Disintegrating Qualitative Research

This paper explores a tension in qualitative psychology between, on the one hand, a deconstructionist framework in which the human subject is understood as positioned in and through competing discourses and, on the other, a humanistic framework in which the integrity of the subject is taken to be both a starting- and end-point of analysis. This paper offers a critique of the tendency for qualitative research to seek to produce integrated `narratives' of experience and argues for the importance of maintaining the vision of a subject in fragments. It does so by taking up the notion of there being `things that can't be said' and suggesting that this refers to two distinct issues: the multiplicity of possible accounts of experience and the way language systematically excludes some `abjected' material. It finishes with an illustrative analysis of an interview text.


Narrative and Emancipation
Much of the appeal of qualitative research to students and academics alike is that it seems to offer a more holistic understanding of the human subject, it seems to oppose the atomizing tendency to be found particularly in psychology, with its reductionist vision of how to explain human behaviour. The impulse amongst qualitative psychologists is to take a kind of moral high ground in which what is left out of conventional psychology is rescued as in fact most significant-the 'subjectivity' of the subject, the meaning-making activity through which people forge their lives, their narrativizing core. There is a great deal to be said in favour of this: much of psychology is indeed reductionist, mechanical, fissiparous. The brain drives the mind, so let us see which bits of the brain light up when the subject is at work on some task; anxiety floods through us, so let us count the words that disappear from our lexicon as a consequence; people run towards trouble, so let us measure their steps. Qualitative psychology enters here as part of a turn in the so-called 'human sciences' towards the rolling-up of experience into narrative form: without discounting the reality of events, the key research question becomes not what happens to a person, but how this is accounted for, how it is put into a frame that makes sense. This humanizes reality, placing the emphasis on the agentic, meaning-making activity of the research (and human) subject.
Over time, the narrative turn in the social sciences has become increasingly important, with much qualitative research sharing its interest in reproducing, or re-presenting, the personal versions of experience that subjects describe. This usually means taking some kind of text that has originated with a person-perhaps in an interview or a diary-and subjecting it to an analytic process that breaks the text down in order to rebuild it in a more convincing way and make of it a more coherent narrative. A person tells a story, but on the whole it is not very well told: it has too many twists and turns, too many characters that contradict each other, too many gaps. The researcher examines this and rewrites the text, presenting it as a set of themes that constitute a whole. However fragmentary these themes might be, the research task is to make some sense of them, to tell, that is, a better story. This story can have various tendrils and connections; indeed, it is a regular and important function of qualitative research to uncover broader contexts which give meaning to the story, for example by reference to societal discourses, or maybe even to the Freudian unconscious (Frosh, Phoenix, & Pattman, 2003;Hollway & Jefferson, 2000). But a story it is nevertheless, something with shape and direction: for instance, in Labov's familiar framework, it is structured into an abstract, an orienting passage, a complicating action, an evaluation and a resolution (Labov & Fanshell, 1977). In the course of this, the agency of the participating subject is restored: he or she is made into a speaker, with a point of view, someone positioned in discourse but nevertheless there, speaking to us from the page. We can identify the person through her or his speech; that is what a subject is. Recovering this subjecthood, this special nature of the individual, is a moral task, after all.
What can be wrong with this? On the face of it, it precisely coincides with an ethical and political act of resistance to the totalizing tendencies of globalization and the continually growing cultural hegemony of Western capitalism. It offers a 'voice', as people usually say, to marginalized individuals and groups whose views and experiences would otherwise be discarded-and, indeed, much qualitative work has had this goal and has consequently focused on groups usually pushed out of view through racism, sexism or other modes of oppressive practice (e.g. Mana, 1995). In many ways, the focus on restoring agency has an excellent philosophical and political grounding in Habermas's (1968Habermas's ( /1987 notion of emancipatory practice, which, whilst embedded in a discussion of psychoanalysis, has resonance for all work that purports to rescue the truth of the subject through attending to and repairing broken narratives. Indeed, it might be suggested that the coalescing of hermeneutic approaches in psychoanalysis and narrative approaches in qualitative research around the notion of emancipation is a significant theme in contemporary social science (Frosh, 2006). Habermas presents a powerful argument to differentiate psychoanalysis from the empirical sciences on the grounds of their differing relationship to a subject-object divide. Empirical (one might say here, quantitative) methods 'aim at disclosing and comprehending reality under the transcendental viewpoint of possible technical control' (Habermas, 1968(Habermas, /1987; psychoanalysis, on the other hand, is premised on the promotion of a meaning-filled link between the analyst and the one 'studied', the object who in being communicated with becomes a subject. More strongly still, psychopathological states are characterized by a kind of alienation in which the subject is separated from her or his own subjecthood: that is, experiencing oneself as an object, split from one's meanings (wishes, desires, etc., in the Freudian scheme) and thus showing symptoms, is precisely what psychoanalysis sets out to remedy. 'Split-off symbols and defended-against motives unfold their force over the heads of subjects,' writes Habermas, 'compelling substitute-gratifications and symbolizations' (p. 255). Psychoanalytic understanding is, therefore, a process not of seeking mechanical causes, but rather of restoring an identity between the subject and her or his own subjecthood. Habermas proposes that: 'the experience of reflection induced by enlightenment is precisely the act through which the subject frees itself from a state in which it had become an object for itself. This specific activity must be accomplished by the subject itself ' (pp. 247-248). The act of interpretation aims to provide for the analysand the opportunity to seize hold of lost or hidden meanings and re-own them; it is consequently a causal process, where cause is located in the reconstitution of transformative meanings and not in the identification of specific psychic facts.
