Adena Rosmarin's comment is part of a wider discussion of E. H. Gombrich's idea of the role of “schema” in perceptions of art—a concept he uses to explain, among other things, how it is that observers can perceive a face in a collection of lines and dots, and, more generally, why it makes sense to say that all art begins not with nothing but with something. A snowman starts with the schema that is snow, for example, and we “work the snow and balance the shapes till we recognize a man.”Footnote ¹ And not only do we work the snow to achieve some approximation of human faces and bodies, but we find human characteristics in—or, more accurately, make characters out of—all sorts of things, living and not. Certainly, we project character onto living nonhumans, such as cats, dogs, birds, and plants, imagining that they share our feelings of love, loneliness, and desire, but we also assign character and characteristics to inanimate objects and assorted phenomena: ships, cars, and hurricanes all get names; we may talk to the toaster, or grumble at a shoelace that, we imagine, stubbornly refuses—refuses!—to untie. Thus it is hardly surprising that literary characters, intended expressly as vehicles for imaginative occupation, take on a similarly outsize quality in our psyches. Words on a page, constructions of language, they invite us to consider them as soulmates and doppelgängers; nightmare—or perhaps ideal—versions of parents, brothers, sisters, friends, and selves: the families we might have chosen if we could; the me (or not me) nobody knows.

In the volume Character, three critics—Toril Moi, Rita Felski, and Amanda Anderson—seek to turn the discussion of literary character in new directions, beginning by surveying the field as it has taken shape over many years: as far back as, in Moi's account, L. C. Knights's 1933 essay, “How Many Children Had Lady Macbeth?” Moi's essay, along with Felski's, seeks to reverse the trend that, from Knights to structuralism and beyond, has urged us to detach from our tendency toward attachment: our desire to “like” and identify with characters, to treat them as persons. Taking as her starting point literary criticism's extended pendulum-swing away from what Jonathan Culler has called the “naturalization” of literary elements such as narrators, regarding them as collections of “semes” or other kinds of linguistic effects, Moi joins many of this past decade's critics in such fields as affect and cognitive studies who have sought to move literary critics back in the direction of treating “characters as objects of identification, sources of emotional response, or agents of moral vision and behavior” (4). Knights, Moi argues, sought to define a literary-critical practice suitable for modernism, which meant doing justice to language's abstract, technical effects rather than its illusory ones, and he did so for professional reasons: to help establish the discipline of English literary studies.

Felski takes a somewhat similar approach, arguing that literary identification is not taken seriously enough by today's scholars. She makes a pitch for multifariousness, suggesting that readers’ capacity for identification transcends such genres as realism or even the idea of character per se, positing that readers identify across boundaries of gender, class, and even (in a gesture that points, here, toward Ivan Kreilkamp's Minor Creatures) species. One can even identify with qualities, such as “vividness and distinctness,” she claims, not citing Gombrich but referring to the minimal makeup of some cartoon characters (78). Characters are both mediated and mediators: they are made of fragments and scraps of culture, and also facilitate our relation to various aspects of it; they need not be unitary or uniform, nor need they be realistic. Felski's essay turns into an anatomy of different kinds of identification and a catalog of examples, with descriptions of how they work and what kinds of connections they can help readers make. Ranging from Emma Bovary to Thelma and Louise, she proposes a variety of kinds of identifications, giving them names: “alignment,” “allegiance,” “recognition,” and “empathy.” If you like classification, this essay (indeed, this book, as I discuss further below) is for you. I found the naming exercise to be somewhat unnecessary, since it recommends that critics think about identification in ways many are already doing—across class, gender, and genre boundaries—advising us as well to identify with a “sensibility” or style rather than a single aspect of a character or person, and suggesting the inclusion of popular culture in our thinking. They also seem, as classifications often do, to separate categories that are not necessarily or even usefully separated, as in distinguishing the evaluative aspect of character—identifying with what a character stands for—from the recognition aspect: seeing oneself reflected. Description and point of view—positioned under “alignment”—might play a role in other categories as well.

Amanda Anderson's essay is another example of doing things with terms, and she follows Felski in inventing new ones rather than trying to do something new with the old. Her contribution, in this case, is “rumination,” a choice that overlaps somewhat unfortunately with Felski's brief discussion of Madame Bovary (1857) and Kreilkamp's book on animals. Anderson argues for the importance of a new category of “interior moral reflection” (131) in realist characters; the term “rumination,” drawn from cognitive science, is meant to signify a slow, mental chewing over of morally complex, value-laden issues by characters over an extended period of time. She means to distinguish this kind of thinking from other kinds, such as the more general “free indirect discourse,” and so defines very specifically what she means by it; in particular, she wants to address a sustained wrestling with moral issues.

