Humans are inveterate storytellers. We make incessant and insistent narrative sense of the world around us and of our place in it—so much so that some scholars have suggested “homo narrans” as a more appropriate identifying description for our species than “homo sapiens”.1  Indeed, a long-standing tradition holds that our very self-identities have an essentially narrative shape: that who each of us is is determined by the stories of our lives, and that in some sense we create our selves by crafting those stories.  In this essay, I focus on an especially compelling case of narrative self-construction: Wordsworth’s Prelude.  I argue that we do need rich, substantive selves of the sort delivered by narratives like The Prelude, both in order to evaluate our past actions and to guide future ones.  However, the very feature which makes Wordsworth’s poem so rhetorically powerful as an autobiography—his invocation of a robust teleological structure, which is imposed on him from infancy by Nature—also prevents us from embracing it as a model for our own self-understanding, because it conflicts sharply with modern views about ontology.  Contemporary advocates of a narrative conception of the self, such as Jerome Bruner, Alasdair MacIntyre and Marya Schectman, drop The Prelude’s objectionable ontological assumptions.  But rather than placing the narrative conception of self on a firm metaphysical foundation, this actually intensifies the threat of fictionalism: the risk that the selves we fashion through stories are mere self-deluding illusions. I conclude by gesturing toward the characters within stories as an alternative literary model which avoids many of these problems.

1. Romantic Autobiography as Teleology

Wordsworth was—at least if you believe The Prelude—a pretty lucky guy.  Endowed at or before birth with “that first great gift, the vital soul” (I.150), he was blessed with a childhood surrounded by Nature—what he calls a “Fair seed-time…Fostered alike by beauty and by fear” (I.301-2): fostered, that is, on the one hand by the gentle beauty of breezes and murmuring brooks, and on the other by the awful sublimity of “sounding cataracts” (II.425) and “Winds thwarting winds” (VI.628, 631).  Through this early “unconscious intercourse with beauty/ Old as creation” (I.562-3), he found that “the earth/ And common face of Nature spake to [him] / Rememberable things” (I.586-588).  This “seed-time” in the countryside bears fruit as Wordsworth enters adolescence and begins to appreciate, not just Nature’s external appearances, but her inward animating force.  He has a growing sense that “A gracious spirit o’er this earth presides, / And o’er the heart of man” (V.491-2).  He begins to experience “the sentiment of Being spread / O’er all that moves and all that seemeth still” (II.401-402).  Increasingly, everything he sees appears to pulse with a single animating Spirit; now, he says,

To every natural form, rock, fruit or flower,
Even the loose stones that cover the highway,
I gave a moral life: I saw them feel,
Or linked them to some feeling: the great mass
Lay bedded in a quickening soul, and all
That I beheld respired with inward meaning. (III. 130-136)

Wordsworth’s growing awareness of Man’s place within an animate, spiritual Nature culminates in his choice to become a Poet, creating works of art that reveal the beauty and power of Nature to his fellow man.  Indeed, Wordsworth concludes that he has been “singled out” as a “dedicated Spirit” (IV.336) for this “holy service” (I.53-4): he is a “Prophet of Nature,” tasked to articulate “what an empire we inherit / As natural beings in the strength of Nature” (III.195-6)—or more precisely, tasked to demonstrate how the imaginative powers endowed in us by Nature enable us, not merely to inherit but to co-create an “empire” of the imagination, by spreading the “sentiment of Being” over the bare earth, and by uncovering Nature’s noble dignity in even (or especially) the most apparently ordinary objects and people.  According to The Prelude, then, Wordsworth is not merely a poet, but a dedicated Spirit of Truth and chosen Prophet of Nature:  a high calling, indeed!

To be sure, it wasn’t always easy.  Wordsworth is forced by familial expectation and economic circumstance to labor in academia and in the urban jungle, under  “That burthen of my own unnatural self, /The heavy weight of many of a weary day / Not mine, and such as were not made for me” (I. 21-23).  He worries that he lacks the gifts and fortitude to become a true Poet.  He has “his unruly times; / His fits when he is neither sick nor well, / Though no distress be near him but his own / Unmanageable thoughts” (I.136-139).  At more than one point, “There was an inner falling off,” when “a swarm / Of heady schemes jostl[ed] each other….all conspired / To lure my mind from firm habitual quest of feeding pleasures, to depress the zeal/ And damp those daily yearnings which had once been mine” (IV.278ff).  And at his “soul’s last and lowest ebb,” after the failure of the French Revolution, with “a heart that had been turned aside/ From Nature’s way by outward accidents,” he falls prey to the worst horror of all–  analytic philosophy:

endlessly perplexed
With impulse, motive, right and wrong, the ground
Of obligation, what the rule and whence
The sanction; till demanding formal proof,
And seeking it in every thing, I lost
All feeling of conviction, and in fine,
Sick, wearied out with contrarieties,
Yielded up moral questions in despair. (XI.290-297)

Even in that dark hour, however, even when he was “bedimmed and changed / Much, as it seemed”—even then his “true self,” “Nature’s self,” remained, like “a clouded and a waning moon” (XI.342-350).  His “office upon earth” was still, as always, to be a Poet of Nature; he only had to remember and believe this.  Ultimately, he concludes that “all/ The terrors, pains and early miseries, / Regrets, vexations, lassitudes” that he experiences along the way play “a needful part, in making up / The calm existence that is mine when I / Am worthy of myself.”  Both the joyful play of his childhood and the vexatious trials of his late adolescence thus form materials for Nature’s labor, in crafting a soul worthy of serving as her Prophet.  “Praise to the end!,” Wordsworth says.  “Thanks to the means which Nature deigned to employ…as best might suit her aim” (I.340-356).

This is a teleological conception of life and of autobiography, with a vengeance.  (A teleological structure is one in which a future goal guides the operative causal processes, much as a carpenter’s ultimate idea of a table guides his cutting and gluing the wood.)  As we’ve seen, The Prelude consistently employs heavily teleological language, with its talk of true callings and destinies; and Wordsworth characterizes the poem as the history of “the discipline / And consummation of a Poet’s mind.”  To comprehend this history, Wordsworth must cultivate a global, retrospective view of his life up to this point.  He recounts that he “rose / As if on wings, and saw beneath me stretched / Vast prospect of the world which I had been / And was” (XIV.379-382).  And what he sees from this high vantage is that, “from a very early age,” Nature “had led me on” (VIII.676-679), serving as his active guide and mentor in molding him into a soul worthy of being her Prophet.  It is this retrospective perspective which enables him to conclude that his life is “in the end / All gratulant, if rightly understood” (XIV.379-387): to conclude, that is, that the result explains and justifies his experiences along the way, by transforming them into means toward a significant end.