Having been offered such a heavyweight banner under which to march towards an emancipatory approach to the reconstitution of personal meanings, it seems carping to suggest that there might be something wrong with the idea that rebuilding narrative coherence is an estimable goal for those who find themselves on the margins of hegemonic discourses. One can argue that just as individuals benefit in the psychotherapeutic domain from being able to speak their stories and have them reflected back in a way that enables them to be owned, so in the political domain it is precisely through the coherent articulation of subjugated narratives that oppressed groups become empowered. The histories of feminism, gay and lesbian rights and black consciousness are clear examples here. Nevertheless, there is something about this turn to narrative that is disconcertingly familiar from a long series of attempts to redefine identities, attempts that very often result in stronger versions of the same. The margins might contest the centre in this way, for sure, but how much changes? How much of what goes on here is a tactical attempt to give confidence to people who might otherwise lose hope, by telling them that their experiences actually make sense? I take as my textual source here a brief critique of the trend towards 'narrativism' in psychoanalysis (as in other social sciences) from Jean Laplanche (2003), who argues that making a coherent narrative can be seen as a defensive process. Summarizing, Laplanche makes a point that can be taken as a general comment on the relationship between psychoanalytic therapy and psychoanalytic understanding but which also has broader implications for the integrating tendency of much qualitative work: The fact that we are confronted with a possibly 'normal' and in any case inevitable defence, that the narration must be correlated with the therapeutic aspect of the treatment, in no way changes the metapsychological understanding that sees in it the guarantee and seal of repression. That is to say, that the properly 'analytic' vector, that of de-translation and the questioning of narrative structures and the ideas connected to them, remains opposed in every treatment to the reconstructive, synthesizing narrative vector. (p. 29) Limits to making sense, to making connections, have to be set. The point here is that it may be consoling, therapeutic even, to have sense made of one's mystifying miseries, one's uncertainties and partial understandings. It can indeed be empowering: we are made into agents; we are subjects with something to talk about. We do not even have to be completely sold on the idea of narrative coherence to accept that it might have an important part to play in an ebb-and-flow process of living in the world: understanding and deconstructing might take place sequentially, so that meanings come into the frame, allow identity to be stabilized and defended, and then once this has been absorbed new challenges occur, things get taken apart and have to be struggled with once more. But even the advocacy of cyclical coherence misses an important point, one that has been central to modernist as well as post-structuralist and postmodern sensitivities and concerns: this is the idea that the human subject is never a whole, is always riven with partial drives, social discourses that frame available modes of experience, ways of being that are contradictory and reflect the shifting allegiances of power as they play across the body and the mind. Freud clearly knew this and articulated it in what Laplanche (1997) calls psychoanalysis's 'Copernican revolution', whereby the subject is no longer capable of being taken as the source and repository of psychic life. What postmodernism adds here is the notion that this 'decentring' can never be reversed through somehow returning agency to the subject, because the riven subject cannot be seen as a whole-there is, simply, no external point from which the true story of the subject can be told. Laplanche's (2003) formulation is pretty exact here: the 'properly analytic vector' is 'that of de-translation and the questioning of narrative structures and the ideas connected to them ' (p. 29). In other words, however much, for therapeutic and strategic reasons, one might want to make a coherent narrative out of a subject's chaotic account, don't believe a word they say.