How much does it help to have a new term for a very specifically defined kind (or topic) of thinking—as, in the previous essay, for a series of narrowly defined aspects of identification? Does it help to use “rumination” to define thinking “with some sort of orienting or nagging focus,” which tends to be circular, repetitive, and restless? (After all, as literary critics, we respond aesthetically as well as analytically: do we really want the Bovaryesque comparison of persons to ruminants?) Is it the case that characters’ “processing” of “profound experiences,” or those experiences themselves, are necessarily moral (133)? If other characters, and Trollope's narrator, describe the Rev. Crawley as “brooding” (146; another animal reference!) why do we need another term, and is the new one more helpful than those in use (such as free indirect discourse or just plain “thinking”)? Anderson makes her case for such delineations—that is what this essay is—and includes illuminating readings of Eliot and Trollope in support of them. But “rumination” turns out to be different in literary use than in psychology, and as applied to Trollope's characters it requires further parsing. In cognitive psychology, Anderson reports, it already has its own “four-factor analysis,” including subsets such as “brooding” and “reflection” (135). For her analysis of Trollope, Anderson adds “second-order rumination”: rumination mediated by another character or a narrator (149). The key factor, however, seems to be obsessiveness: thinking that takes a circling, repetitious form and requires no resolution. The moral quality of the subject matter is Anderson's specific contribution: “rumination,” as she defines it, often stems from “an experience of moral shock or disturbance” (152). Despite the linking of two meanings of the term “character” on which Anderson relies (the personlike and the moral), however, the terms are not necessarily connected, except perhaps in the Victorian period's inconsistent usage (a character can, but need not be, assessed in moral terms). In general, this volume's production of new terms and classifications is at odds not only with Moi's critique of Knights's desire to professionalize the study of literature but also with the book's interest in reinvesting character-study with emotion and attachment. Following the increasingly fine distinctions into which categories are divided is tedious business and saps a reader's energy from attending to the often insightful readings that follow, which could proceed without them—as many fine analyses of thinking in Trollope have done.

Terms change over time; some prove useful for many years, while others outlive their usefulness fairly quickly. “Free indirect discourse” has had a long shelf life, as have metaphor and metonymy; “metalepsis” has been revivified by Elaine Freedgood; Bill Brown managed to redefine the word “thing,” giving it a theory. But the essays in Character express divergent impulses—in the direction of making critical discourse and the teaching of literature more student-friendly and feeling-oriented, on one hand, and of proliferating increasingly fine, not obviously necessary or helpful distinctions, on the other. The introduction of new terms and classifications is exactly the kind of academicization or “professionalization” that, Moi argues, distracted us from the pleasures of identification in the first place.

Any discipline needs a language of its own, and the analysis of fiction no less than poetry requires specialized terms and techniques—more specialized than “like” and “dislike,” though these initial responses may and often do constitute the beginning of more extended analysis. Indeed, what tends to be called the “turn to affect” was perhaps the beginning of a movement away from the kind of technical, text-oriented approach the volume critiques, representing a combination of the need for “discipline” and, frankly, for students—a need that can best be met, the implicit argument seems to be, in the realm of feeling. And so discussions of character might best begin, as perhaps they always have, with the rhetoric of likes and dislikes rather than with the rigors of a formalist approach. Fields such as affect studies and cognitive studies allow us to be disciplinary and to make friends with characters at the same time.

Characters are not persons, but persons can be characters; sometimes, indeed, a single person could be—or contain?—more than one. We may make ourselves into characters for the purpose of telling a story, temporarily simplifying our complex identities (as we may flatter ourselves they are) into a few outstanding traits. We may contain multitudes (or so it may seem to us), which can be deployed at the drop of a hat and are inseparable from our conception of our “selves.” Some of us may imagine that we possess an inner child, or think about ourselves as made up of id, ego, and superego; we might display a home-self when we are home, and a work-self when we are not. We may have gym selves and gourmet-chef selves; selves that stand in line and selves that refuse to; selves that are full of beneficent feeling and others that seethe with an underground-man type of resentment. Are these characters? Or do they, if we round them all up, constitute a single character? Or are we each more like a small crowd—or a large one? Perhaps the collective and the solitary are merely different ways of describing the same thing, their difference the result—to go back to my earlier discussion—of the terms we are accustomed to using: we “are” this or that, we “feel” this way or that way; we are divided; conflicted; of two minds or more. As Anderson asks, “What does talking to oneself mean?” (143). “Rumination,” or any kind of interior conversation, raises questions of identity. Just as all of Jane Eyre's characters may be understood as aspects or projections of their imaginary author, and just as everyone in your dream is you, so too is it possible to use “character” as a term that describes any of the ways in which we imagine ourselves at a given moment, both within literary texts and outside them.