This teleological structure is implicated, not merely in the poem’s descriptive content, but also in its structure and style. The Prelude is constructed so that “the work shall justify itself” (XIV.414), by demonstrating the growth and culmination of Wordsworth’s poetic powers.  It concludes by pointing to itself as evidence of the fulfillment of its goal—as proof of what has been “our guiding object from the first”: to ensure that “my powers [are] so far confirmed…as to make me capable / Of building up a work that should endure (XIV.309-311).  The ‘enduring’ work to which Wordsworth refers here is ostensibly The Recluse, a larger poem that was never actually published.  But as M. H. Abrams argues, because The Prelude was intended to serve as a “portico” to The Recluse and thus as “part of the same building” as it, The Prelude itself becomes “an involuted poem which is about its own genesis—a prelude to itself.”2

As Abrams also argues, The Prelude systematically tweaks the traditional form of autobiography as confessional in order to promote its model of “natural supernaturalism.” Thus, where Augustine conceived of God as leading him through temptations to his allotted role as God’s servant and advocate, guided by prayer and by Christ’s intercessions; here Wordsworth conceives of Nature as leading him through challenges to his allotted role as Nature’s servant and advocate, guided by poetry and imagination.  By substituting Nature for God, imagination for Christ, and poetry for prayer, Abrams argues, Wordsworth updates the template employed by Christian authors like Augustine and Milton for the modern age, ultimately paving the way for the semi-autobiographical memoirs of Proust, Joyce, and others who replace God and Nature with Art.

As with the more classical confessional autobiographies that preceded it, The Prelude’s invocation of a teleological ontology has two crucial implications for the question of self-identity.  First and foremost, it provides a rich, substantive identity for the self who is its focal subject.  According to The Prelude, Wordsworth’s given identity is, from the outset, to be a Poet and Prophet of Nature.  This essential fact provides an overarching interpretive frame which governs the entire course of his life.  It explains what happens to him and what he does: why he was so attracted to Lake Grasmere as a child, say, or his choice to stint academic study in favor of hiking the French Alps.  It determines which features of his life and character are important and which are merely incidental.  And it licenses certain emotional responses to the actions and events of his life: for instance, given that he is to become a Poet of Nature, it is appropriate to feel pride rather than shame at failing to fit into the social and intellectual milieu of Cambridge.  Further, because Wordsworth is still in the midst of life, the self chosen for him by Nature offers a crucial guide for future action: in particular, it establishes that he should create a monumental epic, and he shouldn’t become a cleric or a member of Parliament, as his family would have preferred.

The second crucial consequence of the poem’s teleological ontology is that it makes the issue of self-identity seem primarily an epistemological rather than a metaphysical problem—a problem of self-discovery rather than self-determination.  Like a stream, Wordsworth is naturally headed in a certain direction; and pursuing a different course of action would be “sinning greatly” against Nature (IV.335).  This is not to say that he is mere putty in Nature’s hands: unlike a stream or an acorn, he must actively participate in his own creation.  Nor is there is any guarantee of success: just as many acorns fall on fallow ground, so is Wordsworth often at risk of being stymied or diverted.  (At one point, he likens himself to another “unruly child of mountain birth, / The froward Brook,” who is “Stripped of his voiceand left to dimple down / (Without an effort, and without a will) / A channel paved by man’s officious care” (IV.51-6).)  At several junctures, his fate depends on a stroke of highly contingent luck: most notably, an unexpected bequest that enables him to quit his day job.  But even these external supports merely serve to “clear[] a passage for [him],” so that “the stream / Flow[s] in the bent of Nature” (XIV. 368-9) once again.  All along, there is his “true self,” “Nature’s self,” waiting to be actualized.

2. Analytic Debunking

Wordsworth is thus presented to a reader of The Prelude as a blessed soul, gifted with a high calling and the imaginative powers to achieve it.  To a mainstream analytic philosopher, though, Wordsworth’s faith in his fortunate fate is likely to look like a bad case of wishful thinking and metaphysical confusion. He is lucky only in the sense that he has succeeded in deluding himself into a self-aggrandizing lie.  It may well be right, as Wordsworth claims, that we ‘spread the sentiment of Being’ over the earth, by imbuing the objects and events around us with a moral life.  Hume makes much the same point, in much the same terms, claiming that the faculty of taste “has a productive faculty, and [by] gilding or staining all natural objects with the colours, borrowed from internal sentiment, raises, in a manner, a new creation.”3 It might even be true that Nature participates in or guides this ‘spreading’ or ‘staining’ in the sense that there is some general evolutionary advantage to projecting moral and aesthetic properties onto nature.  But it is most certainly not the case that Nature designates individual people for particular tasks, like being a Poet, and then manipulates their surrounding circumstances—conjuring an advancing storm, say, or orchestrating their discovery of a little boat on a lake—to mold those individuals into agents capable of performing their allotted tasks.

A natural way to respond to the accusation that The Prelude manifests nothing so much as self-serving delusion is to point out that the accusation depends on treating The Prelude in a flat-footedly literal manner, one which ignores the various ways in which Wordsworth the author signals that he is creating a character—the hero of an epic poem—in the service of a larger project of promoting a secular, naturalized, neo-Humean conception of imagination, beauty and morality.  This response is fair enough as a matter of literary analysis.  Indeed, it allows us to identify another source of The Prelude’s rhetorical power: its appropriation and adaptation of heroic tropes from earlier epics like the Odyssey, Aeneid, Divine Comedy, and Paradise Lost.  However, precisely because this response takes The Prelude’s ‘literary’ status so seriously, it also renders The Prelude problematic as a model for a narrative conception of self-identity.  Many people in the past have believed, and many today continue to believe, in a powerful, purposive Agent who selects a particular destiny for each person and guides us toward its fulfillment. However, such a view is not seriously supportable by contemporary intellectual standards.  Even proponents of Intelligent Design do not claim that specific, substantive self-identities are ontologically given or objectively determined; and Intelligent Design is itself at best a highly marginal view in serious academic discussions.  Further, naturalistic teleological conceptions of evolution as a mechanism for explaining apparently adaptive properties of entire species have come under sustained attack over the last thirty years.4 More fundamentally, however, many philosophers—especially, analytic philosophers—argue that a clear-headed examination of the metaphysical facts reveals that there is no self in the substantive sense required to determine a distinctive ‘bent’ or ‘office’ for any particular individual.  Skepticism about substantive self-identities is not a recent development.  Descartes concluded from the possibility that he could be massively deceived—about his surroundings, about possessing a body, even about his memories—that “I am therefore, precisely speaking, only a thinking thing, that is, a mind.”5 On this view, the only feature which distinguishes my self—my ‘I’—from yours is introspective access: you and I are spectators of different passing streams of appearance and thought, which we each interpret for ourselves.  Hume concludes that even the bare Cartesian ego is an illusion; when I look inside of “what I call myself,” he says,