To summarize so far: there is a need to hold on to this dialectic, this movement between fragmentation and integration, the part and the whole, without desperately seeking resolution. Qualitative research lives in the tension between, on the one hand, a deconstructionist framework in which the human subject is understood as positioned in and through competing discourses and, on the other, a humanistic framework in which the integrity of the subject is taken to be both a starting-and end-point of analysis. In the first approach, which has the advantage of being anti-humanist in the sense of not positing a foundational 'essence', and so is less prone to ideological compromises, there is the recurrent problem of agency and resistance: how can one ever become other than what one is positioned as being? If one is 'positioned' by discourse, how does change ever occur? Criticism of some hard-line post-structuralist accounts is relevant here: for example, Benjamin's (1998) complaint that Lacanian psychoanalysis fails to engage with the 'authorship' of the subject and hence to articulate possibilities for people to grasp their futures differently. In the second approach, visible, for example, in some narrative psychology, it is common to chide psychological research for its tendency to fragment the subject and to present qualitative approaches as a way to return to a holistic understanding of the person. This is presented as an advance in terms of both epistemology and morality: something special is pointed to, often silently, that works within the subject to act as an irreducible point of resistance, a place of safety or genuineness, which can never be explained away. However, the romantic view that arises from this is not only at odds with some of the more critical perspectives that bear on the emergence of qualitative psychology, but, if one follows Laplanche's argument, it also serves psychosocial functions best understood as defensive. For this reason, it is important to mount a critique of the tendency to produce integrated 'narratives' of experience and to argue for the importance of maintaining the vision of a subject in fragments. In what follows, however, I want to do this without adopting the extreme position in which fragmentation is celebrated: the history of psychosis is as good a place as any to turn to for evidence that being broken into bits is not a subject position to be advocated (Frosh, 1991). Instead, by employing a sleight of hand in which the opposition to the integrating tendency is presented not as fragmentation but as multiplicity, I will suggest that the aspirations of qualitative work are served best by adopting the notion of an over-determined subject, of a way of being that is 'excessive', too much.

On Not Saying What Can't Be Said
Once in a while, an academic paper seems to strike some kind of nerve amongst readers and to be taken up, albeit gently and often by friends of the author, as if it reflects a general sense of something going on. One such paper of mine first appeared in 2001 in an obscure journal, the International Journal FROSH: DISINTEGRATING QUALITATIVE RESEARCH of Critical Psychology, and was reprinted in the book After Words a year later (Frosh, 2002). This seems to have gained much of its audience by virtue of its title: 'Things That Can't Be Said: Psychoanalysis and the Limits of Language'. In it the discursive turn in psychology was praised for the advances it has brought in restoring meaning to psychological investigation, but then criticized on the grounds that 'there exists a large variety of different psychological experiences of considerable emotional force which lie outside narrative-even outside of what can be spoken' (p. 135). The paper goes on to suggest 'that these experiences can be central elements in people's lives, key components of psychological functioning, and that they have a specific connection with trauma and the processing of traumatically troubling events' (p. 135). Following a development of these themes through investigations of gender, psychotherapy and Holocaust testimony, the concluding rhetoric runs: Things that can't be said are at the core of our experience, we are what they are. Once they are symbolized, they no longer traumatize as much, it is true, and there is a kind of responsibility to do something with them because of this. But this does not make it easy, and there are times when putting things into words is a deep and painful loss. (p. 149) It is pleasing, of course, that this work has to some degree been taken up by others, and that it is part of a broad attempt, by discursive psychologists (e.g. Billig, 1999) as well as by critical psychologists (Parker, 2005a), to look at the way language often references its own 'aporias', its own gaps into which aspects of experience fall. I want to return to these points later. But what is also striking about some of the responses to this paper is that, despite its explicit disavowal of any spiritual goals-any paean to some essential, irreducible element of human subjectivity-this is how the idea is often used. This reading of the extra-discursive celebrates the fact that some things 'can't be said' and takes it as meaning that these things have to be left alone, they have to be seen as that which cannot be reduced, quantified, qualitivized, narrativized or whatever. They can only be pointed towards, left not-known because they are so mystically precious. This seems partly to be a Zeitgeist issue: the search for meaning is always with us, and in its current form the marginal position is to look away from, on the one hand, materialism and, on the other, organized and oppressive religious fundamentalism, and towards something personally whole, spiritually rewarding and philosophically sublime. This kind of marginality again makes moral claims: whatever you might do to the human subject, however you might dis-integrate her or him, something precious remains untouched. Perhaps we are back to the soul here, though few people will admit as much.
