7
The Problem of Speaking for Myself

Philosophers, in my experience, have a horror of self-narrativizing. Although we may be perfectly capable of giving a narrative of our day, the faculty meeting or the recent family visit, being asked to give an autobiographical narrative of our lives in a more serious and comprehensive way strikes most of us with alarm. We know too much about the construction of narratives, and the flattering choices that can be made about where and how to begin, where to leave off, and what to leave out.

Moreover, narratives of first-person experience sometimes receive too much epistemic authority by invoking a rhetorical power (the indefeasible “I was there!”) that elicits our distrust. The successful marketing of contemporary memoirs in itself renders the genre suspect since profit motives cause publishers to underplay concerns about the writer's reliability, leading to some famously disastrous fakes. In general, when any claim is rendered into a first-person report it looks to be unsusceptible to questioning – yet philosophers are trained to reject absolute authorizations of any sort, papal, perceptual, or otherwise. If simple perceptual reports can mobilize philosophers to produce whole industries of epistemic analysis and skeptical hypotheses, then reports of more substantive, meaningful, and complex experiences are likely to fare much, much worse.

Yet the first-person genre remains uniquely powerful, and not just for dubious reasons. And if we can take the first-person point of view as epistemically important without portraying it as immune to challenge, there is less need for skeptical alarm. Moreover, when we reach beyond first-person factual reports and add reflective analysis, the subjective point of view can be a rich site for investigation and insight about the variable conditions of knowing. In other words, the first-person point of view can be a productive avenue for epistemic assessments precisely about the absolute claims of the first-person point of view. Consider this first-person analysis from Maurice Merleau-Ponty (1964: 142) reflecting on his wartime experiences, written just as World War II ended. He writes,

Our being in uniform did not essentially change our way of thinking during the winter of 1939–40. We still had the leisure to think of others as separate lives … all our standards were still those of peacetime. … We lingered over that German lieutenant who had lain dying on the barbed wire, a bullet in his stomach … [looking] long and compassionately at the narrow chest which the uniform barely covered in that near-zero cold, at the ash-blond hair, the delicate hands, as his mother or wife might have done. After June of 1940, however, we really entered the war, for from then on we were no longer permitted to treat the Germans we met in the street, subway, or movies as human beings. If we had done so, if we had wanted to distinguish Nazis from Germans, the peasant or working man beneath the soldier, they would have had only contempt for us and would have considered it a recognition of their government and their victory. … Magnanimity is a rich man's virtue.

Here Merleau-Ponty is recounting not so much his first-person external perceptions as the social context that affected his internal perceptual apparatus, or the way the war changed his sensibilities. Before the war began in earnest, he could be moved to empathize with a young dead Nazi soldier, but afterward, his reactions were forcibly changed. Hence, as he explains, the occupation made Parisians of his milieu

relearn all the childish behavior which our education had rid us of; we had to judge men by the clothes they wore … live side by side with them for four years without living with them for one minute, feel ourselves become not men but “Frenchmen” beneath their glance. From then on our universe of individuals contained that compact gray or green mass. Had we looked more sharply, we could already have found masters and slaves in peacetime society, and we could have learned how each consciousness, no matter how free, sovereign and irreplaceable it may feel, will become immobile and generalized, a “worker” or a “Frenchman,” beneath the gaze of a stranger. (Merleau-Ponty 1964: 142)

It is the rupture of the Nazi occupation that renders more visible to Merleau-Ponty his previous individualist assumptions about how one could simply choose to repudiate the importance of group identities. But in making this point, he is offering not simply a relativist take on contrasting conventions, but also an epistemic criticism of his previous perceptual practices. Group identity categories were there in peacetime as well, with attached ranks and privileges, but his (entitled) belief in having a sovereign consciousness obscured this from sight. He could have seen it then if he had looked “more sharply”; he could have seen the “masters and slaves in peacetime society.” The fact that he did not see these things wasn't because of the vagaries of his neural mechanism, but rather because of what we might today call a learned ignorance, a white and male and majoritarian entitlement to the belief that individual modes of interaction can transcend their situated identities.

As a result of the disruption of the war, Merleau-Ponty becomes reflective about the habits of his interior life, and his practices of perception become visible to him as contingent rather than necessary ways of being and of interacting with others. In a series of post-war essays he describes the changes the war wrought in his thinking about himself, his freedom, and his social relations with others. In particular, he reflects on how national and ethnic identities came to have an all-determining significance under conditions of occupation, controlling the possibilities of interaction at the emotional as well as the physical level. The freedom of the individual to make his or her life, as well as to choose the frame by which he or she formed judgments of others, became more circumscribed by wartime conditions, and in that new experience of constraint the contours and constitutive conditions of the previous period became visible to Merleau-Ponty. This is not just a shift in practice but a reflective awareness about the social conditions that affect the formation of our perceptual practices.

Merleau-Ponty deftly uses the process of self-narrativizing as a way to make a general claim that empathy, and even moral agency, takes its shape from structural conditions. Like other existentialists, Merleau-Ponty found the war and the occupation a challenge to the philosophy of individualism. Existential phenomenologies were then revised to emphasize facticity, embodiment, and limited choices, and leading figures such as Sartre, Beauvoir, and Merleau-Ponty himself began to work on synthesizing existentialism with elements of Marxism and more robust analyses of the sphere of the social.

Merleau-Ponty's post-war personal essays, then, provide a good model for how to think about the practice of memoir from an epistemological point of view. He is giving not a simple perceptual report or narrative of events, put forward as epistemically indefeasible because it is told in the first person, but a theoretical analysis of subjective experience that shows its pitfalls as well as its potential. He offers a reflective consideration of a first-person experience, a critical approach to memoir that explores precisely the changeable nature of our interior lives – the contingency of our very perceptions – and also the ways in which we can be misled into thinking that our subjective life is natural and spontaneous without it being affected by context. But he could not make this very argument in as rich and detailed a way if he was not doing it in, precisely, the first person.

Many feminist and critical race philosophers, including myself, have found Merleau-Ponty's subsequent philosophical works to contain rich resources for showing how oppressive social relations, such as sexism and racism, can be sedimented into our habitual embodiment, imaginary, and perceptual practices. The realm of first-person experiences becomes, then, not the domain of absolute authorization but a site for both discovery and critique.

But in raising critical questions about how our interior life can seem natural and obvious, and its reportage transparent, isn't Merleau-Ponty calling into question the very validity of self-narratives? Via regress, in other words, can't we call into question his 1945 observations as much as he himself called into question his 1939 ones?

The Feminist Debate

There are contrasting feminist views about this question and about the approach in general that we should take to the issue of speaking for oneself. In this chapter I want to clarify what is at stake in this debate by contrasting the work of Sue Campbell and Judith Butler. In her Giving an Account of Oneself (2005), Butler expresses skepticism about the very possibility of transparent self-reports, given their necessarily mediated nature, and urges that skepticism is the key to responsible self-narrativizing. In contrast, Campbell's intervention in the controversy over retrieved memories in her book Relational Remembering argues that we must overturn the long history of disrespecting women as “rememberers” and denying women a “self-narrative position” (2003: 47).

Both Butler and Campbell offer sophisticated, complex, and nuanced arguments that stand out from much of the work on memory in mainstream philosophy since, like Merleau-Ponty, they thematize the way in which power structures can condition our self-narrativizing. But this just leaves us in more of a quandary. If, as Campbell persuasively asserts, the epistemic position of credible self-narration has been arbitrarily denied women for reasons that are more political than legitimately epistemic, then we need to defend women's capacity to speak reliably for ourselves. Yet if Butler is right that the representational claims made in self-narratives are suspect assertions of coherence and transparency that necessarily deny or at least downplay the opacity and elusiveness of the self, then how can we, in good philosophical conscience, champion women's presumptive epistemic authority?

Butler's concerns are not simply the excesses of a deconstructive approach that can overplay the cause for skepticism in everyday life. Rather, they resonate with the sorts of considerations Merleau-Ponty raises as well as with recent accounts in the philosophy of mind about the many ways self-knowledge can be tripped up at the level of “aleifs,” or automatic, habitual beliefs, in our prejudicial perceptual framing, implicit affordances, and other assorted epistemic dysfunctions or complications (e.g. Gendler 2008). The sort of worries that emerge from the world of cognitive science and social psychology may be different from the ones Butler gives, as we'll see, but there is a shared set of concerns about how realistically reliable first-person knowledge can be given the insurmountable problems with our self-knowledge. To repeat, Butler's presentation of these concerns, along with Campbell, addresses in a serious and sustained way considerations of power and politics in the domain of self-knowledge and representation. Besides Campbell and Butler, Lorraine Code (1995), Paul Ricoeur (1995), Susan Brison (2002), Miranda Fricker (2007), and José Medina (2013) have relevant arguments to this debate worth considering.

The debate over how feminism should treat women's self-narrativizing connects to earlier debates between analytic and continental feminists over how feminist theory should approach questions of experience, subjectivity, knowledge, and agency. It appears that that debate has not been entirely put to rest, but has instead transferred to new topics.

Epistemic Reliability

The key question about speaking for oneself is the question of epistemic reliability, but I would urge us to adopt an expansive rather than narrow formulation of this concept. First of all, the idea of speaking for oneself well entails, I would suggest, that one is characterizing some experience or transformation or set of events – some prior self, perhaps – with clarity and veridicality, that one is not totally and completely off-base, out of touch, or, as we say, out to lunch. This is crucial for uptake, what Campbell emphasizes many of us are denied. The point is not simply whether women can speak for themselves, but whether others should take what women say as epistemically reliable or as truthful. Anybody can talk about themselves, but the question is how to assess the content of what they say.

But we need an expansive understanding of this, one that precisely aims not at truth in a narrow sense, or a focus on factual details, but at understanding, in Catherine Elgin's (1997, 1999) sense of a more explanatory discourse than a string of facts with representational content. Merleau-Ponty may misremember various facts about the occupation of Paris, such as the timing of certain events, but what we really want to know is whether his assessment overall evinces a real understanding of what happened, even if he fails to get all of the details right. This issue is especially important to sexual violence: women's and girls' first-person testimonial accounts of sexual violence are regularly discredited and derailed in court on the basis of small-scale factual discrepancies. On the stand in a rape case without corroborating witnesses, the accuser's account is judged by the veracity and consistency of small details that can be checked. Holocaust survivors (women, as it happens) have also been discredited for factual inaccuracies, as if their overall accounts can be dismissed if they thought there were four chimneys instead of one at Auschwitz.1 Thus, small factual untruths are taken to discredit the larger truth claims, as if misremembering the exact time period of a childhood rape or the color of the perpetrator's shirt or the number of chimneys in a death camp means that the story is a likely fabrication.

A truthful account, in Elgin's sense, is less about the small facts that can easily be checked and more about the larger claims that make up a genuinely perceptive insight.2 What we seek is an account of events and experiences, and these are not necessarily built up out of small perceptual details. In fact, the misremembered details may indicate larger truths: a memory of four chimneys rather than one (the correct number) may be how the mind imparts the sensation of being overwhelmed by the enormity of the carnage and the sense of terror and despair experienced by victims, which is in fact a truthful memory. Mistaken details, then, can be a way in which the mind imparts truthful aspects of reality as well as of subjective experience.

So if the debate over self-narrativizing turns on the truthfulness or epistemic reliability of the form, how should we judge reliability if not based on detailed facts? It strikes me that the feminist discussions of this topic nicely foreground the idea that we need to consider more than the usual suspects, more, that is, than sincerity, reliable memory processes, epistemically virtuous motivations, or competent perceptual and cognitive ability. These are necessary, yet wholly insufficient. A rapist might recall a rape in great detail yet “remember” the victim's willingness and consent, interpreting his victim through a frame that entirely misrepresents and misunderstands the nature of the interaction.