I always stumble on some particular perception or other….I can never catch myself at any time without a perception, and never observe anything but the perception.  We are nothing but a bundle or collection of different perceptions, which succeed each other with an inconceivable rapidity, and are in a perpetual flux and movement.6

Locke arguably offers the most optimistic view: at least on one interpretation, he holds that a self encompasses the entire stream of memories accessible to a single consciousness:

For as far as any intelligent being can repeat the idea of any past action with the same consciousness it had of it at first, and with the same consciousness it has of any present action; so far it is the same personal self.7

But even this relatively inclusive view is still radically minimal in comparison with the sort of self exemplified by The Prelude.  In particular, even if Locke intends to include all of the memories and thoughts accessible to a single consciousness as elements of the relevant self, he clearly abjures any principle for generating a more fine-grained self relative to which some memories, thoughts, or actions could count as especially important or “true,” while others are “sinning greatly” or not ‘fitting’ in some other way.  Thus, whether the self is taken to be the consciousness that accesses a collection of memories and thoughts, the collection itself, or some combination thereof, Descartes, Locke, and Hume all assign the constituting elements of the self a single, uniform status.  Indeed, for philosophical purposes they all take all selves are fundamentally equivalent: selves are numerically distinct, and trace different paths through space and time; but the selves themselves are basically fungible—a conclusion that Wordsworth would have strenuously rejected as further evidence of philosophy’s sickly absurdity.

Contemporary philosophers like Derek Parfit have continued within this paradigm. They ask whether the same self would persist across radical transformations like bodily teletransportation or brain fission; and they typically conclude that there is no substantive self in any interesting sense of the term.  Subjected to such counterfactual scrutiny, the choice even to elevate a single strand within the manifold of spatio-temporal possibilities as ‘me’ or ‘mine’ appears quite arbitrary; all there really is are more or less stable distributions of physical and psychological properties.8 Even contemporary analytic defenders of the reality of selves tend to offer fairly minimal, neo-Lockean criteria of psychological continuity, or else physical or biological analogues.  Thus, none of these analyses even purport to address the question that Wordsworth challenges himself to answer, and that most of us ask ourselves at some point: “What is my true self, such that I should pursue and cultivate it?”9

3.  The Ancient Quarrel

At this point, we appear to face a deep conflict: the “ancient quarrel” between beauty and truth, with florid literary fantasy on one side and dour philosophical reductionism on the other.  The robust ontological assumptions that allow Wordsworth to present his life’s story as one of development and fulfillment are so metaphysically dubious by contemporary standards that he can be rescued from the accusation of self-aggrandizing delusion only by treating him as engaged in fiction.  However, the philosophical alternative appears to be just as insupportable.

For consider what it would be like to live as the minimalist self sanctioned by metaphysical investigation.  Such a person would conceive of himself as a bare Cartesian ego,10 and would employ only general rational and normative principles to determine how to act at any moment.  He need not be a Cartesian skeptic: he could exploit a full range of beliefs about the world and his particular trajectory through it, including all the memories that arguably contribute to a Lockean self.  The set of this particular self’s beliefs would be unique, insofar as he traverses a unique trajectory through space and time.  He could even be a free moral and epistemic agent (supposing such freedom to be possible); and he might settle, after sufficient rational reflection, on any of various general normative commitments: utilitarianism, say, or deontology.

However, at any given moment, this self would choose his actions by asking ‘What should one do in these circumstances?’, where ‘these circumstances’ are specified by all of his beliefs about how things are, and where his normative principles are determined by general rational considerations.  Such an agent would make no distinctively personal choices, and have no genuinely personal commitments: he would wear whatever clothes, eat whatever food, perform whatever job, and cultivate whatever friends and family best fulfilled the overall rational calculus of ethics, convenience, and pleasure, given his beliefs about how these factors ought to be weighed and about the available candidates for action.

Although a life lived by these precepts might be morally exemplary, I do not think it would produce a recognizably human self.  Such a self would not have any truly particular commitments, only general rational principles plus the inertia of historical contingency.  Nothing he did or surrounded himself with would be genuinely his own: an object of his own care and concern, for itself.  And without any such personal commitments, it is not clear that such a self would be capable of robust emotional engagement—at least, not if we accept contemporary psychological and philosophical theories of emotion which hold that many emotions (e.g. pride, regret, hope, and fear) are grounded in a perception of threat or benefit to “me and mine,” and by an investment in particular strands of the past and future as distinctively my own.11

Nor is it clear that such a life would actually be livable.  Perhaps the core insight of existentialism, articulated most famously by Sartre in Nausea, is that many of our deepest and most important choices are fundamentally arbitrary in the sense of not being decidable through a rational calculus; and that awareness of their arbitrariness produces excruciating anxiety.  This long-standing philosophical or phenomenological claim has been supported by recent empirical evidence about the so-called “paradox” or “tyranny of choice,” which suggests that when people are offered multiple rationally viable choices, they experience more anxiety, are worse at selecting the objectively ‘best’ choice when there is one, and take less pleasure in their ultimate selection than when they are confronted with fewer options.12 Thus, the philosophical love of depersonalized rationality threatens to produce lives that, if they are practically feasible at all, are significantly worse than those produced by cultivating a distinctive self with a constellation of genuinely personal commitments. As Wordsworth puts it, it seems that philosophy is a “speculative scheme[]—/ That promised to abstract the hopes of Man / Out of his feelings,” (XI.224-226) but that ultimately produces only meaninglessness and existential despair.

4. Narrative Conceptions of the Self

Perhaps the choice is not as stark as I’ve made out: perhaps there is a way to reconcile a respectably parsimonious ontology with a practically viable psychology.  Although analytic metaphysical views of personal identity tend to favor reductive or at most highly minimalist treatments of the self, not all contemporary theorists of personal identity are so resolute.  In particular, a wide range of philosophers, psychologists, anthropologists, historians, and literary critics advocate a “narrative conception of the self.”  Although there are important differences among the views of various narrative theorists, Owen Flanagan provides a useful synopsis of the model’s general contours:

Many have converged on the insight that a narrative conception of self is the ‘essential genre’ of self-representation. One useful way of conceiving of the self is as a kind of structured life, structured in part by its temporal components—its beginning, middle, and ending—and by the relations among the various characters who figure in our lives…The sort of connectedness that constitutes a normatively acceptable self or life is the sort that makes for a contentful story that involves an unfolding rationale for the shape it takes.13

The fundamental intuition is, as we might put it, to create substantive, fine-grained selves by imposing a narrative structure on the teeming complexity of trivial remembered detail that constitutes a Lockean self.  More specifically, the view retains the basic interpretive structure exemplified by poetic autobiographies like Wordsworth’s, while jettisoning their florid ontological assumptions about the guiding hand of God or Nature.  Instead, selves are now taken to be self-constituting, in virtue of telling their own distinctive stories.  Thus, Oliver Sacks claims that “each of us constructs and lives a ‘narrative’…this narrative is us, our identities.”14 Jerome Bruner claims that “self-making is a narrative art”15; , “It is through narrative that we create and re-create selfhood…self is a product of our telling and not some essence to be delved for in the recesses of subjectivity.”16 “In the end,” he says, “we become the autobiographical narratives by which we ‘tell about’ our lives.”17 Similarly, Daniel Dennett writes, “We try to make all of our material cohere into a single good story.  And that story is our autobiography.  The chief fictional character at the center of that autobiography is one’s self.”18 Finally, Marya Schechtman articulates the claim that we construct our own selves by imposing narrative structure on our lives in especially explicit and forceful terms, arguing that “constituting an identity requires that an individual conceive of his life as having the form and the logic of a story…where ‘story’ is understood as a conventional, linear narrative.”19 One must have a full and “explicit narrative” of one’s life, she says, in order “to develop fully as a person”20; “weav[ing] stories of their lives [is what] makes them persons.”21

Most advocates of the narrative conception of self-constitution are not very explicit about precisely what they mean by ‘narrative’: Schechtman, for instance, simply appeals to “a conventional, linear narrative.”  However, they generally endorse the two criteria that Flanagan cites in the passage quoted above: a narrative involves a temporally ordered structure, with a beginning, middle, and end; and that structure is shaped by an “unfolding rationale,” or what Noël Carroll calls a “presiding macro-question.”22 And this combination of features effectively produces a teleological structure.  That is, narratives are structured in terms of a “forward-looking” trajectory to their ends;23 and as a result, the ultimate significance of each of the narrative’s constituent events—and indeed, whether it is worth mentioning at all—is determined by its causal or emotional contribution to moving the story toward its conclusion.  Of course, stories may be told out of temporal order, disrupt our expectations about where they are headed, or withhold significant information until later in the telling.  Likewise, even important events may occur by happenstance; they may make sense only in the sense of fulfilling an “emotional cadence”24 and not in virtue of any straightforwardly causal explanation.  But at least in successful, satisfying narratives, initially surprising events acquire “retrograde necessity,” through which “chance is transmuted into fate.”25

When the general narrative criteria of a temporal dimension and an “unfolding rationale” are applied to the lives of individual persons, then they arguably produce a considerably more specific structure, one that closely mirrors the pattern exemplified in Wordsworth’s poetic autobiography.  In the context of individual persons, the temporal dimension is naturally fixed by a person’s biological life (although there may be important antecedents and consequences extending beyond that life, as in Tristram Shandy.) Similarly, the narrative’s “unfolding rationale” becomes that of the focal subject or protagonist pursuing a goal through obstacles—or, in more traditional terminology, a quest.  Thus, Charles Taylor writes that “we must…understand our lives in narrative form, as a ‘quest’,” as an “inescapable structural requirement of human agency.”26 And Alasdair MacIntyre declares,

The unity of an individual life is the unity of a narrative embodied in a single life.  To ask ‘What is the good for me?’ is to ask how best I might live out that unity and bring it to completion….[The unity of a human life] is the unity of a narrative quest…the only criteria for success or failure in a human life as a whole are the criteria for success or failure in a narrated or to-be-narrated quest…A quest for what?…the good life for man is the life spent in seeking for the good life for man.27

One might wonder why autobiographical narratives need to take as specific a form as that of anything resembling a ‘quest’.  Some theorists, like Jerome Bruner and perhaps Kenneth Burke,28 appear to simply assume that all narratives are inherently structured in terms of agents pursuing goals through obstacles; but this is clearly too restrictive, insofar as it rules out many narrative histories, such as those concerning families and nations.  However, I think we can justify something very close to this restriction within the context of an individual’s biography.  If a person’s life is to be explained in narrative terms, then it must be governed by an overarching, forward-looking explanatory trajectory; and if that explanation is to be plausible and compelling, then it must have some significant causal basis.  But further, if this explanatory-causal role is not to be filled by an external agent who manipulates the biography’s focal subject, as in Augustine’s Confessions or Wordsworth’s Prelude, then it must be occupied by the subject herself.  That is, the subject must be an agent who imposes an explanatory unity on her unfolding life by striving to achieve some goal.   Otherwise, we are left either with no unifying, sense-making ‘rationale’ for the narrative at all, or else with a ‘rationale’ that is presented as merely epiphenomenal: as emerging mysteriously from out of a miasma of blind contingency.

5. Problems for Narrative Accounts

At this point, then, it appears that contemporary versions of the narrative autobiographical conception of the self are poised to achieve the sort of highly particularized, personal teleological structure that is manifested in The Prelude while employing only metaphysically respectable means.  Unless one is prepared to hold that all interpretation is inherently poetic and hence untrue, simply because it goes beyond the “bare facts” of raw psychological experience or physics—unless, that is, one is prepared to declare that all history is inherently fictional—then it appears that we can reconcile the ancient quarrel after all, at least with respect to personal identity.  We can, that is, constitute substantive, meaning-conferring, action-guiding selves by constructing narratives which make overarching sense of our lives, so long as we don’t pretend that events did take place that didn’t actually happen.

I think that a narrative conception of the self has its heart in exactly the right place.  It offers the right genus of analysis for developing a robust self-identity, one that captures the richly textured and highly particularized way in which we do, and arguably must, make sense of our lives.  Specifically, narratives provide an overarching interpretive principle for selecting notable events from the teeming mass of trivia that fills the space-time trajectory of each particular individual.  They relate those events to one another into a unified, coherent arc.  And they thereby imbue them with meaning, assigning them explanatory, evaluative, and emotional significance in virtue of their relation to the narrative’s “unfolding rationale.”  Some such principle for selecting and interpreting specific elements in relation to a coherent whole is necessary if we are going to generate robust, substantive selves from bare Cartesian ‘I’s and their attendant bundles of memories and thoughts.