The line I actually take on 'things that can't be said' is rather different from this. It takes seriously the experience described by many people, perhaps by everyone, that somehow language never quite encompasses reality, that the way we talk about things might indeed have all the performative, effective and constructive functions described in the literature-that we might indeed be positioned by language-and yet the feeling remains that whenever we try to say something completely, the saying of it misses the point. That we keep on trying is a testament both to the importance of putting things into words and to the defensive elements of this narrative: without it, we spiral into nothingness. Lacanians might say of this that the Symbolic defends us against the Real, which by no means implies that we should get rid of the Symbolic, only that it is always a mode of covering-over, however much the world and its subjects depend upon it. It also references Kristeva's notions of the semiotic as an 'underside' of language and particularly of abjection (Kristeva, 1982): that which is excluded from the Symbolic as the disgusting, death-invoking thing that exists prior to language and threatens to plunge the subject back into non-existence. In effect, language both expresses abjection (semiotically) and (symbolically) defends against it. But if something cannot be captured in narrative, then what can be done with it-and how are we to know it at all? My view is that we are not in the presence of a kind of mute pointing-'there is the Real, and there, and there …' Nor are we in the realm of the transcendent, absorbed into a Oneness that cannot be spoken of because it is either too 'Other' to have form, or too immanent to allow us to see it. The reason that some things cannot be said is not that they are mystical and the language in which to express them is absent, but rather that language itself produces gaps and difference, that as we speak and therefore inhabit the Symbolic, we engage in a process of exclusion. This exclusion has two related forms: first, it reflects the partial way in which narrative proceeds, always striving towards a linear account of what is multifarious, uncertain and complex, always seeking to capture something that eludes it; and, second, it actively rejects what cannot be borne, what stands outside language as its radical other.
The first of these two forms of exclusion might be usefully indexed through the idea of multiplicity. There are, simply, too many ways of speaking about things, and to do them justice one would have to use all these different ways, all at once. Something is always left out precisely because something more can be said, and each new way of saying will add a new dimension, often contradicting what has gone before. This might sound overly jubilant about the prospects for language, but it is only another way of saying that our experiences are made up of overlapping bits and pieces, Lacan's (1954Lacan's ( -1955 famous 'bric-à-brac', and that an exhaustive comment on all the potentialities of experience in all possible ways can never be achieved. In this framework, it is not so much that there are things that absolutely cannot be said, but rather that trying to say them all at once is impossible. Why should this be so? Not only because there are so many ways of articulating something, all of them different, none of them absolutely precise; but also because, given the reflexive nature of human consciousness, as one pronounces upon experience, so the experience itself must change. Otherwise, what would be the point of psychoanalysis or any other form of verbal psychotherapy? The speaking of the thing acts as a wager, a point at which something is risked into existence; the very act of doing that, as a mode of Hegelian externalization, means that it is worked on in the presence of the Other-whether that be the 'actual', interlocutive other of the analyst or the imagined other of the speaker her-or himself. This is, as it happens, exactly the paradigm of relational thinking: as Butler (2004) puts it, 'To ask for recognition, or to offer it, is precisely not to ask for recognition for what one already is. It is to solicit a becoming, to instigate a transformation, to petition the future always in relation to the Other' (p. 25). We do not recognize the other and get recognized ourselves through a linguistic process of exchange in which nothing changes; rather, we are called into being through the process of being recognized, and that in itself means we have some of that otherness laid upon us. So Winnicott is displaced by Lacan, and rightly so: the subject is not confirmed in what it is by the other's ministrations, by a process of accurate reflecting back of what is really there; instead, the subject is produced as something else by what the other does. Back in the terms of narrative, one can say that putting something into language changes it; as we speak, so something shifts.

Polymargins
The margins are not, therefore, places for some truer essence of the human subject, and certainly not just for a technology for the recovery of something that the centre (e.g. in the form of mainstream empiricism) has left out. The margins are polymorphous, in this instance meaning that they celebrate additional layers of speech out of which the complexity of experience can emerge, with the key signifier of marginality being the refusal to hold to any one, specific narrative. There is no such whole narrative, not because of the ineffability of the human spirit, but just because that is the way things are. There are always 'plentimaw fish' in the sea of stories (Rushdie, 1990); power is marked precisely by its abolition of these plentimaws, by its insistence on pinning everything down. Being central has the sense of being secure, authoritative, knowing; being marginal suggests vulnerability, fragmentation, quibbling, arguing. It is perhaps no accident in this respect that one of the great documents of a marginal, or at least diasporic, people, the Talmud, purports to be authoritative (the oral Law, given by God on Mount Sinai), but in fact is as polymorphous as one can get. Not only are there variant readings, including two major but different versions (the Babylonian and Jerusalem Talmuds), but every page instantiates the value of argument. The main text lies in the centre, this being the Mishna, which presents a reading of the law as hinted at in the Bible, and the Gemara, which is a commentary on-or, rather, a tussle over-the meaning of the Mishna. Around the edges of the page lie the great commentary of the 11th-century rabbi known as Rashi (some of it in ancient French) and opposite that are those of his followers and other commentators, who much of the time are at odds with one another. All this is studied as if it presents authoritative truth, but this process of study is itself a mode of ironic enactment: students are in fact learning the multiple contradictory debates that sustain the process of argument and critique that in turn maintains a lively, healthy marginality. In a sense, this multi-voiced text, with the arguments it spawns from one generation to the next, is a guerrilla force, a parody. If someone had the 'truth', that would be the end of it, but no-one does; there are only 70 (meaning, innumerable) ways of understanding.