Recall Merleau-Ponty's vividly reflective memories about the occupation of Paris. His self-narrativizing contains an epistemically normative content, but without assuming that the truth of our lives can be captured in full unambiguous perfection. We can improve our cognitive and perceptual practices by reflectively analyzing our narratives of self over time without presuming to achieve absolute clarity and a final, complete account. Moreover, if we castigate the activity of self-narrativizing as irremediably problematic, we lose the very process and vantage point we need to improve our understanding.

From Merleau-Ponty we might make two general observations. First, the sincerity and honesty with which we enter into a process of speaking for ourselves is woefully insufficient for reliable results. The individualistic intellectuals Merleau-Ponty describes before the occupation, among whom he counted himself, and who had interacted with others on a self-consciously egalitarian basis in which the markers of nationality or class identity (uniforms, for example) were deliberately ignored, were ill equipped to see or fully comprehend that this mode of egalitarian-minded (or high-minded) interaction was being made possible by their very race, class, and gender and by the political condition of living within a liberal democracy that, at the moment at least, was not at war. Thus, understanding one's own experiences – indeed, one's own self – with any depth or even adequacy is in no way guaranteed by an orientation toward honesty. The second and related generality we might infer from Merleau-Ponty's account is that rupture or temporal distance is a useful ingredient for the capacity of describing one's self. Like the owl of Minerva, it seems to be that when I have somehow passed through an experience, such as surviving a trauma, I can begin to understand more fully the way I was before, and all the everyday and intimate changes that were wrought by the rupture. Many memoirs of sexual trauma recount just such a “before and after” of bodily habits and degrees of self-consciousness (Brison 2002; Gray-Rosendale 2013; Freedman 2014). “Our ideas about ourselves and our worlds had been forever altered. We'd been broken open to a reality we hadn't recognized before” (Gray-Rosendale 2013: 236).

The Effects of Trauma

The effects of distance or rupture are surely connected to another general aspect of self-narrativizing, and that is its feature of rendering the self – one's own self – into a kind of object to be observed, about which one as a thinking subject can speak. Trauma is defined as an experience that the self cannot make into an object of analysis. Traumatic experiences are by definition those that cannot be incorporated, made sense of, or given a meaningful articulation, which is why, as Freedman (2014) shows, trauma manifests its effects in intrusive thoughts and nightmares and involuntary physical twitches, the mental detritus that exists beyond our accessibly conscious selves. She argues that by turning toward and working through such traumatic after-effects, and rendering their content a part of our conscious lives, the intrusions can sometimes subside.

To speak about, or for, oneself is to make the self an object of one's own intentional consciousness. But to turn one's focus toward a self that has been traumatized and has been experienced thereafter as highly unpredictable can feel like quite a risk: the risk of dropping into an abyss of uncontrollable, frightening emotions and thoughts. It is not a place to which one wants to go alone. Hence, the vital necessity of a dialogic context of some sort, a conversation with others or even just one other. Paul Ricoeur (1995) and Sue Campbell (2003) have both argued that dialogue is a crucial enabling condition for crafting the coherent narratives that present continuous selves, for transforming the blurry successions of sensation into a meaningful story. Susan Brison (2002) and Karyn Freedman (2014) bring this idea into the context of rape.

Brison explores the aftermath of her own rape experience and the challenges she faced in being able to narrativize the event. How does one render the self into an object – discrete, continuous, coherent – in which the before and after are so incommensurable? This challenges the rational credentials of the thinking subject, the one who is doing the describing. Brison argues that her very ability to make sense of the event, and to bring it within the existing narrative of her life, required a dialogic context of supportive and empathetic understanding in which she could regain the voice that had been nearly destroyed. “In order to construct self-narratives we need not only the words with which to tell our stories, but also an audience able and willing to hear us and to understand our words as we intend them” (Brison 2002: 51, emphasis added). We need a compatible community, not one where everyone will agree with us but one where we are not hermeneutically marginalized. Only in the act of thinking aloud within such a space was Brison able to narrativize the trauma of the attack in a way that she could make sense of it. But an audience that does not hear our words as we intend them will not repair the thinking, and knowing, subject who was wrought asunder by trauma. One will not get confirmation of oneself as a being with the capacity to know, to judge, or to understand.

Brison is making a more general philosophical claim about the relational character of the self – indeed, its fundamentally narrative character. Sexual violations pose special obstacles here, since one may well want to resist the relational self one becomes when recounting such experiences: the self who is pitied, disbelieved, or simply the one who has been raped and is known as such by another, to be potentially interpreted forever after through that one event. A dialogic space in which one's rape experience is the topic of discussion painfully pulls one into this identity. In this case, acknowledgment – recognition – can be experienced not as helpful but as a kind of existential horror. Yet, over and over, memoirists assert the vital necessity of interactions with others after the event to engage in making sense of it, of making meaning.

Miranda Fricker's (2007) account of dialogic encounters elaborates on why such a supportive context is necessary. Hermeneutic marginalization, which she defines as the effects of social inequality on the domain of socially available meanings, produces a kind of linguistic desert in regard to certain issues or experiences. Consider, for example, that the word “racism” did not exist prior to the early twentieth century, and the concept of “sexual harassment” emerged only in the 1970s. Such parched linguistic contexts can make it difficult to find the right words to express an experience, or to think through how to understand what experience one has just had. Hermeneutic marginalization and a systematic epistemic injustice toward certain groups are to blame. Encountering excessive skepticism, systematic incredulity without reason, and constant plain rudeness can, Fricker shows, produce not only an absence of clarity on the part of the speaker but also outright confusion. The habitual experience of dismissal atrophies the capacity for knowing, for insightful perception, for understanding, for trying out ideas, or for testing one's judgments. When dismissals and incredulity are organized around group-based identities, then the community of those who commonly, and perhaps exclusively, share an experience is incapacitated from reflecting on that experience, making meaningful sense of it, and creating new terms. The hermeneutic background available to all of us is hence adversely affected.