However, I also believe the narrative conception cannot make good on its promise.  Metaphysically respectable selves cannot be constituted out of the stories we tell about our lives, because the same individual cannot simultaneously play both the role of protagonist within her unfolding life and that of narrator retrospectively interpreting the life she has lived.  The fundamental problem is that, as we’ve seen, narratives are structured in terms of a “forward looking” trajectory to their ends; but even lives that eventually exemplify a narratively compelling shape don’t actually have ends until they actually end.  In its effort to cleanse itself of objectionable ontological assumptions, the contemporary narrative conception actually intensifies the threat of fictionalism confronted by Augustine and Wordsworth.

In this section, I present what I take to be the three most pressing objections to the contemporary narrative model of the self.  In §6, I gesture at an alternative mode of interpretation that I believe can capture many of the narrative conception’s insights while avoiding its primary drawbacks.

5a. Temporal Dependence

The first objection to the narrative conception is that it holds the meaning, value, and even existence of selves hostage to their ultimate temporal end.  If narratives are supposed to constitute selves, and if a good narrative requires a satisfying end, then the only good selves will be those whose lives draw to a satisfying conclusion just at the point of death, leaving neither unfulfilled goals nor narratively extraneous time.29 But lives often end at narratively inopportune moments: as Allen Wheelis says,

That big vague thing, that redemptive fulfillment, is an illusion….A symphony has a climax, a poem builds to a burst of meaning, but we are unfinished business.  No coming together of strands. The game is called because of darkness.30

Given the arbitrary vicissitudes of life’s endings, we need to allow for the possibility of robust, fully actualized selves who die with some of their crucial business left messily undone.  We want to be able to say of such people, not merely that they were on a trajectory toward developing an admirable self if they hadn’t been cut off midstream, but that they actually achieved selfhood.  Conversely, we also want to allow that someone could not only once have been, but could continue to be an amazing, robust, interesting person despite the fact that the world has largely left them behind: even if they are, say, stuck in a nursing home, surrounded by mentally and physically decrepit fellow residents, visited only periodically by desultory relatives.  If we identify selves too closely with the lives they live, we risk losing the ability even to describe such a person’s situation as tragic: because this long, empty tail of their life comes at the end, it plays a central role in determining the narrative arc of their life, and thus in turn in constituting their very self.  The narrative conception thus risks stipulating that each of us gets exactly the life we deserve, insofar as the narrative of that life constitutes who we are.

The problem is not merely that the narrative conception assigns undue weight to a life’s end in tallying up a self’s merits and faults.  According to the narrative conception, the significance of each episode within our lives is determined by its place in the arc of our overall life narrative.  But because that arc is itself a trajectory toward the story’s end, no event can be assigned a determinate significance until that life is over—until we are in a position to stand outside the flow of time, and to “r[i]se / As if on wings, and [see] beneath [us] stretched / Vast prospect of the world which [we] had been.”

Louis Mink articulates the crucial point in a frequently cited passage:

Stories are not lived but told.  Life has no beginnings, middles or ends; there are meetings, but the start of an affair belongs to the story we tell ourselves later, and there are partings, but final partings only in the story.  There are hopes, plans, battles and ideas, but only in retrospective stories are hopes unfulfilled, plans miscarried, battles decisive, and ideas seminal.31

Mink’s basic claim here might be interpreted in a merely epistemic manner: as claiming that we can’t know whether our hopes will be unfulfilled or a parting will be final until the story ends.  Within the context of the narrative model, however, this uncontroversial point takes on a more radical implication.  Because the story of our lives is supposed not merely to reveal but to constitute our selves, the narrative view is committed to claiming that we actually have no identity, no self at all, until our lives are over.  More precisely, insofar as the story of our lives at each moment within our lives is compatible with multiple, dramatically distinct futures, our self-identity at each moment is indeterminate among all the possible significances that the narrators at the end of each of those alternative futures would assign to it.

The narrative dependence of interim events on their contribution to the story’s end doesn’t rule out the possibility of ever making narrative sense of our lives as they unfold.  Rather, like the first-time readers of a novel, we can (and often do) interpret events in “anticipation of retrospection”32: that is, by imagining which futures are most likely to unfold from this point on, and how this moment or possible course of action will appear relative to those potential futures.  However, note that we employ our current values and desires in adjudicating among those futures.  And this requires that we actually be someone right now: someone for whom certain options really are better or worse, even though the self who ultimately tells the story of our lives may have very different values, and even though the option we eventually choose will retrospectively become the narratively compelling one.  But this requirement is not one the narrative conception can meet.  Because, that is, it defines selves via narratives of entire lives, it cannot account for the sense in which we are already actual, here and now, well before our story ends.

5b. Teleological Dependence

The second major challenge for the narrative conception is generated by the fact that narratives are structured and unified in terms of a teleological end as well as a temporal one: a “quest,” as Macintyre and Taylor put it.33  The basic problem is that the lives of a large class of selves, who intuitively count as robust and fulfilled, fail to exemplify the right sort of overarching, goal-directed unity.  Perhaps most obviously, many people organize their lives around multiple goals, each of which plays a significant role in constituting their identity, but each of which bears little relation to the others: for instance, they may pursue romantic ‘goals’ that are intentionally quite distinct from the goals that drive their careers.  Similarly, some people—what Galen Strawson calls “Episodics”34 —move through life without any special commitment to long-term, identity-defining goals at all.  Instead, these people savor each moment, and meet each temporary challenge and opportunity as it comes.  (As examples of Episodics, Strawsoncites himself, along with such luminaries as Montaigne, Stendhal, Woolf, Borges, Murdoch, and Bob Dylan.)  In neither case do we want to conclude that these people cannot have selves, or that their selves must be deeply fractured, simply because there is no narratively compelling connection among the disparate episodes or strands of their lives.

