Resisting Tiny Heroes: Kant on the Mechanism and Scope of Imaginative Resistance

Abstract

Traditionally, theorists suggested that imaginative resistance (e.g., when a reader does not imagine what a literary work entreats her to imagine) is limited to morally repugnant claims. More recently, theorists have argued that the phenomenon of imaginative resistance is wider in scope, extending to descriptive claims (e.g., those that are conceptually contradictory). On both sides, though, theorists have focused on cases where imaginative resistance goes right, tracking something that is wrong with the story—that it is morally repugnant, or conceptually contradictory. I use a rarely cited discussion from Kant to argue that imaginative resistance can also occur when something goes wrong with the reader—namely, when a reader imports their own biases into the story, and resists a descriptive claim as a result. In identifying this new class of claims that can meet imaginative resistance, Kant presses the question: when should we cultivate imaginative resistance and when should we fight it?

Consider the following three fictional “stories,” cooked up by philosophers working on the phenomenon of imaginative resistance:

1. “In killing her baby, Giselda did the right thing; after all, it was a girl.”

2. “They flopped down beneath the great maple. One more item to find, and yet the game seemed lost. Hang on, Sally said. It’s staring us in the face. This is a maple tree we’re under. She grabbed a five-fingered leaf. Here was the oval they needed! They ran off to claim their prize.”

3. “The hero strode across the battlefield with confidence and ease. He knew that he could conquer any country within a few days’ time. A tiny little fellow, he looked the opposing general squarely in the chin and demanded his surrender.”

The first story is provided by Kendall Walton (1994, 37), and it represents a classic case of imaginative resistance. While readers of fiction easily imagine farfetched claims, from talking animals to flying nuns, there are some claims that are more difficult to imagine; when this happens, readers experience imaginative resistance. As far back as David Hume, theorists traditionally have taken imaginative resistance to be a reaction to moral claims—specifically, when a narrator makes a claim at odds with a reader’s moral convictions. In the case of Giselda, the reader resists the claim that it was “right” for Giselda to kill her baby, because of the reader’s own sense of what is right.

More recently, theorists have begun to accept that a reader can imaginatively resist a broader class of claims. The second story, provided by Stephen Yablo (2002, 485), is representative of this broader class. In this case, the reader resists a descriptive claim that is conceptually incompatible with earlier parts of the story, namely that the leaf is “oval.” While clearly not a moral example, the second story, like the first, provides an example where imaginative resistance goes right. In both cases, imaginative resistance is warranted by something that is wrong with the story—that it is morally repugnant, or that it is conceptually contradictory.

The final story is cooked up by me, on the basis of a rarely cited discussion from Immanuel Kant’s Anthropology from a Pragmatic Point of View.¹ In what follows, I argue that it represents a kind of imaginative resistance that has been overlooked in the literature—a literature that has almost exclusively focused on success cases.² Kant points us to cases where a reader imaginatively resists a descriptive claim, not because there is something wrong with the text, but because the reader has imported their own biases into their reading of the story. The imaginative resistance occurs when the reader’s imported biases conflict with some part of the story. Thus, in the third vignette, above, a reader might take it for granted that a hero would be great in stature, subsequently resisting the claim that he is a “tiny little fellow.”

Though Kant has been largely left out of the literature on imaginative resistance,³ I argue that Kant’s brief comments on the phenomenon shed light on its scope. Imaginative resistance does not only occur when something goes wrong with the text; it can also occur when something goes wrong with the reader. Further, even as Kant’s comments point us to a new kind of imaginative resistance, expanding its scope, the theory of the imagination that undergirds those comments offers a unified account of imaginative resistance. Suggesting that the same capacity of the imagination is at work both when we imagine fictional scenes and when we perceive the world around us, Kant explains how slippage between the real world and the fictional world can produce multiple kinds of imaginative resistance. Kant’s account of the imagination, therefore, helps us to theorize why imaginative resistance happens in general, while also highlighting an important class of cases that has thus far been ignored in the literature.

To tease out Kant’s contribution to contemporary debates about imaginative resistance, I proceed as follows: First, I show that Kant’s theory of the imagination provides a broad explanation for why we imaginatively resist in any case. Against the backdrop of this generalized account of imaginative resistance, I then turn to Kant’s overlooked comments on the phenomenon, which suggest that we can imaginatively resist descriptive claims for contingent reasons—such as contingent biases on the part of the reader. I finally argue that, in identifying this class of cases, Kant shows us that imaginative resistance is not in every case a cognitive success. In so doing, he raises a question that has been overlooked in the literature: when should we cultivate imaginative resistance and when should we fight it?