José Medina (2013) makes an important corrective to Fricker's account of this problem. Marginalized individuals, he points out, may often have access to marginalized communities in which they don't experience such epistemic injustice. Within the space of these counter-publics, they can develop ways of conceptualizing and expressing their experiences, even ways of surreptitiously communicating to others of their group when they are in dominant controlled spaces. I agree with Medina, and would suggest that it is surely on the basis of this successful hermeneutic work that we should make efforts to reform the mainstream. In other words, it is not brave individuals who have augmented the parched linguistic deserts of the dominant spaces so much as the collective incursions made by marginalized groups who alter, and improve, the concepts, terms, and meanings available to all.

Once we understand that successful self-narrativizing is dependent not simply on sincerity, but also on what Lorraine Code (1995) called a “rhetorical space” in which concept formation and meaning-making can occur, we must conclude that a concern with truth requires us to be concerned about the political conditions of social interaction. Aristotle believed torture to be a reliable route to truth-telling; consider what a different lesson we have from Fricker, Code, Brison, and Medina. And this contrast does not follow merely from a concern with facts (the torturer's perspective) versus a concern with understanding. In some cases, after all, the facts themselves require hermeneutically expansive opportunities: was I really sexually harassed? Was that rape? Truths that matter are rarely simple matters.

Narrative Selves

The idea that self-narrativizing requires a dialogic process has given rise to theories of the rhetorical self or the narrative self: that is, the self that comes into existence through a practice of creating a story about one's life (see, e.g., Ricoeur 1995). The idea is that selves are a kind of emergent phenomenon of what is an essentially rhetorical process, in which the familiar organization of an individual life – its major experiences, relationships, critical turning points and events – is rendered into narrative form as if for a presentation. The phenomenal reality of life in the moment of experience is a chaotic multiplicity of sensations and thoughts, often elusive and ephemeral, which can yield many possible focal points and interpretations. As Ricoeur says, “It is precisely because of the elusive character of real life that we need the help of fiction to organize life retrospectively” (1995: 162).

In this sense, selves, in the fully formed and meaningful way we have come to think of them, come after, not before, the process of narrativizing. I take it that Ricoeur's point of comparing autobiography to fiction is not to collapse the distinction, but to point out the important similarities. We take the muddy morass of fleeting sensations and organize them into a coherent story, framed for the particular dialogic audience at hand (whether that audience is ourselves or others). Thus both rhetoric and narrative are involved: rhetoric in the form of an address to a specific audience with a specific aim in mind, such as explanation or exculpation or simply to get acquainted; and narrative in the form of an account aiming for some degree of coherence or at least intelligibility, with a beginning and end.

Yet, described in this way, self-narrativizing may invite our epistemic skepticism once again. If the self is so loosely tethered to experience, if experience radically under-determines the construction of self, how can such constructions be judged by their truth content? Plato represented rhetoric as a form of speech aiming only, or ultimately, to persuade, and constitutively vulnerable to obstructing truth. This connotation remains alive when the term is used. We describe a question as “merely rhetorical” to suggest by this that it is not a sincere question aiming for an answer, thus delinking rhetoric from truth. So the idea of a rhetorical self may invoke a process of self formation aiming only for self-aggrandizement, moral innocence, or just an implausible coherence beyond all reality. Surely, intersubjective interactions require some psychoanalytic caution: can we even know what our aim is in explaining ourselves to others? Communicative practice is so often the means by which we establish our position vis-à-vis others, conveying our power, or attempting to ensure our safety, emotional and otherwise.

Ricoeur argues that understanding the rhetorical nature of the self helps to offset these concerns rather than exacerbate them. In acknowledging the fact that we are consciously narrativizing our lives to others, we are made accountable, he argues. Ethical identity “requires a person accountable for his or her acts,” including communicative acts (Ricoeur 1995: 151). Further, in the act of rhetorical construction, we invoke a person with intentions and temporal continuity, a person, in other words, who can then be held accountable. The most effective means to avoid responsibility would be precisely to avoid, or to be incapable of, rendering a narrative identity. There is no possibility of making a promise or of manifesting constancy or fidelity. In fact, it is impossible to build relationality of any sort without the production of an intelligible and meaningful self. Presenting the self is also, and necessarily, an invitation for judgment.

Hence, far from rendering considerations of truth irrelevant, the rhetorical process of narrativizing one's life in a dialogic interaction with others invites a concern with sincerity and with truth. Further, Ricoeur argues that the narrative form simultaneously elucidates the possibility of a different telling: there are always “imaginative variations to which the narrative submits this identity. In truth, the narrative does not merely tolerate these variations, it engenders them, seeks them out” (1995: 148). The possibility of telling the story otherwise, of beginning and ending at different points, of concentrating on different moments, relations, or events, is made apparent in the telling itself, in the constructive and creative nature of the after-the-fact formulation. Rather than a narrative process shutting out the possibility of diverse evaluations and meanings, it makes one alive to the inherent variability of ways of telling the story of one's life. It brings to consciousness the intentional act by which the self comes into being.

On my reading, Ricoeur's account does not eclipse the ethical and epistemic dangers that are made apparent by an acknowledgment that selves are rhetorical productions. Of course, it is possible that we will construct blameless narratives and obscure or deny the intentionality and variability in our story-making. But, unlike some other philosophers, Ricoeur does not take the fleetingness and incoherence of life-in-the-moment as sealing us to an epistemic doom, rendering impossible the construction of meaningful and truthful explanations about who we are and how we came to be. Stories are by their nature revisable, and in offering the story of one's life, one is engaging in a relation with others that simultaneously creates, and acknowledges, ethical possibilities. Fallibilism, or the revisability of our stories about ourselves, does not entail skepticism, unless we assume that our stories must be one-dimensional and fixed for there to be any truth at all.