There is much to be said in favour of this '70' and its wariness of anything that moves us towards adhering to a single truth. This also promotes wariness of the kinds of psychoanalytic approaches to material that find in it a set meaning: for example, psychoanalytic approaches to qualitative research that see 'underneath' a narrative the true significance of what the subject is saying, normally in the form of a set of defences against anxiety. As in the clinical situation, so in the research setting: interpretation of this kind moves us rather briskly towards what Parker (2005b, p. 108) sees as a major problem with Hollway and Jefferson's (2000) 'Free Association Narrative Interview' procedure: that it is individualizing, essentializing, pathologizing and disempowering; the key complaint being that it is organized around a pre-set discourse that imposes an expert account on the research participant in a typical (of psychoanalysis as well as psychology) 'researcher knows best' set of moves. Any fixing of the meaning of the subject in this way risks the loss of the polymorphous tendencies that upset claims to truth and hence maintain the healthy marginality of a resistive subject. Finding a specific psychodynamic source for a subject's 'choice' of location in discourse (e.g. Frosh et al., 2003) now seems to risk a kind of ungrounded reductionism that fixes subjectivity in one, predetermined grid, usually that of a biographical difficulty with one's parents.
I have a different account to give here, which picks up on the Lacanian take on discourse analysis so usefully provided by Parker (2005a) and which will eventually bring into play the second kind of exclusion that constitutes what 'cannot be said'. In the course of his rich paper, Parker comments: The reflexive position of the discourse analyst is an issue here, for when one approaches a text in hermeneutic mode as something we can 'understand' because it is like our own framework (or even because we recognize it as being the mirror opposite of what is familiar to us), this, for Lacanians, would betray the stance we are taking as lying on 'the line of the Imaginary' (imagining that we interpret from outside the text). The task of an analyst is to work on 'the line of the Symbolic' (working within the domain of the text), and to open up the text by disrupting and disorganizing it so that its functions become clearer, including its functions for us. (p. 177) This is not the place to engage in detail with the technicalities that follow from this-after all, Parker warns us: 'If there are elements of technique that can be derived from the account in this paper, they will need to be explicated and warranted each time for each piece of analysis' (p. 178). However, the idea that we should approach a text not as something we can understand, but rather as something waiting to be opened up, is precisely in line with my argument for the polymorphism of marginal, 'disintegrated' qualitative research. Faced with the narrativizing tendency in which we seek to integrate what present themselves to us as fragments, I want to promote the 'disruptive and disorganizing' momentum of a reading strategy that insists on there being too many possible interpretations, too much going on. However, as the meanings that surface thereby slip away, there is a further suggestion of something threatening that gives momentum to all this productivity, undermining meaning at the same time as it makes it happen.
Holding on to 'disrupting and disorganizing' readings is not an easy task, because the sense-making tendencies of interpretive analyses are deeply engrained culturally and-if the account given earlier has anything to itfulfil defensive functions against the terror of personal disintegration. In addition, narrative material is often worked-over so that it does in fact 'make sense', either because it is constrained by the questioning of an interviewer, or because the stories told by a speaker are rehearsed, in whole or in part. Indeed, having expended so much effort in articulating the importance of not being seduced by narrative, in an earlier version of this paper I then produced a loving, rather romantic and holistic interpretation of the text to be presented below-coming 'full circle and ending up facing backwards', in the evocative words of one reviewer. This is perhaps not only a personal failure, but a symptom of the difficulty and of the strength of the pull towards integration; falling into the abyss is not an easy choice.