If this is true, or at least acceptable as a working hypothesis, then it may be useful as a way of describing not only our relation to ourselves but also to those we choose to add to our retinues: I am referring here to pets, or domestic animals, specifically in the book Ivan Kreilkamp has devoted to animals in Victorian fiction. Kreilkamp claims that this is a “pet centered” or animal-centric book, and he frets that animals fail to register as characters: “They are diminished beings, subsidiary, not quite worthy of plots or more developed narratives” (23). Can a novelistic character, he wonders, be “anything that is alive?” (28), arguing in general for the centrality of the domestic animal to the Victorian novel. I am, I must confess, of the view that we can never know what “pet-centric” means, since they can never tell us (except in terms we ourselves imagine, as in children's literature, not discussed in Kreilkamp's book). Their situation is not comparable to that of slaves, servants, or other marginalized persons, as Kreilkamp suggests it is, because they have no voice, in our terms, to suppress—no existence outside their usefulness for humans: as domestic accoutrements, companions, service animals, hand warmers. Indeed, as Marc Shell has pointed out, the idea of extending human treatment to animals can exist alongside the treatment of some humans as animals (Hitler “anthropomorphized” his dog), and the idea of extending tolerance to some or all animals has its own distinct ideological and political justifications.Footnote ²

Absent this kind of historical context, or a coherent account of the kind of political or ethical thinking that undergirds his claims, Kreilkamp argues somewhat abstractly (but with a consistent attitude that an unfairness is being perpetrated) that animals should be given more space in critical analyses of Victorian fiction than they have yet received. For one thing, he suggests, such an approach would allow for the working out of ethical problems that bear on how human beings treat one another. The idea of domesticity, for instance (and, in an overstated extension of his argument, the idea of “fiction” or the Victorian novel per se), requires the presence of domestic animals, as creatures on which humans practice their will to dominate; they represent the precarity and vulnerability of the human, enabling novelists to consider the boundaries between various forms of life. This idea is reasonable, and leads to some plausible readings of texts. But it does not clarify the specificity within Victorian culture (or the sheer weirdness) of the “pet” idea, nor does Kreilkamp fully acknowledge that the boundaries between the animal and the human, and indeed these animals themselves, are cultural constructions, their very DNA manipulated by the humans who own them—a fact that troubles, at the very least, the book's argument for animal agency or “pet-centeredness.” Indeed, the idea of the domestic animal itself both invents and claims to transcend an imaginary and thoroughly ideological boundary. One could, interrogating this concept, sieve out what “domesticity” meant to the Victorians, according to what aspects of the “wild” had to be weeded out. The domestic dog, for instance, must be clean, well shaven, follow orders, and not attack (remember Copperfield: he bites!).

That the Victorians raised this manipulation to new heights—that they obsessively bred innumerable varieties of dogs for their own entertainment, for instance—tells us much about their collective character, at least that of the middle and rising middle classes; it says something about their desire to dominate creatures weaker than themselves, and to fashion those creatures in their own images—commonplace ideas when it comes to pets (and, for that matter, characters). But these issues could be approached in a less sentimental manner. In the centuries leading up to the nineteenth, and within the Victorian period itself, dogs go from being workers—sheepdogs or guard dogs—to commodities: objects for display, emotional attachment, and competition. Breeding enabled the inclusion of desirable traits and the weeding out of those deemed undesirable. “Purity” was valued, and pedigrees existed—as they did for persons—to document the lineage. This kind of thinking of course echoes contemporaneous views of the human race, the difference being the ease with which the DNA of animals could be altered.