Importantly, however, narrative accounts such as Ricoeur's have rarely engaged with the political conditions that affect the possibility of narrative agency. The intersubjective and social context of self-narrativizing should remind philosophers, though it does not always do so, that we should consider the ways in which power or political conditions can affect both the form and the content of our first-person knowledge and self-presentation. This is the essential basis of Butler's critical concerns in Giving an Account of Oneself (2005).

The motivating question of Butler's book is to understand the relationship between the moral and the social in our self-narrativizing practices. Butler asks, what follows from the fact that “the form” that questions of moral philosophy “take changes according to context, and even that context, in some sense, inheres in the form of the question?” (2005: 3). Her point is that there are no neutral universals to which we can appeal. When we ask moral questions of ourselves or of others, we are to some extent mouthing the discursive regime in which we live. Thus Butler is raising skeptical questions about the extent of our moral agency. If we have, as we ought to have, a thoroughgoing social account of the development – from the ground up – of the self, of thought, and of moral action, then what are the implications of this social contextualization for our narrative agency, or our agency in crafting self-narratives?

The answer, according to Butler, is that “giving an account of oneself” is a rather hopeless quest. Hayden White, on the book jacket, offers a useful paraphrase of the central theses of Butler's book: “that the height of self-knowledge may very well consist of the realization that, in matters of the self, insight is perilous, perception is flawed, and judgment is weak.” To some extent, White's blurb may seem overwrought, since in this work Butler avows the legitimacy and necessity of giving an account of oneself, of taking and assigning responsibility, even of attempting to formulate a (partially at least) coherent self. She says that “narrating a life can have a crucial function, especially for those whose involuntary experience of discontinuity afflicts them in profound ways” (Butler 2005: 59). She further states that today she would revise the position she took in her 1997 work, The Psychic Life of Power, a book that gives an inordinately pessimistic view of self-formation as necessarily inaugurated in a primal scene of punishment or self-subjection. The focus of Giving an Account of Oneself is moral responsibility in intersubjective relations, and here she avows the necessity of giving an account not as an act of self-flagellation, but as an act of accepting an account-ability that cannot be reduced to the machinations of the Law.

However, Butler's view of what giving an account actually amounts to retains, as White says, a large dose of epistemic skepticism about self-knowledge. And this is because self-knowledge, or what may be more accurately thought of as self-presentation, necessarily involves a large dose of heteronomy: the subject is grounded in what it did not create or produce. The social conditions of subject formation mean that, for Butler (2005: 7), “There is no ‘I’ that can stand fully apart from the social conditions of its emergence, no ‘I’ that is not implicated in a set of conditioning moral norms, which being norms, have a social character that exceeds a purely personal or idiosyncratic meaning.” How can one give an account of “oneself” when one's self is the nodal point of a set of relations and the effect of social norms?

This is the first reason for Butler's epistemic skepticism toward self-knowledge, but she has a second, equally powerful, reason. Our actual lives are successions of discontinuous and fleeting moments where meaning is chaotic and contradictory. Rendering an intelligible self out of this phenomenological raw material requires the work of projection, imagination, fabrication, and lying, all of which take place within contexts of coercive normalization, by her account. Thus we have the fundamental, irresolvable contradiction: while the dictum to give an account of oneself is a demand for sincerity and honesty, it is impossible to narrativize selves with absolute sincerity and honesty. Hence, Butler provocatively challenges the sort of view Ricoeur develops when she asks, “Is the task to cover over through a narrative means the breakage, the rupture, that is constitutive of the ‘I,’ which quite forcefully binds the elements together as if it were perfectly possible, as if the break could be mended and defensive mastery restored?” (2005: 69).

There is yet a third problem, from her point of view, with the possibility of self-narrativizing. This is that the role of the other – the role that the other is playing in producing our reflected sense of a coherent, narrativizable life – is obscured in the autobiographical account by its focus on the construction of the “I.” Here Butler invokes the idea of the others who provide form to our story before we could form it ourselves – that is, our parents, caregivers, family, social context in general – as well as the immediate addressee of our narrative report, to whom the account is directed, by whom the account is prompted or perhaps coerced, and for whom the account is organizationally constructed. “The ‘I’ is always to some extent dispossessed by the social conditions of its emergence” (Butler 2005: 8). This fact creates, she believes, a necessary, constitutive contradiction whenever I speak for, or as, myself. This aspect of Butler's account – the effect of this other-relation on our speaking for ourselves – will be the one most in contrast with the account we will shortly discuss from Sue Campbell.

Despite the fact that Butler explains her claim about the necessary role of the other in self-construction in the context of discussing Jean Laplanche's psychoanalytic account of transference, in which she universalizes transference across all intersubjective encounters, it does not seem to me that her concerns require buying into these particular theoretical traditions. We might instead make use, for example, of general ideas about the hermeneutic contextual conditions that provide the available concepts for intersubjective narrative work.3 The self that emerges from communicative practice is surely constituted in part by and in the interaction. As Satya Mohanty (1997) has argued, our ability to achieve clarity, to feel anger, including anger on one's own behalf, or to name our experiences, is the effect of the historical happenstance of intersubjective conditions.

From this understanding of the necessarily intersubjective conditions of self-constitution, it should follow that subsequent choices such a self makes cannot be neatly divided into the autonomous and the heteronomous. But Butler goes even further to argue that “the ‘I’ has no story of its own that is not also the story of a relation – or set of relations – to a set of norms” (2005: 8). These norms “establish the viability of the subject” (2005: 9). So now we have to contend not only with interdependence, but also with a more nefarious possibility. Although she departs from the Nietzschean view that she is making use of here, that giving an account of oneself always carries the valence of fear, Butler remains committed to the idea that it is always embroiled in coercive norms. This is the main way in which she theorizes the interjections of power into self-narratives.

Butler holds that in the very giving of an autobiographical narrative we are faced with its contingent production as well as with the distance between the narration of a coherent life story and the real incoherence of its lived reality. We know, in other words, that in self-narration we are making the self into an object, an object that, by nature, we cannot be. Because we know this, at some level, conscientious attempts at thorough self-narration – and here she uses the example of the therapist's couch – encounter the “common predicament” of unraveling. Coherence, self-sufficiency, even temporal organization, resist elaboration. Such a narrative project as “giving an account of oneself” must then necessarily try to “cover over” the impossibility of the project. The I who is offering the narrative, who is giving an account of myself, is structurally distinct from that self that is conjured by the words, that is the subject of the story.