So, heart and head in hand, I shall try again, using the same over-worked text as a challenge. The example is taken from an interview with the novelist A.S. Byatt (Frosh, 2004); the published version is edited, but raises plenty of issues even in that form. In this brief extract, Byatt is responding to a question about her story 'Sugar' (Byatt, 1987), in which she describes a little girl leaping towards her father, who has just come into the room, having been away in the Second World War. Byatt comments: As you can't rely on language you also can't rely on your memory. I have formed an image in my mind of the war, which leads up to this leap. It consists of my father going away surrounded with canvas buckets and he wasn't there and we talked about when he would come back. So there was a long thing in our heads about this person who wasn't there who would come back, and there was the knowledge that he might not come back. And when he did come back, he came back at one in the morning and he came into the bedroom, and that was the moment when the waiting stopped, and the next bit of my life began. I stood up in the bed and leapt over the whole of the next bed and landed on him, and even as a child I thought, 'That's over now.' Now it's only a memory and, having said that, I have partly rescued it but I've also partly forever removed it. I've just told it to you and it wasn't your father and it wasn't your experience, but somewhere in you is now the idea of this child that did this leap. I had this terrible feeling: 'It's over now. Never again will he come back and I will leap.' The scenes that make me weep in literature are always the moment when a lost parent is joined to a child. I weep at the end of Peter Pan and I weep at the end of The Railway Children when the little girl runs along the platform and cries, 'My daddy, my daddy.' (as cited in Frosh, 2004, pp. 147-148) The emotional impact of this piece of text no doubt relates in part to biographical factors in a reader (e.g. one's own experiences of separation and loss) and possibly also to extra-textual and publicly available knowledge about Byatt herself, including her experiences with her own children. But enough of it is textual to make it worth some attention. At the risk of reifying the text too much, I want first to examine its structure as a piece of narrative. The attempt here is to break up the flow of talk/text, marking its various separate ideas and exploring how these are interlinked as a piece of rhetoric, as something best understood as persuasion. That is, the speaker-as it happens, in this case a particularly 'practised' one-is using story-telling conventions to present herself and her argument in a certain way, and one point of interest is to consider just how she does this, and with what effect. Here, I am drawing on some traditions of 'critical narrative analysis' (Emerson & Frosh, 2004) that deliberately distance the reader from the text, at the same time attending to the emotional and conceptual movement of textual themes. Although this process can be extraordinarily long-winded, it is possible to get some way simply by re-transcribing the text into idea units and then organizing these into stanzas (idea units with linked ideas) and strophes (consecutive stanzas working on related notions). These stanzas and strophes are named (an admittedly arbitrary process) and together provide a guide to the flow of text. For Byatt's speech, the result looks like this: Strophe 1: language and memory are unreliable Stanza 1: you can't rely on language and memory As you can't rely on language you also can't rely on your memory. Strophe 4: as the leap occurred she knew it was over Stanza 8: the leap I stood up in the bed and leapt over the whole of the next bed and landed on him, Stanza 9: that's over now and even as a child I thought, 'That's over now.'

Strophe 5: memory rescued and projected in the telling
Stanza 10: now only a memory Now it's only a memory Stanza 11: telling rescues and removes the memory and, having said that, I have partly rescued it but I've also partly forever removed it. Stanza 12: the experience is now in the listener I've just told it to you and it wasn't your father and it wasn't your experience, but somewhere in you is now the idea of this child that did this leap. As one might expect, Byatt's talk here is articulate, emotive and cunningit moves between the problematics of memory and language, personal reminiscence, direct appeals to the immediate listener (the interviewer) and an evocation of a kind of idealized shared nostalgia for childhood, made specifically English through its literary references. In an important sense, her text is 'seductive': it draws a reader or listener in to be moved (as was the case when she spoke, and as happens when it is read out loud to an audience) and convinced that something profound is being said. So let's strip out the text itself, and see what happens with only the arbitrary labels: The thematic headings allow us to note the organization of, and transitions in and through, this piece of narrative. Byatt's context-setter resurrects one made in the story ('Sugar') itself: the unreliability of language and memory, and the specific relationship between the two that generates this unreliability. In 'Sugar' (1987), Byatt writes: 'This event was a storied event, already lived over and over, in imagination and hope, in the invented future. The real thing, the true moment, is as inaccessible as any point along that frantic leap' (p. 248). In the interview held 15 or more years later, Byatt uses a similar formula, structuring her comments with the idea that as the memory is recounted it is both made and 'partly forever removed'. Yet there is something tricky here in this replication of something literary and worked-over that goes before: Byatt has already told us that language and memory are unreliable, so what is this recollection/repetition of the story itself about? She appears to be presenting here something connected to the uncertain nature of the event, even to the subject who experiences it: it was already storied, the act of leaping was already present in her mind before it happened spontaneously; even the significance of her absent father was given as a kind of abstract notion, 'this person who wasn't there who would come back', alongside 'the knowledge that he might not come back'. This seems almost a literal warning of the 'don't believe a word I say' variety: 'I will tell you a story, but of necessity it will be unreliable, because that is the way of things.' The child slips into a set of discourses that are present in children's literature (Peter Pan, The Railway Children) and that give meaning to her own waiting, her experience of the father who is not there and who may or may not return. The leap is presented as a historically real event, but also one made visible through an act of negation: even as a child, the adult speaker claims, she is thinking at that very moment: 'That's over now.' It can never be regained even in the professional activity of someone who makes story-telling her life. The act and the thought about the act contain their own loss-the thing itself has gone, not least when it is resurrected in language, which is an unreliable resource.