A second class of narratively problematic selves does live a strongly goal-oriented life, but in a way that produces spectacularly boring narratives.  These people have as their overarching goal simply to be a certain kind of person: to achieve a particular personality trait, like serenity, for instance, or a certain professional status, like being the town doctor.  When things go as planned, they achieve that crucial, self-defining quality quite early, and simply manifest it in a consistent, ongoing way from then on.  The stereotypical pater familias, farmer, or town doctor coasts through life indefinitely, savoring the pleasures and confronting the challenges of each day and season, but without any particular expectation or hope of substantial change.  Asked to tell the story of their lives, they’d say there wasn’t much to tell, or proudly offer a one-line characterization.  Much like Episodics, they accumulate many anecdotes—variations on an unchanging theme—in lieu of a compelling developmental narrative.  I take it that this is an utterly familiar, even paradigmatic type of selfhood.  But because the narrative conception focuses so strongly on becoming at the expense of mere being, it is forced to disvalue these selves.35 (Indeed, it is notable here that The Prelude ends midway through Wordsworth’s life’s journey: once he has become a poet, there is nothing left to tell, although he continued to tinker with the poem for the rest of his life.)

The third and perhaps most obvious class of persons to pose a problem for the narrative model’s teleological focus are individuals who lack the cognitive resources necessary for composing compelling life narratives or for forming long-term goals.  According to the narrative conception, these deficits are quite dire: Bruner 36 claims that “individuals who have lost the ability to construct narratives have lost their selves”; in support of this claim, he cites Sacks’ description of such people as “scooped out, de-souled.”  However, a closer look at individuals with severe cognitive impairments suggests that this diagnosis is inappropriate.  For instance, William Hirst argues that severe amnesiacs can not only preserve but develop their selves in important ways, “despite the impoverished post-onset narratives” that they generate.37 Individuals with amnesia are capable of developing new behaviors, interests and life ‘themes’, in at least three ways: through “islands of preserved memory” for emotionally and thematically charged events, such as a child’s wedding or a divorce38; through implicit memory which is inaccessible to explicit recall, but which guides current preferences and actions; and through “externalized and collective memories,”39 such as physical and social cues in their environment which consistently remind them of their reality, eventually becoming encoded in implicit memory.  Similarly, children and adults with Asperger’s and high-functioning autism are typically significantly impaired when it comes to narrative generation and comprehension, specifically with respect to causal connections, evaluation, and overall coherence.40 But they still display strong, specific personalities, interests, and commitments, which develop over time; and they certainly do not lack souls—at least no more than the rest of us.  None of this is to minimize the severity of the deficits these people confront.  Rather, it is to emphasize just how strong a claim is involved in the narrative conception’s identification of selves with the narratives that individuals tell about their lives.

So far, I’ve focused on a range of individuals whose lives fail to exemplify the general teleological structure required by narrative. The narrative conception’s commitment to defining lives and selves teleologically is also problematic in another respect.  We saw in §4 that when the “unfolding rationale” that unifies and structures narratives in general is applied to an individual person’s life, it takes the form of an overarching goal pursued by the protagonist.  The problem is that many of the most important aspects of people’s lives and selves arise not from what they do, but from what happens to them: the fascinating woman they encounter a party which they decided to drop by at the last minute, say; or the job that lands in their lap when the company’s preferred candidate decides to change careers; or the debilitating stroke that forces them to relearn basic life skills and finally acquire a measure of empathy.  As the poet Edwin Muir says in concluding his memoir:

What is left to say when one has come to the end of writing about one’s life?  Some kind of development, I suppose, should be expected to emerge, but I am very doubtful of such things, for I cannot bring life into a neat pattern.  If there is a development in my life—and that seems an idle supposition—then it has been brought about more by things outside than by any conscious intention on my part.41

Such influences from “things outside” aren’t merely obstacles to be overcome in pursuit of an ongoing goal.  Rather, they are the sorts of events that fundamentally alter our conception of what our lives can and should be.  They are possibilities we cannot envision for ourselves before they happen, but that we cannot help but make into a fundamental aspect of our selves once they are thrust upon us.

The most tempting way to reconcile a teleological conception of a life’s structure with the fact that in many of our most important aspects we are objects rather than agents, is to posit an external agent—God, Nature, Fate—who manipulates our surrounding circumstances to bring about a goal for us, one of which we may be still ignorant.  But of course, if we embrace this option then we are back with Wordsworth: either ontologically deluded or merely indulging in a playful fiction.  Either way, we lack a serious, epistemically respectable account of our selves in fully agential, teleological terms.

5c. Narrative Reliability

The first two objections I raised to the narrative conception leave open at least the possibility that a wide range of people might fruitfully define their selves in terms of the narratives they tell about their lives—albeit only at their lives’ ends, or in anticipation of such retrospection.  The final objection challenges the possibility that self-generated narratives can ever achieve epistemic respectability—that they can ever escape the twin specters of self-delusion and fictionalism.

The first version of the objection targets the memories from which we construct our self-narratives.  Various theorists have argued that people persistently revise their memories in self-serving ways.42 Such extreme pessimism appears not to be warranted—indeed, subjects are somewhat more likely to remember negative than positive events.43 However, psychological and neurophysiological evidence does strongly suggest first, that the process of recalling an event alters and ‘recodes’ one’s memory in light of the context of recall, and so that we are blocked in principle from an ‘originalist’ access to what really happened, or even our original experience of it.44 Second, it also appears that subjects frequently mold their memories to fit a narrative form.  Thus, people’s spontaneous autobiographies typically conform to specific genres, such as Bildungsroman or tragedy;45 and subjects typically judge the veracity of both their own and other people’s reported memories on the basis of narrative plausibility: on whether those memories involve vivid detail, an intuitively coherent structure, and “characterological” consistency.46

Given just these two respects in which memory is “constructed” rather than given—and there are arguably others—it follows that the mere fact that our memories fit a narrative structure and that we don’t recall any countervailing events doesn’t justify our concluding that that our self-narrative accurately reflects what really happened, or even our past experience.  There is at least as much reason to think that our ex post facto narrative has shaped memory in its image as the reverse.

If narratives are to constitute epistemically respectable selves, it must be possible for them to be checked against some external standard—it cannot just be that “thinking makes it so.”  Among proponents of the narrative conception, Schechtman is one of the few to explicitly address this challenge.  She endorses a “Reality Constraint” on self-constituting narratives, which she proposes to meet by excluding factually inaccurate memories and self-concerning beliefs from the narratives that constitute selves.47 However, given the empirical evidence for pervasive small errors in memory, this proposal produces a substantial risk that the resulting narratives and selves will consist largely of vague generalities, even though we falsely think our selves are richly fine-grained and substantive, and even though we need such substantive selves to evaluate our past and guide our future actions in a meaningful way.48 A second version of the objection targets our experience of ourselves as causally efficacious agents. Daniel Dennett,49 John Bickle,50 Daniel Wegner,51 and others have argued that the phenomenology of agency is an illusion; to support this claim they cite neurophysiological evidence that, for instance, the conscious decision to act occurs after the brain has already initiated action, and that simulating the appearance of causal efficacy often suffices to induce the phenomenology of agency. On their view, we are really just a seething pandemonium of competing “demons” or modular sub-routines, and our self-narratives are just confabulated consolation prizes.  Although they take this argument to tell against the existence of selves under any conception, the narrative model is particularly vulnerable because, as we’ve seen, in the absence of an external Agent, the “unfolding rationale” which unifies a person’s life narrative becomes a protagonist agent or “self-in-charge” pursuing a long-term goal.