I. THE PERCEIVING IMAGINATION

Tamar Gendler’s landmark paper on imaginative resistance theorizes that imaginative resistance occurs because of slippage, or “import-export,” between the fictional world and the actual world (Gendler 2000, 75). Her account appeals primarily to the direction of export—that is, to the impact that fictional imagining can have on our view of the real world. In particular, Gendler suggests that we will not imagine certain claims, because we do not want to export the idea we are being asked to imagine into our understanding of the world around us. For example, we do not want to imagine that Giselda’s female infanticide is right, because we do not want to carry such an idea with us into the real world. Gendler draws this account from the classic cases of imaginative resistance, where the resisted claim is morally repugnant.

A competing explanation suggests that imaginative resistance occurs not because the reader is unwilling to imagine certain claims, but rather because the reader unable to imagine those claims.⁴ This explanation appeals to the other direction of slippage: import. On this view, we resist certain claims when our real-world understanding intrudes on the fictional world (i.e., we import it); it is impossible to imagine these claims because of what we know about the real world. For example, based on what we know about ovals, we find it impossible to imagine that the five-fingered maple leaf is oval. This sort of explanation is typical among those who explore imaginative resistance to non-moral claims (specifically, to descriptive claims that are conceptually incompatible with other parts of the text).⁵

Emine Hande Tuna has argued recently that Kant offers a hybrid position on imaginative resistance, where sometimes we resist because we cannot imagine a claim and other times we resist because we will not imagine a claim. I agree that, for Kant, imaginative resistance can be generated both when we import something into the text (such that we cannot imagine some claim) and when we refuse to export something from the text (such that we will not imagine some claim). I argue, further, that while allowing for different kinds of imaginative resistance, Kant’s theory of the imagination nevertheless offers a unified account of the phenomenon. While some theorists have concluded that we must settle for an “explanatory cosmopolitanism” about imaginative resistance (Liao, Strohminger, and Sripada 2014, 353), Kant’s theory of the imagination identifies a single associative function at the bottom of every case of imaginative resistance. This associative function is at work both when we perceive the world around us and when we visualize fantastic scenes; it explains therefore how slippage between the real world and fictional world gives rise to multiple kinds of imaginative resistance, via both import and export.

Moreover, Kant’s theory of the imagination offers an expanded account of imaginative resistance. In particular, it offers a more capacious view of the “imports” that might generate imaginative resistance. It is not only our preexisting concepts but also our preexisting associations that might generate resistance. This gives the Kantian account an edge over rival theories, in that it can explain the class of cases that I highlight in Section II, where resistance stems from reader bias.

Let us turn, then, to the account of the imagination that will unlock these theoretical benefits. Kant defines the faculty of the imagination as the capacity to represent what is not there (B151; Anth. Ak. 7, 167). Kant’s imagination, therefore, does the ordinary work we would expect of it: the imagination is behind the fantastic sequences that we dream (B278), the fanciful images that we daydream (Anth. Ak. 7, 167), and the fictional scenes that we visualize when reading a literary work (Anth. Ak. 7, 173). Other voices in the history of philosophy—for instance, David Hume, a historical voice that is cited widely in the literature on imaginative resistance—likewise recognize these capacities in the imagination.⁶

Kant flags what he takes to be innovative about his account in a footnote in which he claims that the imagination, our capacity to represent what is not there, is “a necessary ingredient of perception itself” (A120). While Kant may overstate the innovativeness of his account,⁷ the details that he introduces to elaborate the perceptual contribution of the imagination explain why imaginative resistance occurs, in general. Further, as we will see in Section II, they set the stage for the innovation at the heart of this article: Kant’s suggestion that imaginative resistance can stem from contingent biases.

Specifically, Kant elaborates that one facet of the imagination—the reproductive capacity of the imagination—represents what we have previously perceived.⁸ In addition to reproducing a previous perception, the reproductive imagination puts this together with what we are currently perceiving—what Kant calls a “synthesis.”⁹ For example, the imagination might recall the curly-tailed rump I saw just a moment ago and connect it with the pink-snouted face I see now, helping me put together different perspectives on this very same thing. Kant suggests that, without a mechanism whereby the past can inform the present—a mechanism allowing us to represent the past in conjunction with the present—our experience would be one where we are assailed by new sensations at every given moment, unable to appreciate stable identities across the different moments of time (A100f.). But, of course, we do appreciate stable identities. Therefore, Kant surmises, the imagination is at work whenever we perceive (and not just when we fantasize). The imagination (specifically, in its reproductive capacity) is a thoroughgoing player in our uptake of the world around us.¹⁰

The reproductive capacity of the imagination is one that is reflexive, or, as Kant puts it, “blind” (A78). That is, it automatically associates previous states with the ones that we currently perceive.¹¹ In particular, we associate states that are coupled repeatedly in our experience.¹² Sometimes these associations rest on solid foundations; this previous state is another perspective on the same object, or this previous state is a causal antecedent of the present state. But associations do not always track reality. As Hume also recognized, associations can be coincidental. We might associate a past state with a present one, because they happened to coincide in our experience and not because they are really connected (e.g., causally).

Kant provides an account, then, where the capacity to reproduce past states and connect them with the present is an essential part of our perceptual experience. And yet, this same capacity has the potential to introduce misleading content into our experience, information that is not really connected to the situation we are in. On this view of our experience, then, it makes sense to be vigilant about our associations—for example, by trying to identify background associations, investigate associations we are unsure of, and, in some cases, unlearn ungrounded associations (i.e., associations that do not, in fact, track external realities). Unlearning may not be warranted in every single case. Sometimes ungrounded associations are harmless and may even contribute to a rich and colorful mental life. Associating a certain scent with childhood, or associating a song with a loved one, can be deeply meaningful. However, some ungrounded associations are more pernicious. For example, it is widely agreed that in contemporary American society there is a common association of Blackness with criminality, and even with maturity in age. These ungrounded associations contribute, for instance, to harsher punishments for Black children.¹³ It is associations of the latter sort, I argue, that are worth unlearning. I return to this point below.

Here, however, we can circle back to imaginative resistance. Kant can explain why the slippage undergirding imaginative resistance occurs in both directions because, on his view, the imagination plays a role both in constructing experience and in constructing fictional images. Specifically, slippage in the direction of “import” happens because associations that we have picked up from the real world inform how we imagine a fictional story. Alternatively, slippage in the direction of “export” happens because fiction also produces associations, which can in turn affect our uptake of the real world. I elaborate both directions, beginning with the direction of import.

Kant suggests that the reproductive imagination plays a necessary, ongoing role in our experience. The reproductive imagination will be required, then, even as we read literary works: for example, the reproductive imagination helps us appreciate when the plot thickens, by synthesizing new details with previous plot developments. However, this function of the imagination allows for associations built on real-world experience to influence our uptake of the fiction, potentially producing imaginative resistance.

As we saw, the reproductive capacity of the imagination both offers an image (i.e., the previous perception) and synthesizes the past with the present. The imagination behaves similarly with literature, though there is one crucial difference: while we continue to synthesize the (fictional) past and present, the images we connect are not strict reproductions. Kant suggests that a different capacity of the imagination is at work (i.e., another way of representing what is not present): a productive capacity that invents images.¹⁴ However, the reproductive capacity also contributes to the production of these invented images, for Kant insists that the images we invent stitch together empirical elements that we have previously experienced. For example, we cannot imagine a characteristic that is totally new, such as a color we have never seen before (Anth. Ak. 7, 167–8). Rather, even our most fantastic imaginings are combinative, for example, combining the torso of a horse and the horn of a narwhal to form a unicorn. The productive imagination produces the fiction-inspired images, but uses reproduced material to do so.

Associations will be in play, then, even as we construct the fictional scene: some components of the image will come explicitly from the fiction, whereas other details are filled in, brought on the scene because the reader associates them with those explicit components (based on her real-world experience with the sorts of items described in the fiction).¹⁵ When the author’s subsequent discussion conflicts with some component of the constructed image (whether its origin is an explicit textual specification or a reader association), imaginative resistance occurs.¹⁶ In particular, the reader experiences a barrier forming a coherent imaginative sequence. For example, in the maple leaf story, the reader is, of course, fully capable of imagining a five-pointed leaf, as well as imagining something that is oval. However, the reader is not able to form a coherent imaginative sequence, where those images capture one and the same object; their image of the oval is at odds with the image initially formed of the five-pointed leaf. The time slices are imaginable, but a coherent arc is not.

The account can also explain the other direction of imaginative resistance—where the reader imaginatively resists in order to avoid “exporting” what is said in the fiction to one’s real-world experience. Like Gendler, I suggest that literary works add to our “associative repertoire” (2006, 163; see also 150–51). If you were to imagine a princess, for example, it is likely that some of the images that come to mind are images that have come to be associated with princesses due to fictional texts, rather than real-world experiences. Remember here that associations are formed when two states have been coupled in our experience—and not necessarily because they are actually connected (i.e., they may be causally independent of one another). Such couplings can also be produced when we read fictional texts, such that we come to associate princesses with dragons. Or, to offer a more pernicious example, watching crime shows that frequently feature Black criminals might lead one, in real life, to associate a Black passerby with criminal activity—or even, to draw connections between perceptual content that is simply not there (supposing, e.g., that this Black passerby is the cause of those police sirens). In building associations, fiction can also affect our uptake of the real world. Therefore, it is reasonable for readers to resist imagining some sequence if it couples two elements that they would rather not associate (e.g., female infanticide and rightness).