Byatt's apparently well-thought-out account of how narrative works to reconstitute its objects is both supported and belied by its own structure. Support is given to it through the careful construction of the tale: (1) here is an example of how language and memory work; (2) this is what happened; (3) the event is lost but regained in language and disseminated in speech; (4) the loss itself is what fuels her tears at stories of parent-child reunion. At the same time, the frame 'language and memory are unreliable' undermines what happens next, including the psychoanalytically inflected claim that the reason why she weeps at the scenes in literature 'when a lost parent is joined to a child' is that she herself has had this experience (and lost it) of re-joining. This is also a comment about the pervasiveness of the narratives out of which our lives are structured: these stories (including their psychoanalytic gloss) were already there. Byatt's own specifically socially classed experience emerges from Peter Pan (Barrie, 1911(Barrie, /1993 and The Railway Children (Nesbit, 1906(Nesbit, /1993, out of the character's scream 'My daddy, my daddy' that pierces the heart of everyone on the train, and that the little girl re-enacts in her childhood leap, and the mature author retells and makes problematic in adult life. Rhetorically, Byatt's narrative turns on itself so that what is played out is a desire to move the listener as she is imagined to have been moved; it is a kind of seduction into a very particular mode of middle-class and colonial story-telling, where the little English girl calls out and is answered. We have the modernist unreliable narrator at the start ('here's an example of the unreliability of language and memory'), then the tale of tension (will he return or will he be killed?), then the drama attached to melancholy (as the leap occurred she knew it was over), then a knowing piece of theorizing directed at the listener (you have this experience now, and it rings true), then an interpretation (I weep because the lost childhood is represented). Narrative is thus undermined and then reaffirmed: with all this unreliability, it brings back lost things, and we all weep as a result.
Byatt's account here is melancholic in form, in the sense of investing in and thus celebrating loss. It also provides its own commentary on the relationship between multiplicity and fixedness. She seems to be saying that before the event occurred-the leap across the room-there was enjoyment of a multiply storied set of possibilities, lived over and over 'in imagination and hope'; there is even a kind of enjoyment in the idea that her father might not return, as this gives poignancy to the girl's feelings and heightens tension for the listener. When he actually returns and the leap is enacted, this plurality is closed down into something singular: the event itself, in its absoluteness. The feeling 'It's over now. Never again will he come back and I will leap' is described as 'terrible', perhaps specifically because what has been multiplied in the narrative imagination, all the possibilities of retrieval of the lost father, have now been concertinaed into one constraining act. Never again will she wait and imagine; he came, it is done, the moment has passed. Afterwards, other things happen: Byatt weeps when the singularity of her experience is repeated in literature, which both evokes its possibilities and its ending, and which gives cultural meaning to what she constructs as her own story. Memory and literature thus take over from the event itself; memory, like imagination prior to the event, can once again be multiple, partly rescuing the experience and at the same time erasing it by giving it narrative form. The event, however, closes things down, cuts through the varieties of imagination with its inconsolable concreteness: at that moment of utter experience there is nothing that can be said, because the thing itself has taken her over. So an appeal is made to something lost that was in the Real, because it had an absolute sense about it, but this also registers what is most feared and excluded. The singularity of the event is projected as something terrible; the consolation of having shared it with a listener is perhaps also a kind of passing around of a hot potato: 'take this terrible feeling and let it work inside you.' What is slipped into place here is not, therefore, a unitary narrative about how memory works to keep something alive, but a mode of aggressivity, of seduction and expulsion, in which the aesthetics of language are employed skilfully to reproduce a very particular kind of class consciousness, but also to chop up affective space.
