquiry (see also Lacan, 1977). Doubtless tutor identifications as 'one who knows' are influenced by hysteric insistence on the final authority of the master, and in debating how best to present these ideas to students, we anticipate being future addressees of the hysteric's discourse. Tutors will hope to nudge students towards more analytic frames of inquiry (e.g. not 'what am I learning?' but 'who am I becoming?' (Hanley & Brown, 2016, p. 355). Yet experience suggests even here, the desire for certainty continually re-asserts itself -'the signifier always fails to account for the truth' (Zupančič, 2006, p. 166).
Shortly after this excerpt was recorded, one student told us she was fed up with reflexive work in university sessions, wanting to know instead 'When are you going to tell us how to teach?' Such (hysteric) responses are common and reflect an image of learning as technical mastery, with the role of the learner reduced to passive assimilation. Our reading of such responses is that in seeking to appear competent, students invest in a view of knowledge that is stand-alone and self-sufficient, an infallible basis for classroom routines and procedures. Here the desire of the hysteric might be directed at being an 'outstanding teacher', which each portion of authoritative new knowledge promises, then ultimately fails to deliver. The second tutor's excerpt reflects this desire for an object to sit definitely and authoritatively within its perceptual frame (as quoted above, 'authorial intention, reader response theory, grammar, literacy'). However, as Lacan suggests, for a 'master' to appear as 'master' takes our collusion with its appearance of self-sufficiency, and an argument might be made for the presence of the master-slave dynamic, in the sense provided by Lacan's reconceptualisation of Hegel, where it is the slave's and not the master's function to know what the master wants (Lacan, 2007). Tutors and students might be interpreted as alternating the master-slave positions; the tutors are trying to anticipate the students' desire to identify with a learning process in particular ways, whereas the students maintain the master's identity by subscribing to their distribution of knowledge. The pair of tutor's comments mark something of this ambivalence, reflecting our desire for students to recognise their own (analytic) potential for renewing a perceptual frame, whilst seemingly mapping this inquiry onto already familiar ideas and ways of knowing.

What am I doing here? Discourses of the hysteric and analyst
The discourses of the 'master' and 'university' present contingent discursive formations as certainties. We suggested a desire for certainty can inhibit learning. We may be shielded from ambiguity and the repressed elements return in the form of dissatisfaction (Clarke, 2012). In contrast, we can forge a more productive relationship with doubt in the positions of 'hysteric' and particularly, 'analyst'. In psychoanalytic practice, the hysteric (male or female) works with an image of herself made more positive through her capacity to question (Lacan, 2007). Conversely, the analyst occupies the discursive position of the 'one supposed to know'. That is, she appears to possess knowledge the analysand desires and uses this appearance to enable the analysand to learn through questioning herself (2007, p. 43). In university sessions, tutors used a technique where student questions to tutors were answered exclusively with further questions, obliging students to examine what they really wanted to know and how better questioning might lead them there. Asking pertinent questions can be a mere substitute for other forms of mastery ('now ask me a higher-order question!'). However, tutors aimed at particular (analytic) discursive relations, where new ideas, perceptions and ideas could emerge around the function of the analyst, with students gradually taking over this function for themselves.
A tutor diary entry followed an observation of a student's teaching in school and subsequent discussion. This excerpt came towards the end of the year's teaching practice, when the student (Rachel) had also generated a significant amount of reflexive data, relating to her own development into a teacher. At this meeting, she was invited to review all her earlier data in sequence from the start of the academic year, alongside some excerpts from other students.
When she (Rachel) saw the excerpts laid end to end, she noticed something powerful. Right the way through the course, ethics and values mattered to her above all else. Somehow they were more tangible to her now they'd been tested. Now she said she'd never give them up.
I asked her what students 'really' think about becoming teachers. Are we the tutors, getting it right? What should we do differently?
On reflection, though I frequently ask questions like this, I'm not sure what they are supposed to mean. Am I looking for a solution? A debate? Am I simply seeking reassurance? Rachel looked unsure how to answer. Perhaps she was thinking 'will I be graded on my response?' Then she said the course paperwork is very limiting, because it only looks for certain things. It doesn't really reveal the students as individuals, as people. This was said very gently as though I might be offended to hear the course criticized (if it really was criticism). I guess for students the tutors are, first and foremost, figureheads of the institution.
In this excerpt, the tutor wanted to create an analytic, generative space. The tutor's intention is to enable Rachel to recognise changes that have already occurred ('she noticed something powerful'), and to identify with an image of herself as responding positively to change ('tangible by being tested'). The discursive positions shift when the tutor identifies herself in the role of questioner (hysteric), with Rachel positioned as the one who knows about how students learn ('are we getting it right?' 'Am I looking for a solution?').
Here Rachel is being addressed as the master, in the sense of discursive prerogative associated with giving 'feedback' about an educative process. Rachel may be additionally fulfilling the analyst's function of putting her addressee in the position of facing their own split subjectivity (Bracher, 1994) the tutor desires to know what the students really thinks but seems to disavow the question in which the desire is framed ('though I frequently ask questions like this, I'm not sure what they are supposed to mean'). It is Rachel rather than the tutor who creates the conceptual frame in which this question can be understood ('students as individuals, as people'), and the tutor who, perhaps falling back on received ideas about the motivations of students (university), seems not to really hear her answer ('I guess for students the tutors are … ').
Though analytic discourse is not unambiguously present in the data excerpt, we would suggest an avenue for approaching it might include encounters like this, where the tutor (and perhaps student) attempt to assume the analyst's function and much is revealed about what they were attempting from the uncertain results. It is possible to theorise a space where learners begin to recognise how they are positioning the addressee and how this reflects on their own desires for the educative encounter. In particular, we might recognise how the addressee is held accountable for our own sense of wanting knowledge and stable identity. As tutors, we found much of our effort was geared towards helping students to recognise this in their dealings with others and (less comfortably) exploring this possibility in ideas about the tutor role and purpose. The intention of this work was to encourage a conception of teacher knowledge as being less rigid, less absolute and judgemental, more responsive and perceptive, than it might otherwise be (Bracher, 1994).