As I mentioned, this account of export agrees with Gendler, who also recognizes that a reader might resist imagining a claim in order to avoid building an association. However, the Kantian account of imaginative resistance is broader than that of Gendler, since it allows for associations likewise to generate resistance in the direction of import, such that imported associations prevent a reader from forming a coherent imaginative sequence. Kant’s view of the imagination, where the same function is behind picturing fiction and perceiving the real world, explains and elaborates upon both directions of slippage.

II. RESISTING LITERARY DESCRIPTIONS

In addition to offering a general explanation for imaginative resistance, Kant’s account of the imagination broadens the scope of imaginative resistance, as will become clear when we examine his overlooked comments on the phenomenon. As I mentioned above, imaginative resistance was originally treated as a moral phenomenon (call this the “moralizing” position): we resist morally repugnant claims (e.g., the Giselda story), cooperating with descriptive claims even if impossible or farfetched.¹⁷ However, recent theorists have identified descriptive claims that also garner resistance (e.g., the maple leaf story), suggesting that imaginative resistance is broader in scope (call this the “generalizing” position). Kant concurs with the generalizers, offering an example where a reader resists a descriptive claim. But further, Kant’s example broadens the class of descriptive claims that can meet resistance. Rather than suggesting that the reader resists descriptive claims that are conceptually incompatible with other parts of the text, Kant’s reader resists a claim due to contingent associations that they bring to the text.

Kant offers his case of imaginative resistance (or so I argue) in the Anthropology. Here is the passage in full:

Concepts of objects often prompt a spontaneously produced image (through the productive power of the imagination), which we attach to them involuntarily. When we read or have someone tell us about the life and deeds of a great man according to talent, merit, or rank, we are usually led to give him a considerable stature in our imagination; on the other hand, when someone is described as delicate and soft in character we usually form an image of him as smallish and pliable. Not only the peasant but also one fairly acquainted with the world finds it very strange when the hero, whom he had imagined according to deeds narrated of him, is presented to him as a tiny little fellow, and, conversely, when the delicate and soft Hume is presented to him as a husky man. (2006, 283, Ak. 7, 173)

In this passage, Kant describes a wide phenomenon: one that can happen to the peasant and the worldly imaginer alike, either when they read or hear about someone, who is either fictional (a hero) or non-fictional (Hume). Namely, something about the narrative—whether the deeds or personality attributed to the character—“prompt[s] a spontaneously produced image.” In the case of the hero, for example, one involuntarily or reflexively forms an image of him as “considerable” in stature due to the great deeds attributed to him. Subsequently, the description of him as a “tiny little fellow” comes as a surprise. This provides a case of imaginative resistance where the reader resists a descriptive claim (i.e., that the hero is “tiny”).

Further, the reader resists that claim for reasons that are contingent, rather than due to a conceptual incompatibility—broadening the generalizing position. The concept of the great man does not provoke an image of considerable stature due to a necessary relationship between great deeds and great stature (consider Kant’s own height!). Rather, the reader associates great deeds with considerable stature, likely based on past literary examples that reinforced this connection. For this reason, the imagination reflexively produces an image of the hero who is considerable in stature. It is this image that produces resistance when a contrary image (“tiny little fellow”) is described. Like the maple leaf story, the reader is unable to form a coherent imaginative sequence—one connecting the image of a strapping hero and the image of a slight hero as images of one and the same person. However, this inability stems not from explicit details in the text but from the contingent associations that informed the reader’s imagining of the text.

Kant, therefore, points us to an underappreciated kind of imaginative resistance: resistance to descriptive claims for contingent reasons (specifically, reader bias). This kind of imaginative resistance, as I put it above, issues from a problem with the reader rather than a problem with the text; the reader has generated a conflict, rather than successfully identifying one in the text. What should we make of such cases? Should it matter that there are bias-driven cases of imaginative resistance, and that the literature has apparently overlooked them?

One might suspect that the scholarship’s omission of such cases is not an oversight but a practical move: the reason that success cases (stemming from a problem in the text) are more prominent in the literature is because these are the most clear-cut cases of imaginative resistance.¹⁸ Such cases appeal to a variety of different readers (who can be relied upon to oppose female infanticide, or to know that ovals cannot have five fingers), convincing them that imaginative resistance is a phenomenon. Further, in their global appeal, such cases allow theorists to avoid digging into the complexities that might trigger idiosyncratic cases of imaginative resistance (e.g., individual readers’ biases); they can cut to the chase, so to speak, homing in on imaginative resistance proper.¹⁹ It might not surprise theorists of imaginative resistance that bias-driven cases occur, though they have left such cases aside for practical reasons.

However, I argue that leaving aside the bias-driven cases does a disservice to our theories of imaginative resistance. Focusing on success cases limits our explanation for why readers have difficulty imagining some fictional claims. One develops a theory that explains imaginative resistance in those cases where something goes wrong in the text, but not in those cases where something goes wrong with the reader. Attending to the wider scope of imaginative resistance, by contrast, lends more explanatory power to our theories. We can see this by returning to the standard explanations for imaginative resistance that I canvassed in Section I. Standard import-based explanations fail to explain why the reader cannot imagine the “tiny little fellow” claim, and standard export-based explanations fail to explain why the reader would prefer not to imagine it.

First, consider the accounts of import that are popular among generalizers, where imported real-world understanding renders a reader unable to imagine a wide class of claims (moral and descriptive). On these accounts, readers resist when they spot a conflict between different parts of the text. Brian Weatherson, for example, suggests that imaginative resistance occurs when a reader spots a conflict between lower-level facts (“basic phenomenal and functional facts” (2004, 18)) and some higher-level claim (a claim that is made true by certain lower-level facts).²⁰ Their real-world understanding of that higher-level claim (e.g., of what it takes for something to count as an “oval”) allows them to recognize the conflict, generating resistance to the claim.²¹

Weatherson’s account does not explain why imaginative resistance arises in Kant’s case, where there is not a conflict between a lower-level fact and a higher-level claim. As Kant describes it, the conflict is restricted to the lower levels: the reader’s image of the hero’s physical stature (provoked by on-the-ground facts about his deeds) conflicts with a textual claim about that stature. Now, one might suspect that Kant’s case nonetheless relies on higher-level claims: that the character is (or could be described as) a hero, whose deeds are (or could be described as) great. But even if such claims play a role in generating reader resistance, it is not because they conflict with lower-level claims in the text. Heroes can be tiny; great deeds do not require a great stature. Weatherson offers a reasonable account for cases of imaginative resistance that result from conflicting textual claims, as it seems unlikely that an author would pen a lower-level conflict (where, e.g., a leaf has five fingers in one sentence and no fingers in the next). However, as we see in Kant’s case, reader bias can create conflicts that are restricted to the lower levels. Weatherson’s account is tailored to cases of imaginative resistance where something goes wrong with the text; it does not adequately explain cases where something goes wrong with the reader.

My Kantian theory is more capacious. My account specifies the type of conflict that generates resistance in the direction of import: one fails to form a coherent imaginative sequence, due to a conflict between two (or more) imagined time slices. However, my account underdetermines the source of that conflict. Textual details (at whatever level) and reader associations both contribute to our imagining of fictional scenes, and both sources can contribute to a failure to form a coherent imaginative sequence. This failure could arise from textual claims alone, as with the maple leaf story; or, it could arise from the interaction between textual claims and reader associations, as with Kant-style cases.

Kant’s discussion of the tiny hero offers an explanation for resisting descriptive claims that resembles a standard explanation for resisting moral claims: the resistance arises because of something about the reader’s own perspective (e.g., a belief that female infanticide is immoral), rather than a contradiction on the face of the text. Some moralizers have likewise suggested that these moral convictions can be contingent, relative to one’s culture—“alter[ing] from one age to another” (Hume 1757, 236). Do moralizers, then, provide a better explanation for Kant-style cases of imaginative resistance?

Gendler suggests that a reader imaginatively resists when they are unwilling to export some perspective into their uptake of the real world. Specifically, readers resist pop-out passages—passages that, in the readers’ view, ask them not only to imagine that something is true of the story, but also to believe that it is true “in this world” (2006, 160). These are passages that make normative appraisals (e.g., that something is “right”), appraisals that we would also apply to real-world phenomena. When readers resist a pop-out passage, on Gendler’s account, it is because they would prefer not to export what that claim entreats them to believe about the actual world (e.g., that “right” applies to cases of female infanticide). The reader resists when they spot an inconsistency not between two textual claims, but between what the text asks them to believe and their own normative commitments.

