Will Glovinsky: Antagonism and the Form of Self-Accusation

Do novels implicate themselves in cruelty? I want to suggest here that they can, that they have not always done so, and that the history of this development opens up an alternative approach to the questions Bruce Robbins raises in “On the Non-Representation of Atrocity,” his keynote contribution to the V21 special issue of boundary 2 online. Why, Robbins asks, despite widely available reports of brutal colonial violence, did nineteenth-century European novels generally fail to represent atrocity in a self-accusing light? And how do we square that failure with our conviction that the story of atrocity’s emergence as a self-applicable crime remains “untellable without the nineteenth century” (Robbins)? By posing such questions about the “long-term moral history of humankind,” Robbins provocatively builds on V21’s brief for the recovery of presentism as a scholarly and ethical good. But if, as the V21 manifesto contends, presentist criticism also encourages us to “explicitly pursue the politics of form,” then we might reframe Robbins’ inquiries to ask how self-accusation became not a represented content of narrative, but rather the structural effect of specific novelistic forms.

Robbins concludes his keynote by arguing that the absence of atrocity in nineteenth-century novels may be explained by the lack of a robust international public “capable of demanding or enforcing scrutiny of ourselves from outside.” In his response, Zachary Samalin proposes an alternative “genealogy of civilizational self-accusation,” one focused less on disinterested cosmopolitan publics and more on the “unwanted (but inescapable) identification with destructive processes” that lead to acts of atrocity. It is not, in other words, merely critical detachment from abusive power structures but also undesirable attachment to those structures that prepares the way for true self-accusation. How then, Samalin asks, “can we further specify and describe this negative structure of feeling in the novel”?

It seems to me that any such description will sooner or later have to address a key characterological innovation of the Victorian novel: the non-evil, richly psychologized antagonist whose punishment, while in one sense deserved, also reveals the structurally necessary cruelty that underwrites narrative and makes narrative itself non-innocent. For to accuse oneself of atrocity, one must cease to demonize the “other side”; the victims must transform, in one’s eyes, from villains (or rogues, “savages”) to simple antagonists. That distinction is quintessentially literary – a question of imagination and representation – yet it’s a difference, I would argue, that makes or breaks the potential for atrocity as self-accusation. And the realist novel is arguably where that distinction becomes clearest. We might even say that atrocity as self-accusation becomes formally thinkable at the moment when the realist novel forks away from its more melodramatic variants and develops a new sub-genre in which villainy as such has faded away, leaving in its place a more ethically complex figure.[1]

Such antagonists matter to the genealogy of self-accusation because atrocity is most readily recognizable when its victims are separated, if only conceptually, from any evildoing. This condition may not be strictly necessary (atrocities could be and were committed against Nazi SS troops) but as a matter of representation it is remarkable how the concept of atrocity becomes unstable when placed in proximity with anything that might be misconstrued as its justification. Consider how in Oliver Twist (1837) the horrible fate of Bill Sikes, who inadvertently hangs himself while fleeing a mob enraged “with the ecstasy of madmen,” does not read as the near lynching that it is because Sikes has murdered Nancy, and is a villain with whom we need not identify (Dickens 424). Thinking about how mid-century realist authors like Trollope and Eliot ratcheted up our unwanted identification with antagonists – or, as Fredric Jameson has recently argued, used them to persuade us that evil simply doesn’t exist (Jameson 115–17) – can illustrate how the literary embryo of self-accusation takes form as a thousand minute ambivalences, each of them produced through ethical doubts and hesitations in the narrative point of view we are implicitly asked to endorse.

Ambivalences such as: “Mr. Slope was not in all things a bad man. His motives, like those of most men, were mixed” (Trollope 120). Here Trollope opens up the unctuous, grasping antagonist of Barchester Towers (1857) to a measure of readerly identification that, partial and thoroughly unwanted, does not prevent us from relishing his eventual humiliation but does allow us to perceive that humiliation’s underlying cruelty. Slope is hypocritical, vain, and inimical to the novel’s marriage plot between Mr. Arabin and Eleanor Bold, yet he can still be treated unjustly by the novel’s ecclesiastic “good side,” the High Church Grantlys. Within that faction, Eleanor becomes Barsetshire’s exemplary self-accuser. “The abuse of other people,” Trollope writes, “and abuse of a nature that she felt to be unjust […] at last made Mr. Slope’s defence an habitual course of argument with her” (180). And to the extent that we, like the Grantlys, root against Slope – as indeed we must to harmonize with the plot’s social and romantic energies – we sense our implication in the novel’s subtle but unmistakable cruelty.

Hiroshima or Morant Bay it is not. But the satisfaction, and concomitant self-censure, occasioned by the chagrins of a maligned clergyman may well rehearse, at formal and affective levels, the more momentous kinds of self-accusation we tend to associate with twentieth-century novels. And Slope has company: Mrs. Proudie in the Barsetshire novels, Bulstrode and Casaubon in Middlemarch (1871-2), perhaps Hyacinth Kirkpatrick in Wives and Daughters (1864-6). Turning to the appearance of these generously psychologized antagonists to think about the longer history of imperial injustice and the novel might, as Nathan Hensley writes, “push us away from vestigially empiricist models of influence and toward an understanding of how literary presentation can enact thought.” Perhaps this is what it means to read empire formally: pausing in our search for represented content to register the rise of globally resonant modes of cruelty, self-accusation, and guilt in the ethical systems we call novels.

Columbia University


Works Cited


Dickens, Charles. Oliver Twist: Or, the Parish Boy’s Progress. Ed. Philip Horne. Penguin

Classics. New York: Penguin, 2002. Print.

Hensley, Nathan K. “Swinburne’s Oxford Notebook: Violence in/as Form.” boundary 2 online,

vol. 1, no. 2, Oct. 2016, www.boundary2.org/2016/10/nathan-k-hensley-swinburnes-oxford-notebook-violence-inas-form/. Accessed 7 Jan. 2017.

Jameson, Fredric. The Antinomies of Realism. London ; New York: Verso, 2013. Print.

Robbins, Bruce. “On the Non-Representation of Atrocity.” boundary 2 online, vol. 1, no. 2, Oct.

2016, www.boundary2.org/2016/10/bruce-robbins-on-the-non-representation-of-atrocity/.

Accessed 7 Jan. 2017.

Samalin, Zachary. “Genealogies of Self-Accusation.” boundary 2 online, vol. 1, no. 2, Oct. 2016,


Accessed 7 Jan. 2017.

Trollope, Anthony. Barchester Towers. Ed. Robin Gilmour. Penguin Classics. London: Penguin,

  1. Print.

V21. “Manifesto of the V21 Collective.” V21 Collective, 2015, v21collective.org/manifesto-of-

the-v21-collective-ten-theses/. Accessed 8 Jan. 2017.

Woloch, Alex. The One vs. the Many: Minor Characters and the Space of the Protagonist in the

Novel. Princeton, NJ: Princeton Univ. Press, 2003. Print.






[1] Alex Woloch’s The One vs. the Many notably avoids the terms “antagonist” and “villain” (preferring the more neutral “minor” and “eccentric”), but his influential examination of how the unequal distribution of character-space “weigh[s] on the consciousness of the novel” nonetheless informs my sense here of narrative’s non-innocence (Woloch 21).

%d bloggers like this: