I

Many readers first encounter Anne Carson in pieces. Tumblr users post lyrical excerpts of “The Glass Essay” (1995) beneath portraits of anemic young women; the now-defunct @carsonbot tweeted her most brooding lines every two hours; and even some instructors assign only the novel-in-verse of Autobiography of Red (1998), omitting the appendixes. Often left out are the scholarly paratexts that preoccupy Carson’s critics, though some would prefer that “teacher” Carson stay out (Chiasson). Another reviewer dismisses the “serious scholarly commentaries, introductions, footnotes, appendices, postscripts, and even mini-interviews” in Carson’s collections as mere explanations of her poetry: “It’s as if Gertrude Stein, worrying about the effect her hermetic poems will have on some future reader in Australia, composed her own Cliffs Notes to make sure she was understood” (Simic). In other words, Carson’s supposedly CliffsNotes-like self-criticism prematurely resolves the difficulty of her otherwise Steinian verse.

But if CliffsNotes had a bizarro modernist counterpart, surely it would be Stein’s lectures. Think of “Composition as Explanation” (1926), an elliptical self-commentary that deploys conventional explanatory techniques that performatively fail to account for her writing. In what follows, I show how Stein shapes Carson’s critical impulse for interpretive possibilities beyond explanation. As Stein’s lectures do, Carson’s work so often challenges the normative assumption that criticism explains poetry. Her Autobiography of Red may appear to be a critical edition of its novel-in-verse because the book’s meat is thickly marbled with paratext. However, Autobiography of Red’s paratexts simply do not explain the novel-in-verse in the way CliffsNotes explain the Iliad, the Norton Critical Edition explains The Waste Land, or Helen Vendler explains Shakespeare’s sonnets. Typically, critics and scholars explain poems through close reading, theoretical grounding, or literary-historical context—strategies that teach students and colleagues to understand a poem in a certain way. That is not what Autobiography of Red does: reading its critical apparatus would be a poor way to prepare for a quiz on the novel-in-verse.

As Elizabeth Sarah Coles argues, Carson’s works purposefully “err from established modes of scholarly writing” to create “new acts of exchange between the critical gesture and its objects of study” (Anne Carson 35). These exchanges between poetry and criticism are more than merely formal and constitute what Coles elsewhere describes as “a deliberate and fully self-conscious form of intimacy” between Carson and her objects of study (“Sacred Object” 133). For example, the introduction to Autobiography of Red situates the ancient Greek lyric poet Stesichoros as a stepping stone between Homer and Stein. In so doing, Carson offers a new way of reading the “papyrus scraps” of his Geryoneis (and her translation) as Steinian modernist fragments. Carson translates Stein’s writing style into a way of reading the ancient fragments, suggesting they (like Stein) “withhold as much as they tell” (6). This interpretive move resists explanation in favor of critical intimacy, drawing the ancients and the moderns into relation.

In this essay, I enlist object relations psychoanalysis to describe Carson’s imaginative relationship with Stein, a relationship through which Carson marshals a ragtag critical method. Such a relational approach to literature is common among Carson’s literary works. As Reena Sastri observes, “Carson’s work brings to light how reading and inventive response take place in the space between two people,” and her writerly encounters with texts, performances, and visual artworks are “fundamentally collaborative” (308). The use of object relations in describing Carson’s alternative to explanation recalls Eve Kosofsky Sedgwick’s “reparative reading,” grounded in the psychoanalyst Melanie Klein’s theory that our relationships with our caregivers (known as “object relations”) form our psychic structures. Others have shared in Sedgwick’s object relations approach to literary interpretation; Coles, for instance, draws on D. W. Winnicott’s distinction between “finding” and “creating” objects in the external world to illuminate how certain critical readings may “betray, denature or deny” what they find in literary works (“D. W. Winnicott” 5). For my purposes, the psychoanalyst Wilfred Bion’s theory of maternal reverie is useful in describing the mode of collaborative, associative thinking Carson promotes as an alternative to critical explanation. I show how Carson’s critical gestures teach readers to unsettle meaning by pointing to interpretive difficulty without resolving it—a valuable practice for teaching students to attend to textual difficulty in an era where attention is at a premium. I argue that Carson’s mode of reading Stein incorporates Stein’s own sui generis concepts of “identity” and “likeness.” This mode of reading, which I call “reading in reverie,” offers an alternative to critical readings focused on explanation and instead encourages a receptivity that tolerates uncertainty and allows meaning to emerge through provisional linking. Carson models how meaning might emerge through a collaborative, relational engagement with texts—much like the analyst-analysand relationship imagined by post-Kleinian psychoanalysts.

Carson’s conception of the relationship between poetry and criticism distinguishes her from other writers who use Stein as a guidepost—namely, from Language poets, who tend to argue that distinctions between poetry and criticism ought to be dismantled. Ron Silliman once proposed that contemporary literary criticism is “that self-consciousness wch [sic] formerly had been part of the poem itself” but now had become “interned” in the university system (157, 158). Many Language poets sought to restore this self-conscious critical function to poetry itself. Charles Bernstein’s “The Artifice of Absorption” (1987) is a paradigmatic example of a poetic-critical text in the Language strain. To the contrary, Carson’s work retains a distinction between poetry and criticism to show how these genres relate to each other. Whereas Language collapses poetry and criticism, and some of Carson’s critics insist that these two modes be kept separate, Carson arrives at the differences between poetry and criticism by enlivening their similarities. Language poets are perhaps more faithful inheritors of Stein, who preferred compositional unity to introductions and commentaries of the kind Carson includes in her collections. Carson subjects Stein to the same willful misreadings as she does the ancients, so in Carson we as readers find new Steins.