This moralizing explanation does not explain Kant-style cases. While the reader’s perspective (specifically, their broader associations) surely drives such cases of imaginative resistance, I think that this is not immediately apparent to the reader, who feels surprise at the claim that the hero is tiny, rather than, say, embarrassment or self-reproach (“I’ve done it again; why must I always assume that heroes are strapping?”). Associations operate in the background, affecting our uptake of the text but sometimes without our explicit awareness. Therefore, it is not obvious that the reader would experience the “tiny little fellow” description as a pop-out passage that has broader implications for her uptake of the real world. The stakes seem limited, I think, to what is true in the fictional world: is the hero actually tiny? But further, insofar as this claim offers the reader a general lesson, it is that heroes can be tiny. And it is unclear why the reader would prefer to resist that lesson—to block this particular export. After all, the sentence challenges the readers’ uninvestigated, unendorsed associations, rather than, say, their closely held moral beliefs.²² The moralizing account fails to explain convincingly why the reader imaginatively resists in Kant’s case.

Both generalizers and moralizers fail to recognize that the reader can make a more robust contribution to imaginative resistance. Imaginative resistance does not only occur when a reader spots some conceptual or moral problem with the text; resistance also occurs when a reader imports their own associations into their reading of the story, themselves creating the conflict that generates resistance. While such cases elude the standard explanations for imaginative resistance, my Kantian account of imaginative resistance explains them: on this view, a reader imagines a scene based on author specifications and their own imported associations, and can themselves generate resistance in those cases when their imported details conflict with a subsequent claim in the text.

I have argued that my Kantian account of imaginative resistance is broader than rival theories; therefore, it has more explanatory power than those theories, explaining imaginative resistance to descriptive claims due to reader bias. However, this argument fails if the Kant-style cases that I have been discussing are not genuine cases of imaginative resistance. Indeed, one might object that Kant’s discussion exemplifies not imaginative resistance but rather a milder phenomenon, say, hermeneutic recalibration.²³ Imaginative resistance proper is persistent, whereas in hermeneutic recalibration, the reader changes their understanding of the story following their initial surprise; for example, they recalibrate their image of the hero, so they no longer resists the claim that he is tiny. To answer this objection, I explain why resistance can persist in some Kant-style cases.

I argue that this sort of resistance is, for the reader, in some cases indistinguishable from resisting a claim due to a contradiction on the face of the text. That is, the reader can (mis)take there to be an incompatibility in the details of the text because of a conflict between their own image of the scene and a new textual claim. On the Kantian account of imagining fiction, we form images that both respond to the constraints of the story, and exceed those constraints (as when filling in details). However, the components of those formed images do not betray their lineage; the image is not stamped, say, with the information that this component was dictated by the text, while that component was fabricated by the reader.²⁴ Based on this image, then, it is easy to confuse what the author has specified with what the reader has added—and to think the author has erred when saying something incompatible with the reader’s own supplements. And it is also easy to make this mistake, I suggest, because the reader does not take ownership over the association that involuntarily provoked the image; the reader cannot find some explicit belief (e.g., an endorsed moral conviction) that explains why the reader added this component. In Kant’s example, the reader might wonder: “Didn’t the author mention earlier that the hero was quite tall?” or “Wasn’t a considerable stature required for those heroic exploits?” If the reader indeed (mis)takes the author to have made conceptually incompatible claims, then resistance due to associations is in this case equally as persistent as in recognized cases of imaginative resistance, like resistance to the maple leaf story.

To be sure, resistance of this magnitude is not explicitly present in Kant’s own example. Kant does not suggest that the reader doubles down on their imaginative resistance, insisting that the text contains a contradiction; he does not develop the example further than describing the reader’s initial surprise. A final example of imaginative resistance of the kind Kant analyzes, however, will bring out the purchase and significance of his account. In 2012, some Hunger Games readers objected to the film adaptation of the book, because the filmmaker cast a Black actress as Rue, an innocent child who befriends the heroine. These readers objected that the casting choice did not cohere with their mental image of Rue. One reader tweeted, for example, “Awkward moment when Rue is some black girl and not the little blonde innocent girl you picture” (Holmes, 2012). Even beyond this initial “awkwardness,” several objectors insisted that the portrayal of Rue in the film was factually inaccurate—that the book had described her as white and blond.²⁵ There is no such description in the book; in fact, it attributes “dark brown skin” to Rue.²⁶ Rather, these readers had formed an image of Rue as white and blond based on their own association of youth and innocence with whiteness.²⁷ They resisted the film’s portrayal as a result, maintaining that the filmmaker had made a mistake.

These Hunger Games readers made precisely the mistake described above, prolonging their resistance to the film portrayal of Rue. They did not recognize their bias as the reason that they imagined Rue as white and blond, but rather supposed that these features of their mental image of Rue must have come from the author’s specifications. Therefore, in some Kant-style cases—where a reader resists a descriptive claim due to contingent associations—imaginative resistance persists, even beyond a reader’s initial surprise.