Butler is right about the phenomenal complexity and elusiveness of experience, and about the necessarily social constitution of the self, yet her argument fails nonetheless. Her analysis overplays the disconnect between the self that narrates and the self that experiences, so much so that the “I” is permanently disabled. Because norms not of my making are those that bring me into being, she believes I cannot narrate or acknowledge them in any way except as undermining the credence and coherence (i.e. intelligibility) of my speech. I can only point to the constitutive inadequacy of my self-presentation. Butler says these heteronomous norms are “the condition of my speech, but I cannot fully thematize these conditions within the terms of my speech. I am interrupted by my own social origin” (2005: 82). Yet notice that this analysis presumptively separates the “I” from its social origin and then presents that “I” as bereft of social determination. The sphere of the social is portrayed as an interruption of my narrative, and this is why the “I” remains opaque, defying “narrative capture” (Butler 2005: 80). In finding disabling the fact that “my” speech is not entirely my own, Butler is herself positing an ideal norm here of an autonomous self. Were our selves not constituted interrelationally and socially by heteronomous conditions, then we could self-narrativize with success, on her view. She says “one can give and take recognition only on the condition that one becomes disoriented from oneself by something which is not oneself, that one undergoes a de-centering and ‘fails’ to achieve self-identity” (Butler 2005: 42). The judgment of failure here presumes an implausible concept of a non-relational or non-social self-identity.

The mere fact of the disjuncture between the “I” who narrates and the self that is narrated is not a particularly strong argument for the inability to narrate a life; it's at least a defeasible problem within the terms of narration. Perfect knowledge is not required for partial success, and partial success does not merit the rather catastrophic tones of narrative failure that Butler describes. One can acknowledge that the speaking “I” has limited agency and autonomy, and that the self of which it speaks is a partial construction, and that the “I” is always containing within itself the contingency of the social. Nor do I think her claim, influenced by Nietzsche and Foucault, that self-narration is the result of a coercive prompt is a plausibly general characterization. We are not always standing before a grand jury trying to explain ourselves; sometimes we are in a genuinely loving and supportive discursive space. And even if no discursive space is altogether free from power or desire or the operations of transference, there remain significant distinctions between the political valence of diverse spaces, distinctions that Brison's (2002) and Fricker's (2007) work, for example, helpfully describes. Butler may be right that the moral and epistemic agency of “selves” cannot be ultimately disentangled, but she has a tendency to view this necessarily social dimension (or what she calls social inscription) as necessarily oppressive, a claim I believe lacks philosophical warrant.4

Yet Butler is right to urge a focus on the normative context of self-narration and the disjuncture between the way linguistic utterance characterizes an event and the way it is lived. “That which I am,” she says, “defies narrative structure” (Butler 2005: 80). She is right that the narrative structure imposes an order we cannot have experienced, and that ambiguity is lost (2005: 68–9). This is the basis for her claim that attempts at self-knowledge fail. What she calls our “opacity,” or resistance to transparent representation, “resists all final illumination” (2005: 80). Yet notice that where Ricoeur finds in the act of self-narration a strong potential for acknowledging contingency and variability in our self-understanding, Butler portrays an inevitable failure.

Interestingly, Butler's scrupulous attention to the exact conditions of representation is born out of her concern with a truthful account of the foundations of things. She suggests that the idea of giving an account of oneself implies false claims about the conditions of both agency and selfhood, about how an account originates and what makes it possible. Her arguments raise important and troubling questions concerning the epistemic legitimacy of claims about ourselves and our experiences. The necessarily heteronomous character of the norms within which we are embedded calls into question, for Butler, what we are doing when we judge or interpret. Who can take responsibility for the kind of judging and interpreting that occurs? How can we possibly know if our judgments or interpretations are plausible, legitimate, germane? The answer is: we cannot. There is an indefeasible partiality (or incompleteness) and ambiguity that characterizes all meaning and knowledge. This is not an implausible claim on Butler's part.

Butler makes two moral observations in regard to the fact of this indefeasible partiality and ambiguity. First, that when we condemn, that is, when we judge as if unambiguously, we are doing a violence. “Judgment,” she claims, “can be a way to fail to own one's limitations and thus provides no felicitous basis for a reciprocal recognition of human beings as opaque to themselves, partially blind, constitutively limited” (Butler 2005: 46). Judgments that are given unequivocally always conceal their inevitable uncertainty, which indicates that they have different aims than truthfulness. We commonly call this self-righteousness, since it effects a sharp separation between ourselves and that which we condemn, a clear line of demarcation that purges the abject, as it were. Better to offer a more measured account, with an always fallibilist self-consciousness, a recognition of our own opacity to ourselves, and here is where Butler finds the seeds of moral hope. Again, this is not an unfamiliar claim: that epistemic humility leads to good ethical outcomes because it enhances our willingness to listen rather than speak, and to hold back from decisive or permanent denunciation.

Do Butler's claims here make sense? And how do they play out in the context of the general, even generic, epistemic disauthorization of women, or the specific disauthorization of survivors? Can normative judgment cohabit with a recognition of our relational agency and constitution within power?

Relational Selves

Sue Campbell's Relational Remembering: Rethinking the Memory Wars (2003) can read like a rejoinder to Butler, though it was published two years earlier. Campbell's main focus, as the title suggests, is on memory, a critical issue in relation to sexual violation, but in assessing memory, she develops a constructivist account that emphasizes the intersubjective context of self-narrativizing and, hence, of the self. Thus, in important respects, her approach accords with Butler's, yet she comes to radically different conclusions.