Conclusion
In this paper, we put Lacanian concepts to work in our discussion of university tutor identities, where the formation of pedagogic knowledge involves both students and tutors 'adopting a new order of symbolic relations to the world' (Lacan, 2000, p. 78). In reference to Habermas and Foucault, we suggested that new symbolisations can be articulated in terms of continuity or change, where educative objects seem to fit within existing schema or demand alternative kinds of response. In discussion of the last data excerpt, we illustrated how such changes can play out in tutor-student interactions, when neither person is quite sure of the true nature of their collaborative purpose, and learning results from grappling with this uncertainty. Such dialogues can be both reassuring and unsettling, as educative objects continuously reshape themselves to our desires for the educative encounter, whilst ultimately proving elusive. As Lacan (1977) says, in one of his more flamboyant moods, I would now like to make clear, astonishing as the formula may seem to you, that its status of being, which is so elusive, so unsubstantial, is given to the unconscious by the procedure of its discoverer. (p. 33) And Lacan (2000) again, The subject does not have to find the object of his desire … He must on the contrary refind the object … Of course, he never does refind it, and this is precisely what the reality principle consists in. The subject never refinds … anything but another object that answers more or less satisfactorily to the needs in question. (p. 85, last emphasis added) In fixing our expectations on a particular external object (like 'good teacher'), we are committing ourselves to seeking it in particular ways. Yet the intended object and the inquiring perspective will exert pressure upon and unsettle one another, obliging us to reposition the motivational stories that tell us who we are and what it is that we really value.
In this paper, we presented an account of student and tutor activity, as a continuous attempt to define oneself differently through the unfolding of discourse. This attempt was referenced to the discursive functions of the master, university, hysteric and analyst, which differently focus the learner's desire on educative experiences and understandings. We suggested that a learner sometimes craves certainty, a stable sense of self and knowledge (master, university) above all else, and this impulse fuels her engagement with educative experiences. We also offered an alternative view, the key finding of this paper. As university tutors involved in teacher development, we recognise the importance of doubt in a process of learning (hysteric, analyst). That is, a successful learner can establish a productive relationship with doubt, by analysing her own sense of learning within limits and how these might be surpassed.

Disclosure statement
No potential conflict of interest was reported by the authors.