Carson’s Stein figures prominently in Float (2016), which collects twenty-two chapbooks falling loosely from an acetate sleeve. Float’s chapbooks have no page numbers and no particular order: “Reading,” its back cover promises, “can be freefall.” We can take this promise literally—Float’s chapbooks slip easily from their sleeve over the side of my desk. These slim, bluish volumes claim a variety of genres, both critical and creative. You may recognize lectures, lyrical and scholarly essays, translations, drama in verse and prose, and a few poetic genres, including a cycle of sonnets. We meet Carson’s Steins in three chapbooks—“Lecture on Pronouns in the Form of Fifteen Sonnets,” a “how-to” essay, and a staged piece titled “Uncle Falling: A Pair of Lyric Lectures with Shared Chorus.” The first chapbook shows how critical genres (such as the lecture) and poetic genres (such as the sonnet) depend on relationships of address created by the pronoun—that little piece of grammar that Stein so loved. Stein’s concept of “identity”—that is, one in relation to others—can help us as readers trace the lines Carson draws between genre, relationality, and identity. In my next exhibit, the “how-to” essay, Carson toys with a related term, likeness, drawing out the “like” at its root to mean both “resemblance” and “enjoyment.” By teaching us to see likenesses (or similarities) in Stein’s work, Carson also teaches us how to like (or enjoy) Stein’s writing.

But Stein can be difficult to like. To bring us around, Carson focuses our attention on how we feel when reading Stein. Carson’s attitude toward Stein is therefore reverie in practice, offering an alternative to explanatory criticism. Instead, we learn to tolerate a frustrating lack of coherence, waiting for the last possible moment to impose our own interpretation. Float takes up poetic and critical genres—the sonnet, the lecture—to expose the relationships each makes possible between writer and reader. In Float, genre works like the analytic situation. It is a frame within which analyst and analysand make connections and, ultimately, transform their thinking.

II

The chapbook titled “Possessive Used as Drink (Me): A Lecture on Pronouns in the Form of Fifteen Sonnets” compares lectures and sonnets according to their use of pronouns of address. In the odd title, “Possessive” refers to pronouns denoting ownership, and “Drink (Me)” alludes to the tag on the bottle of liquid that Lewis Carroll’s Alice drinks at the outset of her adventures in Wonderland, shrinking her to the size of a blade of grass. By transforming her body, the magical drink also transforms Alice’s relation to the world, precipitating the identity crises that bedevil her in Wonderland. Carson’s lecture-sonnets teach us that pronouns wield similarly transformative powers. Pronouns shape relationships among speakers and the people and things in their world. In poetry, lectures, and everyday life, pronouns forge relationships of possession, address, deixis, gender, and so on.

Pronouns of address mark both sonnets and lectures. Sonnets usually address a singular other, while lectures address a public audience; both genres seek transformation in the addressee. Shakespeare’s Fair Youth sonnets, which Carson cites in her lecture, stage an I-thou relationship between speaker and beloved. A lecturer’s “you” is not so intimate. A lecturer addresses the public to stimulate their attention. In her lecture titled “Poetry and Grammar,” for example, Stein invites her audience into her central inquiry by means of address: “What is poetry and if you know what poetry is what is prose” (313). However, the disappearance of thou in English collapsed the intimate pronoun of the sonnet and the public pronoun of the lecture into the very same you. Carson’s lecture-sonnets in Float sometimes mourn this loss of the intimate register: “Drop’t Sonnet” tenderly addresses a beloved whose affection “cried out to be addressed / as / thine” (lines 31–33). Yet the lack of an intimate register authorizes the lecture-sonnets to unsettle conventional relationships of address in both poetic and critical modes as the lecture-sonnets toggle between the intimacy of the sonnet and the formality of the lecture.

Carson’s lecture-sonnets playfully expand queer pronoun practices to include pronouns of address. Identifying oneself with gendered pronouns is an expression of gender identity. Because gendered pronouns are in the third person, the now-common queer practice of sharing preferred pronouns creates a social triangle: I tell you my pronouns are “she/her,” but you use these pronouns only when you talk about me with a third person. In recent years, nonbinary people have experimented with neopronouns, using a favorite noun (such as kitten/kittens/kittenself) instead of a third-person pronoun. The queer theater artist Taylor Mac even uses the name “judy” as judy’s third-person pronoun. If third-person pronouns have the power to change perceptions of gender, Carson imagines that second-person pronouns can reform relationships between persons.

One lecture-sonnet, for instance, proposes an oddly specific new pronoun of address—a pronoun specifically of “addressing Gertrude Stein.” It is a fitting tribute to invent a neopronoun just for addressing Stein, who claims to prefer pronouns over nouns because they have a “greater possibility of being something” than plain old nouns do (“Poetry and Grammar” 316). Here is Carson’s lecture-sonnet, titled “Sonnet of Addressing Gertrude Stein,” in full:

Carson presents a neopronoun for addressing Gertrude Stein. Such specificity is usually the work of names, not pronouns. And what is the pronoun? “Here”? The colon after “with”? The phrase “a dog”? Or the full phrase “a dog you’ve never had before has died”? Multiple readings are available, but I focus on the latter two, in which Carson converts a noun phrase (in my reading, “a dog”) into a pronoun of address.

So why might Carson choose “a dog” to stand in for Stein—or, more precisely, for an address to Stein? The lecture-sonnet tweaks a phrase from Stein’s 1931 prose poem “Sentences and Paragraphs.” Stein’s original phrase reads, “A dog which you have never had before has sighed” (23; my emphasis). In the middle period of Stein’s writing, “a dog” invokes what Stein called “identity,” or the person defined by ongoing relationships with others. Take for example her better-known aphorism, “I am I because my little dog knows me” (Geographical History 112). “I” am defined through my relationship with my “little dog” and his continuously forming knowledge of “me.” “I” depend on my existence in my little dog’s memory. According to Stein, “identity” is the ordinary way that we experience ourselves going about “the business of living” (“What Are Masterpieces” 358). In Carson’s sonnet-lecture, “a dog” fulfills the function that “thou” once did in English sonnets as an intimate form of address rather than the formal “you.” “Gertrude Stein” moves from the “I” in “I am I” to the place of “a dog”—an intimate other that makes the speaker’s identity meaningful.

Identity’s complement is “entity,” Stein’s term for the self-in-itself and “not in relation” (“What Are Masterpieces” 358). Stein associates pronouns and proper names with entity and ordinary nouns with identity. Shedding light on the identity/entity distinction are scholars who situate Stein’s writing in the context of contemporary Anglo-American philosophy of reference. Jennifer Ashton explains how Stein prefers proper names to nouns because, whereas names refer to a particular person or thing, nouns refer to a class of things and consequently suffer diminished meaning (see ch. 2). For example, one might refer to Stein’s beloved pet as “Basket” or merely a “dog.” “Basket” calls up the dog’s particularity, but a “dog” just invokes Basket as one of many dogs.Footnote ¹ Stein likes pronouns just as much as names because pronouns “wear their referential function on their sleeve,” as Marjorie Levinson puts it (205). Referring to Basket as “you” or “him” keeps him near, too. Most damningly for Stein, nouns also depend on a subject to identify them, so they languish in subjective memory. Alternatively, the entities indicated by names and pronouns exist independent of a subject.

Carson raises Stein’s identity/entity distinction only to trouble it when she declares that an ordinary noun phrase (“a dog”) can be a pronoun—following queer artists like Mac in transforming names and nouns into pronouns. Carson’s pronouns do more than stand in for a particular thing; they enthusiastically create relations among speakers and their referents. When I use a second-person pronoun, for example, I create a relationship between us: “you” become “you” because I address you. And like nouns, gendered pronouns also express the relation of someone to a class; I mark Stein as a woman when I refer to her as “she.” If we read the lecture-sonnet to suggest that “a dog” could be a pronoun for addressing Stein, then perhaps “a dog” acts as a pronoun in Stein’s writing too, standing in for an intimate other.

In its sestet, Carson’s lecture-sonnet turns doubly away from identity. First, the pronoun “which” evaporates into a stanza break and the lecture-sonnet’s volta. Then, Carson modifies Stein’s original phrase: we find not only that this is “a dog you’ve never had before” but also that it has “died,” not just “sighed.” After the volta, it seems that to address and to be identified is to lose that which “you’ve never had before.” Identity depends on a dual absence: a relation between two subjects, neither of whom has content in itself but only in relation. The sonnet’s I-thou relation conceals the thinness of both subjects alone. In this sestet, I hear Shakespeare’s declaration in sonnet 13: “Oh, that you were yourself! but, love, you are / No longer yours than yourself here live” (lines 1–2). The fair youth cannot exist only unto himself forever and can continue to live only if he procreates. The only way to create an identity is to have and make relations. I have no identity without a little dog, a fair youth, a son, or an audience to listen to my lecture.

III

Let’s review: for Stein, identity is a oneness made between two. Stein’s second portrait of Picasso, titled “If I Told Him: A Completed Portrait of Picasso,” expresses the difficulty of rendering anyone’s identity in words: “Exact resemblance to exact resemblance the exact resemblance as exact as a resemblance, exactly as resembling, exactly resembling, exactly in resemblance exactly a resemblance, exactly and resemblance” (506). It is the job of a portraitist to render an “exact resemblance” of the subject, but in Stein’s portrait, the phrase degrades into oxymoron as it repeats. A “resemblance” is a relationship in which two people or things are seen to share certain properties, so if two people (say, Picasso and Napoleon) exactly resemble each other, that would mean that Picasso and Napoleon share all the same properties. If Picasso and Napoleon share all the same properties, they are identical—the very same person. However, if Picasso and Napoleon are the same person, it would not make sense to say that they resemble each other, because a resemblance is a relationship between two things. Therefore, an “exact resemblance” cannot exist. A portrait is a mere resemblance; its subject’s identity is disclosed in the portrait’s failure to be exact.

Another of Float’s chapbooks, “How to Like ‘If I Told Him: A Completed Portrait of Picasso’ by Gertrude Stein,” teaches us how to read Stein’s portrait of Picasso. The essay’s title puns on the double meaning of “like,” which can denote resemblance (as a preposition) or enjoyment (as a verb). “To complete a portrait,” Carson writes, “is to learn to like your likeness.” The converse is true for readers of Stein’s portrait: only when we understand how to like (that is, enjoy) Stein’s portrait can we appreciate the dynamics of likeness (that is, resemblance). The essay begins with an attentive close reading of the portrait. Carson’s essay does not explain identity, resemblance, and likeness in Stein’s portrait but rather models the attitude of reverie that readers must cultivate to see resemblances.

To illustrate the compositional principle of likeness, Carson figures the portrait maker (here, Stein) as a mirror. Yet if Stein is a mirror, she is a strange one. Picasso looks into her and finds Napoleon; Napoleon looks into her and finds Picasso. A mirror promises the exact resemblance of a reflection, but this one yields only likeness. For Carson, the second “complete” portrait of Picasso surpasses the first precisely because it privileges mere likeness over exactness. In comparing Picasso with Napoleon, Stein’s second portrait renders her subject as knowable only through his likeness to another. Stein’s first portrait had done the opposite in presenting Picasso as a solitary genius—as “one who was working”—in entity rather than identity (Stein, “Picasso” 283). Hence the pun in “If I told him would he like it” (“If” 506): Stein wonders whether Picasso might like the second portrait better because it better renders his likeness.

Confronted with so much likeness, readers of the portrait may “fear that we have fallen into the mirror,” Carson writes. But for Carson and Stein, a likeness contains its own undoing—likeness always swerves toward difference, Carson assures us. Melissa Feuerstein highlights how Carson conceptualizes metaphor (the recognition of likeness) as a form of productive error, noting that while “metaphor establishes equivalences … the equivalences themselves are essentially false, its ‘mistakes’ its truth” (227). The error of metaphor creates a fault line in the “plane surface / of ordinary language,” Carson writes in Men in the Off Hours (2000), and the mind can “experience itself // in the act of making a mistake” (30). The identity paradox comes full circle when we recognize that Stein’s mirror does not merely reflect Picasso and Napoleon to each other—it also reflects Stein herself as she sees her own faulty thinking in their resemblance. In crafting their likenesses, Stein asserts her identity as a portraitist, an identity defined by her relationship to her subjects.

Carson reassures us that their “exact resemblance” actually keeps Picasso and Napoleon apart as their likeness “swerves toward difference.” And what is the difference between Picasso and Napoleon? It is a funny question because Stein does not tell us how they are similar. Is it that they are both “famous gentlemen,” as Carson puts it? Famously short men? Famously short and consequently egomaniacal men? As for differences, Stein does not mention the obvious: Picasso was a twentieth-century Spanish painter, and Napoleon I was a nineteenth-century French general. The sole difference she points to is that of sequence. Stein writes, “Who came first, Napoleon the first” (“If” 507)—in other words, Picasso succeeds Napoleon I. Except for sequence, Stein jokes, we may be unable to distinguish between these two men.

In a work from around the same period, “Patriarchal Poetry” (1927), Stein satirizes the kind of poetic canon making that T. S. Eliot proposes in “Tradition and the Individual Talent”—that the poet’s order of arrival on the literary scene determines the meaning of his poetry. She makes strange the idea that the order in which things appear determines their meaning. It turns out that patriarchal poetry, like empire, obtains meaning through the succession of sons and the search for “their origin and their history” (“Patriarchal Poetry” 576). Indeed, the sequence of poetic sons conventionally explains the content of a poem: “Patriarchal poetry or indeed an explanation” (“Patriarchal Poetry” 583). As an alternative, perhaps the poet or portraitist must make connections and see resemblances among persons. Punning on the portrait’s repeated phrase, “I land,” Carson writes, “No gentleman is an island, her mirror implies, at least if Gertrude Stein happens to be sailing by.” In this portrait, identities are formed by relation, and the portraitist promises to reveal these likenesses to us.

Such an imaginative approach has some precedent when it comes to reading Stein. Johanna Winant has pointed out that no other modernist has received quite so much attention in the form of attempts at explaining her poems. Unlike criticism of other modernist poets, which can relieve some of the difficulty of reading them, attempts to explain the difficulty of Stein’s poems often end in frustration. Winant argues that this is partly because Stein’s writing sought a “theorization of explanation” (96), a concept that Linda Voris’s writing on Stein’s “An Elucidation” (1927) has also explored. According to Winant, “the resulting literary difficulty can be clarified but not solved” in criticism of Stein’s work (97). Stein’s most difficult writing produces a concoction of “astonishment and fatigue” powerful enough to paralyze readers’ interpretive capacity, as Sianne Ngai has argued (253). Yet many contemporary American poets have written about their attempts to manage the frustration of reading Gertrude Stein—not only Carson but also Wayne Koestenbaum, Tracie Morris, Anne Waldman, and John Ashbery. The poets advise fellow readers of the enormous “patience” that reading Stein’s “wild experiment” demands (Waldman). Koestenbaum admits that Stein’s writing is difficult enough to remember, let alone understand. Still, Koestenbaum makes the case that reading Stein is both clarifying and “nice”: he compares it to “one long rinsing or cleansing operation” (305). Ashbery notes that the “almost physical pain” that Stanzas in Meditation (1956) inflicts is worth the “inrush of clarity” that follows (252). Like an ice bath, Stein’s writing strengthens us as it challenges us to tolerate her difficulty.

To describe this attitude, which I call “reverie,” I recruit object relations psychoanalysis, pioneered by Melanie Klein, and particularly her intellectual heir Wilfred Bion’s theory of thinking. Literary scholars including Mary Jacobus, Jacqueline Rose, Rudi Vermote, and Adam Frank have turned to Klein and Bion to locate a distinctive conception of “thinking” in Anglo-American poetic modernism, a kind of thinking that is relational and affective rather than asocial and detached. Thinking first develops in our relationship with a caregiver who receives and processes our emotional experiences with an attitude of relaxed attention that Bion calls “reverie.” Though Bion’s paper “A Theory of Thinking” (1962) postdates Stein’s death by sixteen years, Frank argues that Stein shares with Bion “a commitment to investigating the emotional conditions for the emergence of new knowledge” (106). I argue here that Carson teaches us how to read Stein in reverie to promote this kind of emotionally charged thinking. Additionally, Bion’s theory of thinking is inspired by modern life and modernist literature, including the work of his analysand Samuel Beckett.Footnote ² Bion’s theory of thinking is uniquely suited to describe the kind of attention that reading Stein’s modernist writing demands and the habits of reading that Carson hopes to show her readers.

Klein is known for her work in child psychoanalysis and her departure from the Freudian focus on the Oedipal complex, especially as developed by Klein’s contemporary and rival Anna Freud after her father’s death.Footnote ³ For my purposes, the most relevant feature of Klein’s work is her explanation of how a person’s “objects” (that is, a significant person or part of a person who is subjected to libidinal drives) are internalized in the mind. A person’s libidinal relationships with these internalized objects form that person’s personality—hence the term object relations psychoanalysis. Klein dramatizes the initial absorption of objects in the scene of breastfeeding, where the primary players are mother and baby. In this drama, the Kleinian baby understands the world only in terms of good and bad in relation to its self-preservation needs: hunger is bad, fullness is good. For a helpless baby, hunger is an intolerable feeling equivalent to fear of death. The baby cries to project its overwhelming feeling into its mother and communicate its distress to the mother by making the mother distressed, too. When one succeeds in projecting an unwanted feeling into another person, Klein calls this dynamic “projective identification” in her 1946 paper “Notes on Some Schizoid Mechanisms.” Once the mother has received her baby’s projections, the mother can either understand its fear and nurse her baby, or refuse her baby’s projections and decline to nurse her baby. The baby psychically splits its mother accordingly into two objects: both the Madonna-like “good breast,” which nurses the baby with good milk, and the denying “bad breast,” which seems to inflict hunger on the baby (Klein 100, 101). The baby “introjects” (that is, internalizes in its mind) good and bad objects just as it takes the mother’s milk inside of its stomach. But the Kleinian mother in this scene is just an ordinary woman who attends to her baby’s needs but sometimes falls asleep, runs out of milk, attends to a sibling, or picks up the phone in another room. Maturation, for Klein, occurs when “loved and hated aspects of the mother are no longer felt to be so widely separated”—that is, when the infant realizes that good mother and bad mother are one real person (105).

Bion’s theory of thinking proceeds from the Kleinian drama of breastfeeding. Bion sees the frustrated impulse to feed as the baby’s first “thought,” for which an apparatus of thinking “has to be called into existence” (“Theory” 111). Unlike Klein, who characterized projective identification as an always primitive defense mechanism, Bion writes that projective identification can be employed as a “realistic activity” that can be “reasonably calculated to arouse in the mother feelings of which the baby wishes to be rid” (114). In other words, projective identification can be communicative. Unable to speak, a baby can arouse feelings of frustration in its mother when it can no longer tolerate these feelings itself, and the baby’s mother, if she is a “well-balanced” one, “can accept these and respond therapeutically” by feeding the baby and helping it manage its fears (114). As Vermote writes, “When the mother is able to tolerate the infant’s projections, the infant can use the mother as an instrument for psychic processing of sense-data, as it does not yet possess this ability. In this way, the projections may become psychic and the personality develops consciousness” (76). In other words, our earliest conscious thinking is conducted together with our mothers—not alone.

But how can a mother “tolerate the infant’s projections?” After all, anyone who has been around a crying baby knows it is not easy to remain calm in the face of such intense projection. Bion defines a “well-balanced” mother as one with the capacity for “reverie”—that is, a receptive mental state characterized by relaxed attention (“Theory” 116). Maternal reverie creates the conditions for what Bion describes throughout Learning from Experience (1962) as the relationship between “[c]ontainer and contained” (113), a dynamic in which the mother’s mind (the container) receives and processes the infant’s projections (the contained). Lacking an “apparatus for thinking,” a baby nonetheless has thoughts, defined as unsatisfied needs (“Theory” 112). A mother receives her baby’s needs and meets them with satisfactions. The paradigmatic scene is breastfeeding: a mother answers her baby’s cries with the satisfaction of breast milk. In so doing, she thinks her baby’s thoughts and teaches the baby how to think, according to Bion. This relational process develops into what Learning from Experience names “alpha function,” or our capacity to transform our raw, unprocessed experiences (“beta elements”) into digestible, thinkable forms (“alpha elements”). Through this transformation, the infant gradually develops its own capacity for symbolic thinking. This earliest conversation plays out in the language of need: a requirement for thinking, unique to Bion’s theory, is the tolerance of frustration. As Coles writes, Carson’s criticism shores up “the value … of recognizing what is not grasped, not resolved, not wanted, not yet even read, in the object of interpretation” (“Sacred Object” 142). In this sense, Carson’s critical attention resembles the maternal reverie: Carson receives the work of another—difficult, unresolved, fragmentary—without attempting to master or explain it.