Campbell develops her account of memory through an analysis of the debates that were generated by the False Memory Syndrome Foundation (FMSF). The FMSF was founded in 1992 as a lobby on behalf of parents accused of sexual abuse. It claimed that accusers had been led by “suggestive therapies” to construct memories of non-existent events. In some cases, the FMSF blamed politically and economically motivated therapists for unduly influencing clients, but it also pointed to new empirical studies about the suggestibility of memory to raise doubts about accusations, and a number of creditable scientists supported its thesis that false memories constituted a “syndrome.” The FMSF created a Scientific and Professional Advisory Board whose members could provide expert testimony in trials as well as pursue experiments to test the thesis of suggestibility. This Board conveyed its views in numerous public forums and popularly accessible publications. Hence, despite the fact that the FMSF had a polemical agenda funded and supported by accused parents, its project gained credence from memory studies conducted by scientists that provided evidence that the nature of human memory is highly suggestible under even the best of conditions. Not only are our memories unreliable, researchers were suggesting, but new “memories” can actually be implanted through subtle suggestion under receptive conditions.

These studies and their public dissemination helped to create widespread incredulity toward the mounting claims of child sexual abuse crowding the daytime talk shows in the 1990s. As Campbell shows, the studies about memory suggestibility entered into a social context that, owing to two important components, was already inclined toward viewing accusers’ claims with scepticism. One of these components was the persistence of “harmful stereotypes about women's passivity” that lent support to the claim that female accusers were highly receptive to suggestion, and the second, related, component was a tendency toward individualist notions of subjectivity or self-formation that created a predilection “to regard the social dimension of remembering only negatively, as a kind of threat or contaminant to memory” (Campbell 2003: 8, my emphasis). The overly confident conclusions put forward about the nature of memory by such star scientists as Elizabeth Loftus were significantly magnified in the public domain of discourse (as well as within philosophy, I'd suggest) by the related ideas already powerful in this context about women's inherent suggestibility and the properly individual nature of remembering.

The memory wars portrayed a scenario in which the “facts” as determined by objective and apolitical scientific methods bravely stood their ground against a politicized domain of feminist therapy and over-emotional activism. Politics and values were thought to influence only one side of the debate. Campbell's view, however, was that no part of the debate should be depoliticized. She argued that “if we attempt to depoliticize the false memory debates [by, for example, focusing only on the scientific adequacy of the memory studies], we will neglect the effects of power on the lives of rememberers and on which accounts of memory secure scientific and public allegiance” (2003: 15). In other words, many publics, including scientific ones, are susceptible to value-inflected judgments about the plausibility of hypotheses, the assessment and interpretation of evidence, and the validity of conceptual frameworks. All must come under a politically informed analysis, not in order to reach a value-free conclusion, but to develop a reflexivity about the interpretive, social conditions within which judgments are made.

In truth, Campbell was less interested in the false memory debates than in the way in which these debates assumed and in some cases argued for a particular account of memory construction that had ideological resonance with longstanding sexist views and also, she believed, was simply false. The debates over memory called out for a political analysis not only in order to understand how one account of memory wins out over another, but also, and more importantly, in order to understand the necessary conditions of good memory. Her claim was that memory processes are affected by their social context, including their political context, and thus whether women (or anyone) can remember with any epistemic adequacy is socially, and not just neurologically, enhanced or disabled.

Campbell used the term “relational remembering” to name her alternative account which views memory as “an appropriately relational capacity” (2003: 16, emphasis added). Memories are best understood not as transcripts or video-tapes, but as experiences that require interpretation. Memories are always selective, involving a determination of significance. The capacity to remember is enhanced when an experience is deemed significant enough to be repeated, and when its report receives uptake or credibility from others. In a sense, we can sometimes choose to remember, based on these contextual conditions, and sometimes choose to forget. “Strategies of motivated forgetting,” Campbell says, are actually necessary for memory to work properly since we cannot remember everything. And forgetting may also help us navigate traumas that undermine our agency, crushing us with painful and intrusive thoughts, deflating our self-regard, and endangering crucial relationships (2003: 48–9). Even optimal processes of interpreting and narrativizing our memories will always involve intersubjective interactions that affect how we determine significance, how we interpret events, and the impact that the memory of those events has on our lives. The idea that optimal processes would produce comprehensively complete memories beyond interpretation or social influence makes little sense.

Memory is not always or even often an individual affair. We check our childhood memories against our siblings’ and assess our relationship history with friends. Sometimes we reject the memory claims of others, sometimes we learn new things, and sometimes we see the events we did remember in a new light. It can be helpful to discuss our memories with those we were close to when the events occurred, but it can also be helpful to think through our memories with others, who might share related memories in their lives or just help us think through our questions and concerns. When certain “rememberers,” as Campbell puts it, are excluded from these sorts of common intersubjective interpretive processes, because they are claiming abuse, or just because of their female identity, this obstructs their ability to develop their own or contribute to others’ memory processes. When rememberers are portrayed as unconcerned with the truth of their past, their skills are undermined. On Campbell's account, then, the necessity of interpretation in memory processes and the interactive, social nature of how humans remember do not obviate the possibility of truthful memory, nor present us with a priori grounds for skepticism (2003: 18).

Memory selection and interpretation involves what Campbell calls “a sense of self.” To have a sense of one's self requires “opportunities to understand yourself in relation to your past, opportunities to plan and to act on your intentions, and some self-regarding emotions and attitudes” (2003: 29). These ingredients are what make possible the narrative self, or continuous self, within which memory processes occur to select, interpret, and also evaluate the character of our “memory experiences.” Yet each of these requires a social context; even self-regarding emotions and attitudes require an individuation and differentiation from a contrast class of others. Citing the research of Elizabeth Waites, Campbell explains that “memory capacities are socialized from infancy” (2003: 37). Children learn how to organize their memories, focus their attention in such a way as to make some experiences memorable, and encode their experiences with meaning. Thus, children develop the capacity to be self-constituting rememberers who can rely on their memory processes for their sense of self through these social relations, relations that necessarily involve “interests, emotions, and judgments” (2003: 37).

Thus, like Butler, Campbell makes sociality a constitutive condition of selfhood, self-understanding, and self-narration. Yet for Campbell the social nature of selves does not, in and of itself, counsel epistemic skepticism or even undermine autonomy. The question most germane to an epistemic assessment of our self-narration is not “Was it developed in a social context of meaning formation?”, but “What are the specific conditions of sociality in a given context?” In other words, not whether a self was socially formed but how one was. Some social contexts provide epistemic resources that enhance our capacities rather than undermining them.

Campbell argues that our fundamental sociality should call into question the very notion of memory suggestibility that Loftus and other psychologists put forth. The concept of “suggestibility” or social contagion implies that there are non-social memories, non-social narratives of experience, or experiences that are not, so to speak, suggestible.

The implication of Campbell's view is that we cannot isolate or separate the autonomous social and non-social contributions to subject formation if we take the self to be – as Brison, Campbell, Ricoeur, and Butler all argue – essentially relational and social. Butler's skepticism, like Loftus', is based on an odd attachment to foundational aspirations, making a self that is free from the influence of social norms the implicit sine qua non of integrity and the very possibility of sincerity and epistemic reliability. On her view, the fundamental sociality of the self is a prima facie epistemic problem. For Campbell and others, the social nature of the self is the means by which it achieves agency and the very capacity for autonomy or intentional action.

One might be tempted to resolve the differences between Butler and Campbell by reference to a levels distinction. That is, we might be tempted to say that the sort of concern Butler (2005) raises is really only of concern when we are doing an ontology of the self, in which case our epistemic standards are going to be both high and exacting. By contrast, in the memory dispute that Campbell (2003) is addressing, we want to know something altogether different: does the FMSF have a legitimate concern about the suggestibility of memory under specific conditions? In other words, there may be a rather pedestrian sense in which memories can be false, which is altogether different than the philosophical sense in which self-narration is always implicated in questionable claims about the individual nature of self-formation and the possibility of self-knowledge. But I want to resist this easy answer. We need a consistent account of self-narration across these domains, and memory is critical, of course, to the capacity to narrate our lives. How we understand the very basic nature of the self and of self-formation will, as Campbell shows, have a large impact on how we craft memory studies and on how we interpret the data they provide.

The issue of sexual violence provides an especially clear case of this debate over memory and self-narration, given the long suppression of open discussion over these kinds of memories, the epistemic discrediting of women, the subsequent paucity of the hermeneutic community that might be able to form adequate terms and concepts, and also the complex feedback loop between interpretation and experience in this domain, as earlier chapters have explored. Women, among others, have often not been in a good position to develop their memory capacities or a sense of self. The solution to be found is not through more individualism for women but through changing the social context in which memory processing occurs.

Further, we can acknowledge the special challenges in this domain without dropping all concern about the truthfulness of our memories. We might still, for example, maintain skepticism toward the idea that, if a person has five of the eight symptoms of childhood sexual abuse survivors listed in the manual A Courage to Heal (Bass & Davis 1988), then they should form the belief that they were abused. We can, in other words, accept the social approach to memory formation that Campbell puts forward in her “relational remembering” account without losing our capacity to make epistemic distinctions between better and worse memory claims that have been reached within those social processes. The very point of assessing and overturning the epistemic discrediting of women rememberers is to improve memory processes in a way that will assist the whole society.

The troubling implication of Butler's account concerns the role of social norms in judgment. If all we are is a nexus of relations, the end-point of the effects of social norms, then on what basis do we condemn sexual violence at all, including adult–child sexual relations? Butler's counsel against judgment calls seems based on epistemic standards that are unrealistically high as well as disconnected from the best information we have about actual memory processes.

Alternatively, via Campbell's approach, we can develop epistemic evaluations of competing claims of self-narration in the following way: we can ask, what exactly has been the political and epistemic character of the social process in which the person's experience has been interpreted and given meaning? Has there been identity oppression or hermeneutic marginalization, in Fricker's (2007) sense? What have been the political conditions of the rhetorical space, in Code's (1995) sense? Using Brison (2002), we might further argue that, given that subject formation does not always occur under the same conditions, we need variegated rather than a single uniform analysis of its epistemic effects on self-narration and memory. The question is whether the effort of self-knowledge has access to adequate terms and concepts to express and understand its experiences. On this basis we can then judge the particular self in a given instance, rather than all selves under all conditions. Butler's (2005) approach lends itself to an abstract decontextualization of the problems of self-narration; I am suggesting that we can build another approach through the work of Brison, Campbell, Code, and Fricker that builds in a political reflexivity without either a general epistemic skepticism or sacrificing our ability to remain concerned about the epistemic status of the claims of the self.

A permanent fallibilism about self-narrativizing is, of course, always a good idea. Merleau-Ponty raises this as well, even in the midst of his sure-footed 1945 observations. He asks, “But are we not here the dupes of our own emotions? If, ten years hence, we reread these pages and so many others, what will we think of them?” (Merleau-Ponty 1964: 150). His answer is instructive:

Assuredly … those five years have not taught us to think ill of what we once judged to be good, and in the eyes of conscience it is still absurd to hide a truth because it harms one's country, to kill a man because he lives on the other side of the river, to treat another person as a means rather than an end. … The War and the Occupation only taught us that values remain nominal and indeed have no value without an economic and political infrastructure to make them participate in existence. … In man's co-existence with man, of which these years have made us aware, morals, doctrines, thoughts and customs, laws, works, and words all express each other; everything signifies everything. And outside this unique fulguration of existence there is nothing. (1964: 152)

Self-narration, self-formation, no less than ethics, occurs in the midst of relations, background social conditions, politics, life. This is not an a priori problem for judgment, perception, truth. Rather, we need to learn to have a more expansive understanding of the coherence relations that can constitute epistemic validity when we try to speak for ourselves, to narrate our lives. Thus, creating the conditions within which women can speak for themselves even when they speak of sexual violations remains a critically important feminist project.

Notes