III. RESISTING IMAGINATIVE RESISTANCE

In addition to providing a case of especially persistent resistance due to association, the Hunger Games example raises a question that has gone unappreciated in the literature. Both generalizers and moralizers suggest that something goes right when a reader imaginatively resists, whether it is identifying a contradiction on the face of the text, or a claim at odds with a reader’s deeply held moral convictions (it is improper to “enter into such [vicious] sentiments,” says Hume).²⁸ Yet, the Hunger Games example provides a compelling illustration that imaginative resistance is not always a success. Therefore, it presses the question I articulated at the outset of this article: when should we cultivate imaginative resistance, and when should we fight it?

Gendler suggests that imaginative resistance is a success or achievement in that it blocks the exportation of morally repugnant ideas. For example, imaginatively resisting the Giselda case prevents the reader from exporting the idea that it is right to kill female infants. Similarly, a reader might avoid exporting racial biases by imaginatively resisting an allegory where the narrator claims that “the mice who had white fur were hardworking and industrious, but the mice who had black fur were slothful and shiftless” (Gendler 2000, 73). In these cases, the reader does not export the repugnant views expressed in the literature, such that they affect the reader’s uptake of the real world. Here, imaginative resistance insulates the reader from moral corruption, benefitting a reader’s moral outlook. Gendler, therefore, offers a class of cases where imaginative resistance would seem to be worth cultivating: cases where imaginative resistance is a moral success.

However, recent defenses of immoralism challenge the claim that imaginatively resisting morally repugnant claims carries such moral benefits. Immoralists hold that morally defective works can nonetheless have positive effects—including, on some versions of immoralism, positive moral effects.²⁹ On Daniel Jacobson’s version of immoralism, for instance, imagining morally deviant claims helps us better to understand the moral flaws we encounter in the real world, and makes us better positioned to converse with and reform morally flawed people (Jacobson 1997, 193). For example, imagining (rather than resisting) the mice allegory might help a reader understand racial biases, and be better positioned to confront and challenge such biases upon encountering them in the real world. If immoral art can better our moral outlook, as immoralists suggest, then it is difficult to see how resisting such works would amount to a success. The immoralist position, then, pushes us to ask not when imaginative resistance is a moral success but if it is ever a moral success.

While Kant’s discussion of the tiny hero points us to a class of cases that are definitively not a success, I suggest that his more general account of imaginative association helps to answer the immoralist challenge. Sometimes imaginative resistance is a moral success, and Kant can help us to see when, and why. Recall that, for Kant, associations are formed when two states frequently coincide in our experience. I suggested, further, that literature can produce such associations. If this is so, then exposure to stories like the mice allegory—or, indeed, crime procedurals that frequently feature Black criminals—can poison one’s associative repertoire. That is, these stories can build harmful associations that one brings into the real world, such that encountering a Black person might automatically call to mind ideas of slothfulness or criminality. Further, I take it that such associations would form even if one had a critical moment after consuming such works, where one recognizes that the show reinforces harmful stereotypes. On the Kantian account, associations function reflexively; associated elements spring to mind independently of our considered judgments. On the Kantian account, then, there is reason to cultivate resistance to works that build harmful associations.

However, I would like to make two caveats here that, indeed, might do some justice to the immoralist position that immoral artworks can produce moral benefits. First, because associations are formed by frequent exposure, one could engage with immoral works rarely and reap some of the benefits that the immoralist identifies. However, frequent consumption of immoral works poisons one’s associative repertoire, by building harmful associations. Second, consuming immoral artworks alongside works that challenge harmful associations may temper the harmful associations that would form if one consumed immoral artworks alone. That is, a more varied artistic diet would ensure that one does not always encounter the two states in tandem (as in the immoral works), such that an association is formed. Ultimately, Kant’s account of imaginative association speaks against consuming immoral artworks frequently and exclusively. It does not recommend that we cultivate resistance to any one work, but rather that we cultivate resistance to frequent and exclusive imagining of harmful associations.

Conversely, when should we fight imaginative resistance? The Kant-style cases on which I have focused offer a clear instance where imaginative resistance is a failure—worth fighting, if at all possible. In these cases, a reader imports their own biases into the story, which turn out to conflict with it. These cases represent a failure, first, in that the reader does not know what is going on in the world of the story, for they are mistaken about some component of the story (thinking the hero is strapping, or that Rue is white and blond). But this local failure, I think, contributes to a broader failure: by imaginatively resisting the descriptive claim that challenges ungrounded associations, the reader maintains that association. The reader has an opportunity to better their outlook, but does not take that opportunity. If imagining some claim can better a reader’s outlook,³⁰ then they should fight their resistance to that claim.