There is a coda to all this, which can also be derived from Byatt's text and in particular by the framing of her leap as the moment in which everything changes. Tracking the narrative visually here, the parabola of the leap is its highest point: there is the war and the absent father that lead up to it; there is the telling that leads down from it; and there is a frame in which literature is made to stand in as the source of meaning. But the highest point, the 'moment when the waiting stopped, and the next bit of my life began', cannot be pinned down; the leap is an inaccessible set of points that rise up from the girl's bed and end in the father's arms. It is in that inaccessible place of movement that the subject exists; that is what the story and the spoken narrative are about. Yet it is also precisely here, in this parabola, that the subject disappears. This does not mean that nothing can be recovered from it: all such moments of return produce weeping in people schooled in a certain kind of narrative sensibility; and these moments certainly can be spoken about. But they also both defend against and invoke the dissolution that can be found in texts of this kind, in which narrative resources are employed to hold together something that is continually falling apart; and with every speaking of them, every multiple reading and invoked identification, we end up saying: 'That's over now.'

Another Voice
Does this analysis work in the way that I hoped, moving against narrative integrity even with a text that is crafted in order to produce such integrity? Does it do more than re-position me as a 'subject supposed to know', to deploy the conventional Lacanian formula concerning the analyst in the transference (Lacan, 1972(Lacan, -1973(Lacan, /1982? Trying to read this against the grain, and with help from others, I remain unsure. Here is a brief commentary from a colleague, reproduced with permission, with which to end, framing everything that has gone before as a question about what can be achieved in the face of the intensity of the push and pull towards making sense. That it is written by someone else is significant: perhaps each one of us remains trapped in our own narrative-making, unable to fully embrace disintegration without literally falling apart. Or perhaps I am … Well, I just don't know. I really don't know. Did you pull it off this time? Maybe, perhaps. You are certainly more critical of Byatt this time, more aware of her narrative strategies, or even the 'technologies of affect' that she is employing. And you give us a very thorough, and if you like 'multiple' set of readings, offering us an array of different ways to understand the text. But ... in doing this, do you not, once again, set yourself up (yourself here is the analyst/researcher) as the one who knows? And isn't this the point where your multiplicity collapses? Isn't this one of the fundamental traps that Lacan orientates us towards, and that you yourself reference earlier in the paper? It is all very well showing speakers/texts where the gaps are in their discourse through which they emerge or fall as subjects. But in pointing this out you set yourself up as the one who knows, you speak from an outside (that alluring place of the insightful, skilful analyst/researcher), and as Lacan says, that place is an illusion. Because, I think by the end, we are seduced by your reading, precisely by the way that you show us how not to be seduced by hers. You show us, systematically, and clearly, exactly where Byatt's text opens up and where she falls into her own aporias, but in doing so, you cover over your own. You produce a very coherent, understandable, sensible and touching reading in the name of disruption and disorganization. And yes, in the end it becomes just one reading, despite the multiple options on offer. What is the difference between the researcher/analyst disrupting and disorganizing the text, and an analysis that shows how the text itself has its own aporias, its own disruptive and disorganizing moments? If it is only the latter that is attended to, then surely this comes from a position of coherence and the project is once more lost. Unfortunately I have no idea how not to lose this project! Perhaps there are some texts that are just so seductive, so full of narrative pulls that come from certain types of childhood stories, for instance, that they are impossible to resist. Perhaps it is the very choice of text that betrays you. To me, you say, 'What would happen if I really disrupted the text?' and then you say, 'Well actually, forget about that, because there are so many beautiful and moving things to say about this text, so many beautiful and moving ways of saying that she, Byatt, disrupts herself, contradicts herself, covers herself over, produces narrative out of aporias, shows us the aporias in order to seduce us, that now I am seduced by my own voice, my own narrative trajectory into telling you all about them.' But (and it is an important but) in many ways I don't think I fully advocate the Lacanian trajectory, and I have a feeling that neither do you. Why disrupt everything all the time? Of course, what you can't do is claim that you are going to stay on the side of multiplicity, and then bottle out. Maybe you need to tell us that you again tried, and failed, and that there is something important in this process that you have learnt about yourself, and language and narrative. Perhaps one learns that some things just must be covered over, and must stay covered, and that disruption itself can be a silly childish trick. Let's face it, you would not want to see a Lacanian analyst in real life, so what are you doing trying to advocate this as a way forward? Perhaps we have to go back to poetics, even if all the poet's teeth appear to have fallen out; to trying to think aesthetics and politics together. Embrace failure! It's the only way forwards ...