If we combine both versions of the objection, it begins to seem quite likely that the narratives we tell of our overall lives are largely confabulated.  Because we want to be the agents of our own destinies rather than passive victims of blind contingency, we tend to impute causal efficacy to ourselves as we pass through the various events that make up our lives, even when that efficacy is lacking.52 And because narratives are one of the most powerful means by which humans interpret the world and our place in it, we strive to impose a coherent narrative structure on the sequence of events we experience, even when that structure is lacking.  Together, the pressures to agency and to narrative coherence lead us to recall our past selves as protagonists in pursuit of a stable long-term goal.  We tend to forget events that don’t fit that narrative; and our narrative is reinforced each time we recall a particular event while retelling our life story.  The result is that we sincerely remember ourselves as having been, and hence take ourselves currently to be, more consistent, efficacious, and goal-directed than we really are.  We interpret ourselves as having coherent histories and stable selves when the reality is likely considerably more messy and inchoate.

6.  From Narratives to Characters

Some advocates of the narrative conception openly embrace the fictionalist conclusion.  Most famously, Dennett claims that “we are all virtuoso novelists,” putting “the best ‘faces’ on [our disunified behavior that] we can.”53 Richard Rorty goes even further, encouraging us to take the liberatory consequences of fictionalism seriously: one of Freud’s great legacies, he says, was to “help[] us become increasingly ironic, playful, free, and inventive in our choice of self descriptions.”54 The problem is that an interest in truth—a Reality Constraint—is more than just analytic philosophical fetishism.  Who I really am and what I really did matters enormously for assessing the moral implications of my past actions and for deciding what I should do in the future.  The rash of legal trials in the 1990’s centering around ‘recovered’ adult memories of childhood sexual abuse provides an especially stark demonstration of both the legal and ethical stakes of getting one’s past narrative right, and of the ease with which even well-meaning people can get it dramatically wrong.  But here, resolute philosophical minimalism offers little help.  To evaluate our past actions and guide future ones in a meaningful way, we need substantive selves that go beyond the bare ontological minimum.

The narrative account’s failure can be traced to two basic assumptions.  First, the narrative account identifies selves with the lives they live.  But lives are essentially temporal in a way that selves are not: they exist only over a span of time, while selves exist at each moment.  And second, it interprets lives in terms of a specific structure, which arcs in a “forward-looking trajectory” to a concluding resolution.  But many lives fail to exemplify this structure in any satisfying way.  Taken together, these assumptions also produce a kind of schizophrenic split even in those selves whose lives do ultimately fit the narrative mold.  On the one hand, the narrative account posits a knowing teller: someone who stands outside of time, surveys the entire scope of a life’s constituent events, and knits them together into a unified whole.  On the other hand, it posits a protagonist, embedded within life as it unfolds, pursuing a relatively fixed goal through largely unforeseen obstacles.  But even the best of us cannot be both of these at once; at most, we can live in “anticipation of retrospection,” attempting to imagine how things will appear to the narrators we will have become by the time our lives end.

I want to suggest that if we drop these two core assumptions, then we can retain what is most appealing about the narrative account, and more generally about an account inspired by literature, while avoiding its most damning faults.  Obviously, I can provide only the barest sketch of how such an account would go here.  But briefly, the idea is this.  Narrative is just one species within a broader genus of what Louis Mink calls “the configurational mode of comprehension”55: the interpretive act of “seeing-things-together,”56 by selecting and structuring a coherent unity out of teeming multiplicity.  On the narrative account, the constituent elements are always remembered events; and the unifying structure is always temporal, and usually causal.  Events and narratives are certainly crucial components in our understanding and construction of ourselves—albeit more so for some people, and at some stages of their lives, than for others.  But many of the self-narratives we care about most are vignettes rather than full autobiographies: short stories about particularly poignant, revealing, or influential episodes within our lives.  And these short stories themselves stand in need of selection, explanation and unification into a coherent overall self.  Further, intuitively, selves also include many other elements that don’t fit smoothly into a narrative account: most notably, standing commitments and dispositions to think, act and respond in certain ways.  Indeed, these are likely to be the first things you mention when asked to describe someone, even if they never end up playing a significant role in advancing their life’s “unfolding rationale.”

Thus, in place of the temporal and teleological structure of narrative, I propose focusing on the stylistic coherence of a character, or on what Iris Murdoch calls a “texture of being.”57 On this model, a self is a distinctive way in which a particular ‘I’ inhabits, interprets, and engages with the world—a particular nexus of dispositions, memories, interests, and commitments that flesh out the bare skeletal ego in virtue of which multiple tokens of ‘I’ count as referring a single entity.  Like a narrative, this nexus imposes a selective, unifying structure on a mass of more or less disparate elements.  But both the kind of structure and the relation between interpretation and self differ between the two models.

Like a narrative, a character selects just a subset of an individual’s feature as relevantly identity-constituting.  But rather than asking whether a certain event advances the plot’s “unfolding rationale,” the character conception selects directly for whether a feature significantly contributes to making that person into a particular, distinctive self.   Some features, like having hair, may be too unremarkable and ubiquitous to contribute to defining that person’s character; while other features might be excluded as isolated accidents of history, or temporary conditions imposed from without.  Further, among those features that are relevantly character-determining, some are more important than others, either in the sense of being more prominent or notable or else in explaining more of that person’s further features.  The ‘norm for selfhood’ is not that one eventually satisfies a longstanding “quest,” but rather that these structures of relative prominence and centrality interact to produce a coherent organizational gestalt, so that the person’s characteristics as a whole hang together in an intuitively comprehensible, even aesthetically pleasing way.

A character is also like a narrative in that it allows for diversity of interpretation about how one selects, explains, and structures this collection of disparate elements into a unified whole.  Most contemporary proponents of the narrative model take the diversity of interpretation very seriously, by assigning an essential role to the choice of how to interpret oneself in constituting the very self who is interpreted.  The character conception could simply follow the narrative model here.  But then it too would lose the resources to satisfy the Reality Constraint: to distinguish epistemically respectable self-interpretations from self-deluding ones.  A better option is to say that a self just is a distinctive, structured complex of dispositions, memories, interests, and commitments; and that one’s self-interpretation is a distinct issue from one’s actual self.  We can also say that self-interpretations are subject to the usual standards for judging theories: they should explain as wide a range of features as possible, as simply as possible.  However, this still leaves considerable room for legitimate interpretive differences.  And these differences make an especially important difference in guiding one’s choice of future actions—indeed, often the choice among distinct courses of action just is a choice about which self-interpretation one wants to embrace and actualize.

Because it appeals to a much more flexible organizational structure, and because it drops the narrative account’s radical conflation of interpreter and interpreted, the character conception fares considerably better with respect to each of the objections raised in §5.  First and most obviously, by dropping the narrative account’s focus on temporal conclusions, the character conception does not hold make the identity of a self at each moment hostage to the tale ultimately told by a retrospective narrator.  Instead, a self can simply be who she is at every moment; and she can know who she is at each moment by reflecting on the commitments and memories that matter most to her, and by investigating how well her actual habits cohere with that self-conception.

Second, by dropping the narrative account’s focus on teleological ends, the character conception avoids having to conceive of selves as split between a knowing teller and an engaged protagonist.  Instead, a self can be simultaneously and integrally a center of interpretation and a locus of agency.  We cannot abandon all commitment to the self as an integrated, actively self-promoting entity, on pain of abandoning the notion of selfhood altogether.  Nor, contra Dennett, Bickle and company, should we abandon this commitment: however it happens, people often do make and carry out decisions, and do display a stable core of behavioral dispositions, such that others can predict and hold them responsible for their actions at least over moderate periods of time.  However, where the narrative model is forced to posit a implausible executive “self-in-charge,” the character conception can hold that an individual’s past and future behaviors are her ‘own’ so long as they can be interpreted as arising out of and fitting in in an intuitively coherent way with her other dispositions, commitments and habits.

Indeed, on this view, contra Dennett, selves are not fictional even in the fairly unobjectionable sense of being “just an abstraction.”58 Rather, as Owen Flanagan says, “one’s self exists as a complex dispositional structure in the brain and is often involved in structuring experience, albeit unconsciously.”59 Further, one’s primary task in achieving selfhood is not to attain some far-off goal, but to make oneself into an ever more coherent, rich, and aesthetically pleasing whole, by gradually training one’s brain and body into a certain style or “texture of being.”  As Nietzsche puts it:

One thing is needful.—To ‘give style’  to one’s character—a great and rare art! It is practiced by those who survey all the strengths and weaknesses of their nature and then fit them into an artistic plan until every one of them appears as art and reason and even weaknesses delight the eye. Here a large mass of second nature has been added; there a piece of original nature has been removed — both times through long practice and daily work at it.…In the end, when the work is finished, it becomes evident how the constraint of a single taste governed and formed everything large and small. Whether this taste was good or bad is less important than one might suppose, if only it was a single taste!60

Although there is much in Nietzsche’s broader account of self-actualization that I reject, I think his emphasis here on the aesthetic dimension of selves, and on the way in which we make ourselves organically and piecemeal, through the slow accumulation and dissipation of small habits, is illuminating.

The third advantage of the character conception is that, because it employs a criterion of coherence that is so much more flexible than that employed by the narrative conception, it can encompass a much wider range of selves, including ‘Episodics’ and those whose lives are relatively static.  For some selves, a single, overarching feature—being a Poet of Nature, or the town doctor, say—may be highly central, explaining nearly everything about them.  But for most selves, multiple overlapping but distinct strands of centrality will collect subsets of habits, commitments, and memories into kernels of personality.  Similarly, the character conception can also make sense of people who lack the cognitive resources to form explicit long-range goals or narratives.  The narrative model follows philosophical tradition in making the self a highly reflective interpreter; but many individuals who plausibly have selves lack such a high level of self-awareness.  On the character conception, having a self merely requires having a stable, coherent “self-presentation”61: a way of presenting ourselves to, and more generally engaging with and responding to, the world.  Of course, most of us also have a fairly robust self-representation.  However, according to the character conception this higher-order representation is just one element within a larger package, albeit one with an especially significant influence on our future actions and development.62

Finally, shifting away from self-representation through narrative and toward self-presentation as character also allows us to address the worries about narrative reliability.  It is entirely plausible, even likely, that other people represent at least some aspects of our character more accurately than we do. Further, as we saw, our present memories are often unreliable guides to our original experiences.  On the narrative model, such inaccuracies threaten the very existence of selves.  On the character conception, by contrast, we are free to appeal to other people’s reports, along with any other epistemically respectable source of evidence, in order to probe the accuracy of our self-representations.  Self-interpretation thus becomes a kind of “personal science,” which is “directed toward the evolution and preservation of a meaningful sense of self.”63 And when we do discover gaps between our self-presentation and self-representation, we can address them either by training ourselves into habits and commitments that we like better, or else by reconciling ourselves to a new theory of who we are.

7. Conclusion

I have focused on narrative theories of the self, not because I believe they are utterly wrong-headed, but because I think they are deeply natural, and because their failure points us in a better direction.  Narratives are indeed a crucial tool by which many of us make sense of our lives.  The problem comes in identifying selves too directly with the lives they live.  If we drop the insistence on life-long autobiographies in favor of many short overlapping stories, we can hew more closely to the role narratives typically play in everyday self-representations; but then we also stand in need of a new criterion for unifying those stories into a coherent self.

I have also focused primarily on the specter of fictionalism: the risk that the selves generated by narratives will lack epistemic respectability.  A related risk is that the narrative theory will lead us to devalue lives and selves that don’t exemplify a compelling narrative structure, and specifically the model of a romantic hero embarked on a grand quest.  At least some of us might end up achieving richer, more complex selves if we asked ourselves, not what “office upon earth” has been fixed for us, but how to sculpt a coherent, pleasing unity from the contradictory impulses we encounter competing within ourselves.  As Wordsworth says, “There is a dark/ Inscrutable workmanship that reconciles / Discordant elements, makes them cling together / In one society” (I.341-4). But where Wordsworth invoked Nature to engineer this reconciliation for him, we must shoulder that burden for ourselves, in the here and now.64