What is perhaps most striking about the idea of the pet—or domestic animal, that is—is what it says about the Victorians’ Frankensteinian ability to bring their animal fantasies to life, and Kreilkamp's book is, and can only be, about those fantasies. He says as much himself: “What is the sheep, the cow, or the dog in its domestication but a projection or extension of human desires and needs?” (117). “The cries and prints of dogs, cows, horses, and ponies matter only insofar as they can be manipulated and read by human beings for human purposes” (142), he writes, but then follows this immediately with the claim that the hound of the Baskervilles’ wail, in Conan Doyle's story, “articulates the sorrow of the animal's failure to signify effectively” (143). Whose sorrow? Despite Kreilkamp's acknowledgment that all this is projection, that is, there is more than a residual sentimentality in his argument: in the idea, for instance, that there is some inherent tragedy in our inability to apprehend or correctly “read” animal subjectivity—a concept that is itself a human projection, relying on the notion that animals are versions of us. Even stranger is the notion that animals could or should be central characters in novels, their thoughts and feelings somehow accurately expressed—as if, outside of fantasyland, this were possible. We will never know the true feeling—and of course this is not the right term—of Emily Brontë's dog Keeper, which endured violence at her hands yet remained “loyal” (my quotation marks) to the end. Marveling at the animal's selflessness is an exercise in exactly the kind of anthropocentrism Kreilkamp appears, in the book's more clear-eyed moments, to want to overturn.

If a dog in Great Expectations (1861) is “a precarious or threatened identity . . . vulnerable to the precariousness that defines animal life” (76–77), that is because “dog” in this instance is a metaphor for something to kick. If Waldo, in Oliver Schreiner's Story of an African Farm (1883), says he feels “like an animal” (176), is he gesturing toward another mode of being, or merely toward a common cultural understanding of what the term “animal” signifies? Kreilkamp often strays into this mystified territory, mistaking human gestures toward animals as somehow boundary-crossing, when there is nothing on either side of the line that is not human fabrication. This book might be an occasion to consider the way in which animals in life and in fiction are bred—we might say invented, as in the classificatory terms discussed in the earlier portion of this essay—for the purpose of highlighting certain traits: physical, affective, or both. But even a trait, such as bravery or loyalty, is merely an aspect of “character” as the term is commonly understood, rather than forming the whole of it, suggesting domestic animals’ purpose as emotional props subject to the whims of their owners, as the story of Keeper, emptied of its sentimental content, illustrates.

Here's an animal-centric story: a Nova episode about the history of the dog speculates that the domestic version came into existence when, long ago, wolves cozied up to humans for easy access to food and protection. At one point, the show details an experiment that seeks to differentiate between a dog's response to food and a dog's response to love. I won't elaborate all the bells and whistles the experiment involves (and it does involve both), but you will not be surprised to learn that the experimenters concluded that the dogs responded more strongly to promises of affection (from them!). It seems pretty clear to me, contradicting everything I have said so far and taking the animal point of view, that these dogs knew very well where their next sausage was coming from, and that the difference between love and sausages was, to them, invisible. The most powerful illustration of “domestication” may be the mystification the idea itself introduces, suggesting that a cordon sanitaire around the home, or idea of the home, can be successful, and that we are fully in control of who or what enters, leaves, or belongs.

John Frow begins his book about character with a discussion of the term's ambivalence. Character is a textual construct but one that encourages the overlooking of its textuality; it is modeled on the form of the person—a reality—but is also imaginary. A person, he goes on to say, has a body, and that body too is both real and imaginary: “Its key coordinates are the boundaries between the inside and the outside; between the self and the other; between male and female; between the clean and the polluted; and between the living and the dead.”Footnote ³ Another name for “imaginary,” that is, might be “cultural”; while we can agree on the physical existence of the person or the body, disagreements about their meaning are inevitable. I think one of the best responses to the stubbornness that seems endemic to both sides of the character question (the textual and the affective) has been provided, once again, by Rosmarin, who reminds us of Gombrich's point that all thinking is classification and all classification is mistaken. Clouds can resemble faces; we can choose to see sonnets and dramatic monologues, different as they are, as forms of poetry; we can read a novel as a poem, or a monologue as if someone is speaking. “Anything can be taken for anything else if our purpose is sufficiently compelling and our standards of acceptance sufficiently tolerant,” she writes.Footnote ⁴ Trying out different modes of thinking with regard to characters, animals, genres, or things can be edifying, as we have all experienced in our work as critics, perhaps most of all when keeping in mind, as Rosmarin puts it, that as we “choose, correct, invent, or define,” the best we can achieve—and what enables us to practice criticism—are the “desired mistakes” that make up our own classifications, our own thinking.