Thinkers must tolerate the frustration of uncertainty to forge new links instead of reaching for prefabricated interpretations. For Bion, “reverie” also describes the relaxed attention with which a psychoanalyst ought to receive a patient’s train of associations and projections with minimal intrusion of “judgment” (Learning 109). The reverie allows meaning to proliferate in the patient’s associations, out of which patient and analyst can together make connections among recurring elements in the associations. To make such connections is, in Bion’s vocabulary, to “link.” Linking names the connective activity that joins one thought to another, and it extends from the infant’s earliest bonds to “the most sophisticated forms of verbal communication and the arts” (“Attacks” 108). Psychosis entails destructive attacks on this linking function (“Attacks” 93). Unable to tolerate frustration, the psychotic patient assaults the connections that would bind experience into thought, leaving fragments that cannot be thought at all. The analyst must maintain relaxed attention when a patient’s associations do not yet make sense. If the analyst gives in to the “irritable searching after certainty” and makes forced interpretations (“Grid” 11), then the patient may suffer from the premature foreclosure of meaning. Bion alludes to John Keats’s famous December 1817 letter to his brothers, in which the poet describes “Negative Capability, that is, when a man is capable of being in uncertainties, mysteries, doubts, without any irritable reaching after fact and reason” (57). Equally important for Bion’s analyst is the ability to dispense with “memory and desire,” a phrase Bion used to describe how preconceptions about the patient’s past or expectations about therapeutic outcomes can interfere with the analyst’s capacity for reverie. These mental activities threaten to interrupt the analytic reverie like “lilacs out of the dead land” (Eliot, Waste Land, line 2). Bion’s analyst is concerned not with what happened or with what will happen but only with “what is happening” (Bion, “Notes” 380).

Merely reading Stein’s fragmentary writing necessitates a capacity for reverie, because it feels to many readers like containing a baby’s fragmentary projections. Waldman, who characterizes Stanzas in Meditation as the writing of a “willful child,” taught a class at Naropa University called Mind Gertrude Stein, which calls up both “mind” as the seat of thought and “mind” as looking after a child. Morris calls her critical-creative response to Stein’s Tender Buttons (1914) a “handholding,” as though the poet would help little Stein cross the street. Ashbery characterizes Stein’s writing as “a child’s incantation against grownups” (250). Koestenbaum, pointing out that Carl Van Vechten and Alice Toklas affectionately nicknamed Stein “Baby Woojums,” claims that “Stein’s paradigm of the writer was the baby” (299). The “author as infant” model frees Koestenbaum, as a reader of Stein, to “interpret or not to interpret her as I see fit” (298). Readers of Stein contain her fragments, and in Bion’s model, they “think” the fragments that her writing presents by linking them together.

In “How to Like ‘If I Told Him,’” Carson models reading in reverie for those who would like to read Stein, as Carson’s reading of Stein permits patterns to emerge in her fragmentary portrait. This way of reading recalls what Bion describes as “kaleidoscopic change” in the analyst’s reverie of the patient’s material, as elements surface together repeatedly in different configurations that Bion calls “constant conjunctions” (“Grid” 11). Carson’s essay moves nimbly from the portrait’s title to its final recitation of “what history teaches,” linking each new element of the poem with the last. The portrait’s nouns—Napoleon, Picasso, kings and queens, shutters, mirrors, and trains—become playthings in Carson’s reverie. If Stein is a mirror, Carson puts her in a kaleidoscope. Nouns shift around like colored pieces of glass, creating points of contact that depart soon after they arrive.

Abruptly, Carson writes that the portrait’s meaning deserts her when Stein repeats “train”:

The portrait zooms “out of sight” like a train that “switches tracks.” Yet Carson patiently “ride[s] along a while in good faith.” Carson’s essay not only teaches us about particular meanings of Stein’s portrait but also trains us to hold the attitude that we ought to take when the meaning fails to let us board—an attitude of openness and unconditional positive regard, like an analyst listening to a patient’s associations (or you could call it a “train of thought”). In her reverie Carson sees resemblances, makes links, and finally identifies with Stein: “If she told us would we like it. Yes, we would. Or at least I do. To complete a portrait is to learn to like your likeness albeit miraculously,” the essay concludes.

What makes Carson’s reverie so compelling is its double function: Carson contains simultaneously Stein’s fragmentary baby mind and new readers of Stein who (like new mothers) might feel overwhelmed. In this recursive structure, Carson metabolizes Stein’s projections while offering a containing model for readers—who are themselves trying to learn how to relate to difficulty. In this way, the essay performs a double act of holding as it contains Stein and it contains us. These layered container-contained relationships characterize Carson’s chapbooks across Float, which draws on pedagogical genres—lectures, instructional fragments, even scripts—to stage the conditions under which thinking might begin. When meaning temporarily departs during the dizzying “train” repetitions of “If I Told Him,” Carson does not explain the departure or rescue the reader from it. Instead, she models how to stay: not through anxious grasping for interpretation, but by waiting at the station, so to speak. She absorbs Stein’s resistance and our potential frustration with it, showing that the proper response to textual difficulty is not resolution but relation. In doing so, Carson teaches us how to read her, embodying the attitude of reverie she seeks to cultivate in us. It is not only criticism but instruction in a way of being with literature.

Another chapbook, “Uncle Falling,” opens with a chorus of “[f]our seated Gertrude Steins” accompanying two lyric lectures. The stage directions have the Steins wear masks of the Picasso portrait and “laugh ‘like a beefsteak.’” Carson lifts the phrase from Mabel Dodge Luhan’s description of Stein’s laugh, a remark she had already examined in “The Gender of Sound” (1995). Stein’s laugh embodies a refusal to submit to patriarchal aesthetic norms: because they “did not know what to make of her prose,” Carson writes, Stein’s critics tend to exaggerate her “abnormal attributes” and lesbian-coded physical appearance (“Gender” 121). In this light, the famous story of Picasso’s portrait reveals its misogynistic shading. As the story goes, Picasso had difficulty painting Stein’s face, and after months of sittings, he irritably wiped it out and left for Spain, claiming that he could no longer see her face at all. Upon his return to Paris in 1906, he repainted her face before ever seeing her again, this time in the manner of an Iberian mask, an inspiration for what would soon be called cubism. It is not quite her: the nose is too straight, the eyes are too large and elliptical, the skin looks more like stone than flesh. When confronted with the opinion that Stein did not resemble her portrait, Picasso supposedly said, “she will” (Stein, Autobiography 669). Carson’s stage direction transforms the portrait back into a mask as her four Steins wear the face Picasso imposed, while laughing the beefsteak laugh her critics could not domesticate. The mask is his, but the laugh is hers.

This defiant Stein is a provocative casting choice for the chorus in a pair of staged lectures that tell two stories of Carson’s actual patriarchs—her great-uncle and her father. While telling these stories, the lecturers and the chorus call into question how literary and syntactic conventions structure family history. The first lecture, titled “Uncle Harry,” narrates the life, illness, and death of Carson’s paternal great-uncle Harry. Lecturer I knows little about Harry except that he was a hermit. Carson fills the gaps by reading in reverie, linking Harry’s life to Stein’s on the slender grounds that both were born in 1874. The second lecture, titled “Falling,” gathers ideas about falling from a range of texts—Carson’s father’s wartime diary among them. Lecturer II recounts reading, after the father’s death, the diary he kept as a prisoner of war in Pomerania—the only thing he ever wrote.

The lecturers and chorus piece together the few memories and fragmentary texts these dead men left to Carson. In both lectures, the connections can be weak and the narratives thin, since Carson seems to have little of substance to say about her patriarchs. In the first lecture, we learn that Carson’s great-uncle liked to look at a landscape, whereas Stein said, “I like a view but I like to sit with my back turned to it” (qtd. in “Uncle Falling”). In the second lecture, lists of Carson’s father’s favorite wines, Russian words, and other ephemera from the diary reveal little about the man himself. Carson praises her father’s “observant, rueful, spare” style and merely “factual” account without wartime sentimentality. Yet she resists the pious narrativizing that such a text would normally invite from a filial reader. Carson’s own family history emerges as just as fragmented and opaque as Tender Buttons. The Stein chorus helps her read these familial remains as if they were Stein’s textual fragments:

Coming from the Steins’ chattering mouths, the diary sounds like “Food” in Tender Buttons. Names of foods and wine are paired with singsong rhyming phrases without much reasoning. Carson finds “a path here into the intimate places of a man” by loosely connecting these fragments as she would the lines of a Stein portrait.

In these “lyric lectures,” Carson models reading in reverie by placing the chorus of chirping Steins in the position of the baby or analysand and the two lecturers in the position of mother or analyst. Linking is, for Carson, an exciting, “miraculous,” and “terrifically perilous” activity—it is the reward of lecturing, analyzing, mothering, reading. After the recitation of the lists, Lecturer II interrupts herself to lay out what I understand as an important compositional principle throughout Carson’s work:

Lecturer II’s description of lecturing as “linking” thoughts together resembles Bion’s theory of thinking. Thinking requires forging links, even if they are “frail” and contingent. Lecturer II justifies lectures that call attention to this aspect of thinking instead of presenting “what we call sense” or “reasoning” as the discovery of premade, necessary, or essential connections among thoughts. What is more, Carson shows this thinking as an unfolding process that requires toleration of initial frustration and uncertainty. When she writes that “the best connections are the ones that draw attention to their own frailty,” Carson describes just the “perilous” kind of thinking that Bion theorized, a kind of thinking that demands that we renounce infantile omniscience and reveals the thinking apparatus for what it is—a tenuous, diaphanous web.

Float shows us that fragile relationships among persons—what Stein would call “identity”—resemble the fragile connection between that thought and this thought. The interaction between the lecturers and the chorus of Steins dramatizes thinking as a social activity, an aspect retained from its origin in the exchange between baby and mother. Here, Stein is not poetic forebear or parent but babbling baby. Carson’s poetic-critical output emerges as Carson receives Stein’s projected material and processes it through her reverie. For Carson’s Stein, identity is not an object of thought but thinking with others. Carson habitually thinks in collaboration, and in fact has taught an “artistic collaboration” class called The Egocircus at New York University with her husband, Robert Currie (Carson, “Throwing”). The performance of the lecture introduces yet another player into the collaboration. “Uncle Falling” has been staged many times, and, as Carson herself has said, each experience is uniquely “deepening” (“Interview”). Each performance, then, is a lesson in how to read in reverie.