While Kant’s example of imaginative resistance illustrates when imaginative resistance should be combatted, his broader account of the imagination tells us how we might do so. I make my way to this Kantian strategy by first examining a different set of strategies for overcoming imaginative resistance, offered by Daniel Altshuler and Emar Maier (2020; manuscript). On their account, when a “face value” reading of the text fails—that is, when it is difficult to imagine that its claims state straightforward truths about the fictional world—a reader can adopt two different strategies. First, a reader might reread a claim that they initially resisted as indirectly reporting the (mistaken) speech of a character, where “Here was the oval they needed,” for example, reports what Sally said during the scavenger hunt. Second, a reader might take the narrator to be an unreliable narrator with mistaken views about the fictional world—a character who may not have been named yet—rather than an omniscient, impersonal narrator, as the reader initially supposed. Both strategies reinterpret the claim as issuing from the (flawed) outlook of some character, making it more imaginable as a result. It is easier to imagine that a character is mistaken about the shape of the five-fingered leaf than it is to imagine that such a leaf is oval.

While these are reasonable strategies for overcoming some cases of imaginative resistance, they miss the mark when it comes to Kant-style cases. Consider why classifying the “oval” claim as indirect speech looks like a reasonable strategy for overcoming imaginative resistance in the maple leaf story. As Altshuler and Maier argue, one overcomes resistance in this case by developing a deeper appreciation for the story—noticing that Sally’s speech had previously been reported (“Hang on, Sally said”), and that the indexicals in the resisted claim (“Here was the oval they needed”) signal a speaker. One overcomes the resistance by forming a coherent narrative arc that is attuned to the finer details of the story.

Now consider the tiny hero story. On Altshuler and Maier’s strategies, a reader might overcome resistance to this story by taking a speaking character or an unreliable narrator to have mistaken beliefs about the hero’s stature. While this reinterpretation might very well allow the reader to form a coherent narrative arc, they remain deeply mistaken about the world of the fiction; the reader maintains that the hero is strapping, and now they also think there is some character who is confused about his stature. This strategy does not take the reader closer to the finer details of the story, but further away from them. And that is because the strategy misdiagnoses the source of the mistake. There is not a contradiction on the face of the text; there is not a character who is mistaken. It is the reader who has made a mistake, generating an inconsistency by importing their own biases into the story.

This sort of imaginative resistance calls for a different, and earlier, coping strategy. Rather than reinterpreting because a “face value” reading fails, the reader must first investigate whether the face value reading in fact fails. Similar to Altshuler and Maier’s strategies, I suggest the reader must carefully reread the text—but, here, they need not comb the resisted claim for signs of indirect speech or a personal narrator, but rather go to the initial passages with which the subsequent claim was supposed to conflict. Was the hero in fact described as strapping—Rue, as white and blond? By employing this strategy, the reader is able to resolve the conflict by realizing there is no conflict, at least not on the face of the text.

Beyond the relatively low stakes result of correctly understanding the story, the reader employing this strategy also unearths the association that generated the resistance. For if the conflict did not come from the text, it is something that the reader generated themself. The reader assumed that the hero was strapping, or that the innocent child was white. Rather than operating in the background, the association is unearthed, where it can become the target of critical scrutiny. Further, in resolving the (supposed) conflict, one has an experience that challenges one’s association—an experience that decouples the associated elements (e.g., great deeds and considerable stature), experiencing one in the absence of the other. Correcting their reading of the story, the reader chips away at ungrounded associations.

Indeed, Kant’s account of imaginative association offers a broader strategy for overcoming bias-driven resistance, which targets the biases generating that resistance. If associations are formed by the frequent coupling of two states in one’s experience, it stands to reason that they will be unlearned by seeking out experiences that decouple previous associations. For example, instead of watching crime shows that couple Blackness with criminality, a reader might seek out first-personal accounts from Black authors, providing a richer view of Black experiences. Surprisingly, overcoming bias-driven resistance might mean cultivating imaginative resistance in some cases (refusing to engage with the crime procedurals), even as one attempts to overcome it in others (engaging with artworks that challenge the ungrounded association).

We have seen the theoretical importance of recognizing Kant-style cases of imaginative resistance, where a reader’s associations generate their resistance to some textual claim. Namely, attending to such cases yields broader and better accounts of imaginative resistance. But attending to the wider scope of imaginative resistance also has normative importance. Focusing on success cases prevents us from seeing that, when, and how one should combat imaginative resistance. Conversely, attending to those cases where imaginative resistance is a failure provides a path not only to combatting to imaginative resistance, but also to challenging the associations that can taint our uptake of fiction and reality alike.³¹

Footnotes

REFERENCES