Early in Zola’s Germinal, a young horse is lowered into a mine.footnote1 Tied beneath the cage that carries the workers down and sent into a near bottomless deep, it takes fright and nearly dies; it will not get up from the floor. Another working horse comes over, sniffs the terrified creature and then whinnies. After a short time the young horse is made to stand up (with a flick of the whip) and the animals are led off together.

Are these horses characters and, if so, how do they become so? John Frow would argue that they are. Not because a horse in a fiction is a priori a character, though animal-characters are hardly a novelty, but because of how they are represented in linguistic and narrative terms. Both animals have proper names (Bataille and Trompette), a sex and the pronoun that indicates this. The miners who are watching treat the equine encounter as social—‘Ah, old Bataille is a one . . . Look at him talking to his mate’—and it seems the reader should follow their lead. The narrator, in the same spirit, describes Bataille as ‘crafty’ and ‘wistful’; and Trompette’s journey into the mine repeats the pattern of a human narrative: the young man Etienne’s first day at the mine, which also includes a terrifying journey in the cage and then the camaraderie and reassurance of other workers he meets. Bataille seems to be offering Trompette equine solidarity.

In the plainly titled Character and Person, John Frow argues that characters have to be understood as both textual (and intertextual) entities and as quasi-persons with whom we, as readers and viewers, may identify, have affective relationships and speculate about ‘as if’ they were persons. The experience of this type of relationship is not, after all, the province of what Viktor Shklovsky called ‘ontological naivety’. Character has to be understood in relation to person, a locution he chooses as less discursively loaded than ‘self’ or ‘individual’, and a concept he aims to complicate. Exclusively textual ideas of character (particularly of the structuralist and post-structuralist varieties) do not include readers’ own role in character-making, how we can and do develop complex affective relations with characters. On the other hand, what he terms ‘ethical’ accounts of character, which encourage readers to think of characters primarily as people, refuse the full textual and intertextual specificity of character, and ignore its complexity as well as that of person itself. Although the narrowly textual account is not inherently mechanical and reductive, there is limited interest in describing Gwendolen Harleth in Daniel Deronda as an obstacle or false quest in Daniel’s own Bildung. But it is equally uninteresting to spend too much time wondering why Gwendolen continues her engagement with Malinger Grandcourt after discovering the long-standing relationship he has had with Lydia Glasher. Both approaches inhibit more interesting questions. In the latter case, the answers we come up with tend either to universal tautology—that’s the kind of person she is; she’s selfish; she feels trapped by her family—or to trite historical generalization, probably, in this case, concerning the plight of women in the later 19th century.

Character and person are ‘ontologically discontinuous and logically interdependent’, Frow argues, and their match and mismatch are central to the book’s argument and practice. Both character and person have various modalities of being: a particular character in a play or novel or film exists in and through genre and other social typologies, for example, and through various bodily scripts and schemas that endow some of their gestures and movements with meaning and not others. We use many of the same knowledges and strategies to identify and understand both persons and characters: knowledge of social and psychological types, various forms of language competence that enable us to understand—say—the peculiar properties of proper names and the shifting reference of pronouns. If we focus on the relations between character and person and no longer take character as given or personhood as self-evident, Frow hopes, different questions will suggest themselves. How do readers come to see or recognize characters in the first place? How and why do we grant quasi-personhood to certain kinds of textual entities but not others? What is the reader’s role in making character?

Character is ‘the central category in literary theory and its least adequately understood’, Frow argued, as long ago as 1986, developing an idea of how psychoanalytic accounts of audience identification in film might apply to literary texts. That, he says, was ‘the germ from which [Character and Person] sprouted’, and the book can be seen as fusing two of his distinctive, longstanding commitments: theorizing the processes of reading and interpretation and developing rigorous formal concepts that can have valency across disciplinary fields. His book, Genre (2006) can be viewed as a companion text in this sense. Frow describes himself as working at the boundaries of literary and cultural studies. He has defended the place of literary objects and practices within cultural studies and has, in his own work, focused on the production and circulation of cultural value across various fields. Although examples from film and video gaming feature here, his main focus is literature and, above all, the novel.