IV

Carson’s creative interpretation of Steinian “identity” allows her to transform relationships between critical and creative genres in her writing. Stein recognized that identity is configured in our relationships with others. In Float, Carson suggests how grammar and genre create possible identities by establishing modes of relation among speakers and addressees, writers and readers. Relating, however, requires the emotional condition that I have called “reverie,” after object relations psychoanalysis—that is, an openness to making connections, to forming identity. By modeling “reverie” in her creative-critical writing about Stein, Carson demonstrates an alternative approach to reading and interpreting Stein’s difficult, inexplicable work, such as “If I Told Him.” Reverie is a method of reading that critical writing might also adopt, modifying the relationship between, say, a lecturer and an audience. Finally, “Uncle Falling” underscores the importance of reverie as a skill outside a purely literary context: we might use the capacity for reverie that reading Stein has strengthened to tolerate the relational ambiguity that a family of origin invariably leaves in its wake.

In a recent interview, Carson recalls entering a “reverie” while translating Ancient Greek as a student (“Throwing”). Once upon a time, before digital technology dominated the classroom, students like Carson would need to look up unknown words in a lexicon, flipping pages in search of the right definition until they found it and copied it down by hand. There is a moment of not yet knowing between identifying an unknown word in a text and locating it in the lexicon. Carson calls such a moment of hesitation “reverie.” Carson complains that today’s digitized classrooms do not leave room for such an experience. The computer’s lexicon supplies students with a missing word with only a tap or click on a text. “[Y]ou don’t even get the history of the word and a chance to float around among its possible other senses,” Carson says (“Throwing”). Students may learn to read Ancient Greek this way, but they do not experience the slow, uncertain process of thinking and its imperative to remain calm in the “absent presence” of meaning.

Ordinary scenes of reading and writing have changed since Carson flipped pages in her lexicon fifty years ago. N. Katherine Hayles’s distinction between “hyper attention” and “deep attention” is useful for conceptualizing what has changed. “Deep attention” requires “a high tolerance for long focus times” on information in a single medium (187); in contrast, “[h]yper attention excels at negotiating rapidly changing environments in which multiple foci compete for attention” (188). Deep attention is generally the mode of attention we ask students to bring to the literature classroom; hyper attention is the mode of attention we use to navigate our contemporary world, noisy with the dings of our smartphones. Literature teachers cannot expect students to come prepared with deep attention but must teach this skill through explicit instruction.

As scholars and teachers of literature, we must continue to create the conditions for reverie in the classroom. The recent conversation on “aesthetic education” reminds us to instruct our students to pay deep attention to literary texts. Proponents of aesthetic education encourage scholar-teachers to bring our students’ attention to a text’s sensory and emotional features. According to Nicholas Gaskill and Kate Stanley, such features make literature what it is. Meaning inheres as much in what a literary text “offers to be felt as in what it sets out to say” (129). Gaskill and Stanley join two senses of feeling here (that is, sensation and emotion) in their claim that a literary text demands of its reader “an emotionally infused cognition that resists abstraction from the text’s sensory particulars (131). Yet if aesthetic education fails to theorize the movement between sensation and emotion, it risks fetishizing the aesthetic by treating certain affective responses as universal outcomes of formal procedures. As Michael Clune carefully argues, aesthetic judgment names “a space in which the boundaries between subject and object are reorganized so as to enable perceptions, discoveries, and actions otherwise impossible” (71). Aesthetic judgment, for Clune, necessitates a transformation of the self through sustained contact with resistance.

Relational approaches, such as Bion’s theory of thinking, can clarify how feeling begets feeling—that is, how sensing begets emoting—in the context of an object relationship. In his model, the data of raw experience must be contained, metabolized, and transformed into symbolic form before it becomes thinkable as emotional content. Crucially, this transformation first happens in the relationship between infant and caregiver. In the scene of reading, an analogous transformation goes on in the relation between reader and text: the reader notices the sensory features of the work and, with patient attention, gradually assembles them into an interpretation. The text is not merely a vessel for aesthetic experience, nor is it an inert object for analysis. Rather, it is an affectively charged, formally resistant presence with which the reader enters into an unfolding relation. Such an account of reading as relating clarifies what is at stake in ongoing debates about method. The affective quality of the relationship between reader and text is not a distraction from method but the medium through which interpretation becomes possible.

Carson models reading in reverie as she contains the “baby mind of Stein” and her readers’ doubts.Footnote ⁴ Carson does not try to translate Stein into argumentative clarity or resolve her resistance into interpretation. It would be fruitless. As Koestenbaum writes, Stein’s resistance to any “critical vocabulary that might contain her” forces readers to inhabit the perilous moment of hesitation before a thought finds its satisfaction (319). Reading Stein becomes an exercise of reading in reverie—a capacity to remain in uncertainty without reaching for explanation prematurely.

If, as Bion suggests, thinking arises through the maternal containing of frustration, then students must feel contained themselves in order to contain textual difficulty. As a matter of practice, we do our students a disservice when we expect them to produce well-formed “readings” of a text during classroom discussion. This is especially true when teaching difficult texts like Stein’s. The challenge is not interpreting her but helping students feel comfortable in uncertainty. Instead of aiming for argumentative clarity, we might encourage students to make “light, loose” connections among textual features, trusting that conversation can gradually massage frustration and observation into something like interpretation. Instructors ought to consciously contain students’ frustration upon encountering textual difficulty. When students express frustration with a difficult text, look at their response together: What specific textual elements provoke this reaction? What might this tell us about how the text operates? Teaching students to read in reverie teaches them how to think by offering the emotional conditions in which thinking begins.